Tesla, Mobileye, Alibaba, and Didi have denied speculation that they have been involved in conducting illegal map-making activities after China’s top security agency said it found an incident of spying in which a foreign enterprise used a local company to collect sensitive map data.

The unidentified international firm illegally conducted geographic mapping activities in the name of implementing autonomous driving research projects, China’s state security ministry said on Wednesday in a post on the popular social media platform WeChat. It was assisted through outsourcing the activities to a Chinese company that has a government license for geographic surveying and mapping, according to the post. It added that the Chinese government had pursued legal action against the entities over the incident.

Tesla responded to the post by Beijing with Grace Tao, vice president in charge of external relations in China, reposting the article on the Chinese microblogging platform Weibo, with the comment: ”Compliance is the bottom line for business operations” (our translation).

Mobileye, which has partnered with Chinese automakers including Geely-owned Zeekr and state-owned FAW Group, followed suit immediately. “We have continuously invested in data compliance and regulatory adherence, including in China, where we operate under the supervision of licensed entities in full accordance with legal requirements,” Intel’s self-driving car unit said in a statement published Thursday on Weibo.

On the same day, China’s Alibaba Cloud and Didi also said they had nothing to do with the incident. Alibaba’s cloud arm terminated its mapping permit three days ahead of the release of the case, according to a statement by the Department of Natural Resources of Zhejiang Province, while Didi has been testing autonomous vehicles in a number of major Chinese cities. The two companies were also among the targets of Beijing’s regulatory crackdown against the country’s tech sector during 2020 and 2022.

China has for years required automakers and self-driving car developers to secure licenses or use authorized companies for mapping data used in vehicles. Some electric vehicle makers and tech companies, such as Xpeng Motors and Huawei, have pivoted to make assisted driving technologies without the assistance of high-definition maps for cost-saving reasons and because of regulatory issues.

Guangzhou Automobile Group (GAC), a long-time manufacturing partner of Toyota and Honda in China, has hired a veteran marketing executive from BMW as the chief operating officer of its overseas business unit, as it seeks to accelerate growth in Europe and beyond.

Why it matters: Making the announcement at this year’s Paris Motor Show on Monday, GAC joins an array of Chinese automakers tapping Europe amid ever-intensifying competition at home and following the imposition of nearly 40% tariffs on Chinese EVs by the European Commission.

READ MORE:  EU anti-subsidy EV probe: What Chinese automakers have done in Europe and what’s next

Details: Thomas Schemera, a longtime executive at BMW and Hyundai, will join GAC International, the overseas business subsidiary of the state-owned automaker, as the global chief operating officer and senior vice president. GAC aims to establish positive consumer resonance in the European market with this move and inject new impetus into the company’s global development, according to a Chinese statement.

• German-born Schemera held various roles at BMW over three decades. He started his career at the German carmaker in the late 1980s as an engineer of computer-aided manufacturing and briefly led the development of sales and dealerships for the manufacturer in China between 2005 and 2008, according to his LinkedIn profile.
• After his departure from BMW, he served as global chief marketing officer and executive vice president at Hyundai from 2021 to 2023, a role that included bolstering communication of Hyundai’s N brand and its dedicated battery electric vehicle lineup IONIQ, according to a Hyundai announcement.
• GAC will begin sales of a sports utility vehicle, Aion V, in several European markets in mid-2025. It has a driving range of 520 kilometers (323 miles) per charge and will be priced at no more than €40,000 ($43,748). To compare, a Tesla Model Y starts at €42,900 with a driving range of 244 miles in Germany.
• Meanwhile, the Chinese automaker is looking at the possibility of making EVs locally to avoid EU tariffs, an executive told Reuters. Still, speaking to Chinese media outlet Caixin on Tuesday, Feng Xingya, the general manager of GAC, explained that the company’s pursuit of local production would be premised on strong demand.
• GAC’s developments show its serious intention to dig deep roots in Europe, expanding its dealership and service networks with a timeline to sell cars and provide maintenance repair all over the continent in 2028. A transit warehouse for Europe will come into service next year to ensure a more timely supply of car components.

Context: GAC is among several Chinese automakers to steal the show in Paris with debuts of their price-competitive and technology-advanced EVs, and announcements displaying a continued ambition to crack the market despite rising protectionism.

• Stellantis-backed Leapmotor showcased its B10 compact SUV, a global model designed for tech-savvy consumers with assisted driving and infotainment functions powered by Qualcomm’s latest processors. It began taking reservations for a budget EV and a large SUV at sticker prices of €18,900 and €36,400 last month.
• At a press conference during the show, Xpeng Motors announced the pre-sale price for its P7+ sports sedan starts at RMB 209,800 ($29,493) in China. Meanwhile, BYD and Huawei-backed Seres showcased their newest models – the Yangwang U8, and the Aito M7, –  respectively.
• Chinese EV imports will face additional duties in the 17-35.3% range on top of an existing 10% levy likely starting from November. Beijing is still pushing for negotiations on alternative solutions, while Brussels is poised to publish tariff implementation  regulations by Oct. 30.

READ MORE: Chinese carmakers to become dominant globally despite tariffs – AlixPartners

China’s GAC Group said on Oct. 10 it planned to invest $27 million in self-driving car startup Pony.ai through its wholly-owned subsidiary GAC Capital, a move expected to “create synergies” between the two companies and speed up the development of GAC’s automated driving functions. The investment marks the latest bet by a Chinese automaker to boost their autonomous driving capacity in the hope of catching up with Tesla, which on the same day said it aims to produce its driverless taxi, the Cybercab, in 2026.

GAC and Pony.ai first forged a partnership in 2018, TechNode reported, followed by Pony’s undisclosed investment in GAC’ ride-hailing platform OnTime in April 2022, when the two companies said they would together operate an autonomous ride-hailing fleet. Meanwhile, Pony.ai has been backed by a Toyota investment worth $400 million since 2020, and the two companies, along with GAC, a manufacturing partner of Toyota in China, announced a RMB 1 billion ($139 million) joint venture to make robocars last August.

State-owned GAC has also been a strategic investor in WeRide since late 2021, a self-driving car company with regional headquarters in the southern Chinese city of Guangzhou. Both WeRide and Pony.ai are looking to sell shares in the US and have obtained approval for overseas listings from Chinese regulators. Pony.ai said on Monday its autonomous ride-hailing vehicles have traveled a total of 40 million kilometers (25 million miles), of which around 4 million km were entirely driverless.

More Chinese automakers are following Tesla’s lead with their robocars. Xpeng Motors’ new model for autonomous ride-hailing services is expected to enter production in 2026, chief executive He Xiaopeng posted on Chinese microblogging platform Weibo on Oct. 10. On the same day, Tesla CEO Elon Musk said the company’s long-awaited robotaxi, with two gull-wing doors and no steering wheel or pedals, will eventually cost less than $30,000, Reuters reported.

READ MORE: The Chinese startup battling for robotaxi supremacy

Xpeng Motors is shifting towards a  camera-based sensor system and away from expensive lidar, on Thursday providing a handful of new details on its assisted driving technology. A new electric sedan also made its debut at an event, bringing the company into direct competition with popular models including the Tesla Model 3, BYD’s Han, and the Xiaomi SU7.

Why it matters: The P7+, a completely new model following in the footsteps of its first sedan P7 launched four years ago, is expected to go on sale at a competitive price point and replicate the initial success of its cheaper-than-expected and technologically advanced MONA M03.

• Xpeng reduced the cost of making its vehicles by 25% with this B-class model “through technological advancements,” expecting it to become a top-seller in the second half of this year, chief executive He Xiaopeng told investors during an earnings call in May.

New image sensors: CEO He called Xpeng’s new lidar-less, vision-based approach, “the Eagle Eye,” saying it will be the first in the industry to harness the capabilities of the so-called lateral overflow integration capacitors (LOFIC), which allow its vehicle’s cameras to capture scenes with a large brightness discrepancy.

• Image sensors usually have a limited dynamic range of brightness that can be photographed, creating overexposed or underexposed images in difficult lighting conditions such as nightscapes and backlight. However, those with a LOFIC structure have a wider dynamic range, meaning objects could be captured with a more balanced exposure.
• This makes it possible to avoid blowing out the front vehicle when the cameras on an Xpeng’s car try to recognize it at points of extreme illumination such as at the exit of a tunnel, according to a slide presentation seen on Thursday by TechNode. Xpeng said the P7+ will integrate 26 sensors, mainly cameras, as well as dual NVIDIA’s DRIVE Orin chips, but no lidar units.
• The technology is being adopted in more premium and newer consumer electronics devices due to the sophistication of the circuit’s structure, with Sony and Samsung among the major suppliers, while Chinese firms are making their replacements. Smartphone maker Honor in March announced its Magic 6 Ultimate would feature a main camera using image sensors with a LOFIC structure from Omnivision.

A sporty sedan: The P7+ is a sporty hatchback-style sedan with a gently sloping rear roofline, and a spacious interior with an impressive headroom of 973 millimeters for passengers in the rear row. With a wheelbase of three meters, it has a boot capacity of 725 liters which can rise to 2,221 liters with the rear seats down.

• All the variants come with an electric motor mounted on the rear wheel with 180 kW or 230 kW of peak output. It is powered by an affordable iron-based battery pack featuring 60.7 or 76.3 kilowatt-hour (kWh) capacity, achieving a driving range of 602 kilometers and 710 km, respectively.
• The fives-seater boasts premium comfort for both driver and passengers, as all the seats can be heated and ventilated with massage options as standard. It also provides a folding table and an eight-inch infotainment screen for rear passengers.
• The P7+ competes with rivals’ models such as the Tesla Model 3, Xiaomi’s SU7, and the BYD Han, which are priced from RMB 231,900 ($32,764), RMB 215,900 and RMB 165,800, respectively. Xpeng will reveal the pre-sale price of the car at the Paris Motor Show next Monday. A redesigned P7 went on sale last March at a price range between RMB 249,900 and RMB 339,900.

Context: Xpeng is among the Chinese EV makers pivoting to self-driving technology that uses low-cost cameras, following the lead of Tesla, over the past two years. Its advanced driver assistance system (ADAS) will be able to work on all Chinese roadways in the fourth quarter of this year, providing users with “door-to-door” automated driving across the country, the company said in July.

Chinese automakers are escalating their battle with Tesla for a bigger share in the country’s electric vehicle market. As SAIC and Changan said on Thursday, they have benchmarked their new sports utility vehicles against the US automaker’s Model Y crossover.

Changan’s premium EV brand Avatr launched (in Chinese) its first model with both full battery electric and extended-range hybrid (EREV) options in a move to expand its lineup and improve sales against the backdrop of slowing demand for pure EVs. The EREV version of the Avatr 07 has a starting price of RMB 219,900 ($31,358) and a 38 kilowatt-hour (kWh) battery pack that can charge from 10% to 80% in 15 minutes. Meanwhile, the entry-level, all-electric variant costs RMB 10,000 higher and travels 650 kilometers (404 miles) on a full charge.

Avatr, jointly developed by state-owned Changan, battery maker CATL, and Huawei, offers Huawei’s Advanced Driving System (ADS) and the HarmonyOS, the tech giant’s alternative to Google’s Android operating system, as standard features on the mid-size SUV, as it has done with previous models. The EV maker, a major investor in Huawei’s automotive business unit, delivered only about 36,400 vehicles in the first eight months of this year with two models on offer, the large SUV Avatr 11 and the 12 sedan.

Also on Thursday, IM Motors introduced a redesigned LS6 all-electric crossover, which the SAIC-controlled luxury EV maker said was the first model in its class to feature a four-wheel, steer-by-wire system that produces more exact steering ratios than a traditional mechanical steering system. This technology, something Tesla has been keen on, offers enhanced comfort and driving dynamics. All the LS6 variants feature an advanced driver assistance system (ADAS) that enables functions such as lane switching on city streets without relying on high-definition maps, similar to Tesla’s Full Self-Driving software (FSD).

The cheapest LS6 now costs RMB 216,900, representing a 6% cut from its predecessor, while the driving range has been boosted by 12% to 625 km. The top-end, dual-motor five-seater travels 750 km on a single charge and can sprint from 0 to 100 km/m (62 mph) in 3.5 seconds, with a sticker price of RMB 279,900. IM Motors, a partner to Volkswagen’s Audi brand, has four models on sale and delivered 34,719 vehicles from January to August.

READ MORE: Huawei’s answer to the Tesla Model Y is coming

Huawei is coming after the Tesla Model Y with its second EV model under the Luxeed brand in a partnership with Chinese automaker Chery, featuring its advanced driver assistance system (ADAS), a longer driving range, and a roomier interior, all at a similar price. 

The entry-level Luxeed R7 travels 667 kilometers (415 miles) on a single charge, powered by an 82 kilowatt-hour (kWh) battery pack from Chinese supplier CALB at a price tag of RMB 259,800 ($37,022), which is RMB 9,900 more than Tesla’s Model Y. It drives autonomously on Chinese expressways using the basic version of Huawei’s Advanced Driving System (ADS) without a lidar unit for remote sensing.

The higher-end versions of the compact crossover have a maximum driving range of 802 km and can navigate Chinese urban roads, for a cost of between RMB 299,800 and RMB 339,800. Customers pay RMB 10,000 on top to access the more premium Huawei ADS 3.0 software. By comparison, Tesla sells the top-end Model Y at a sticker price of RMB 348,000 and is looking to roll out its so-called “Full Self-Driving” software in the country in the first quarter of 2025, TechNode reported.

The five-seater vehicle is said to have more head and leg room than the Model Y and a boot capacity of 837 liters which rises to 2,130 liters with the rear seats folded. Still, it faces strong competition not only from the US automaker but also from local rivals, including NIO’s sub-brand Onvo and Geely-controlled Zeekr which launched similar offerings for less money last week. 

Huawei and Chery delivered only 18,091 units of the Luxeed S7 sedan from January to August, compared with more than 254,045 Aito-branded EVs sold by Huawei and partner Seres. Tesla shipped 354,693 Model Y vehicles from its Shanghai Gigafactory over the same period, according to figures released by the China Passenger Car Association, and sales of the world’s top-selling EV model totaled 1.22 million units last year, said market research firm JATO Dynamics.

READ MORE: Huawei, Chery introduce second joint model following setback

China’s Zeekr on Sept. 20 unveiled the 7X, its first family sports utility vehicle with off-road capabilities and enhanced safety features, with a cheaper price tag than Tesla’s Model Y letting it compete against the world’s top-selling model in its home market.

Zeekr said more than 58,000 customers gave refundable deposits to reserve a place in the production line queue following the car’s debut on Aug. 30. The New York-listed electric vehicle maker is ready to deliver at least 3,000 units in the remaining days of September by ramping up manufacturing rates as demand grows, vice president Jason Lin said in an interview with TechNode.

Why it matters: Geely’s stand-alone EV brand has a tough task ahead to ensure the Zeekr 7X is successful in the highly competitive mainstream market. At least four new models from Chinese car brands, including Zeekr, NIO’s Onvo, and Dongfeng’s Voyah, are testing themselves against the benchmark in the class – the Tesla Model Y.

Details: The Zeekr 7X comes in three variants with a number of creature comforts on offer, including heated front seats with massage functions and large, automatic sun blinds on the rear-side windows and car roof.

• All the variants also boast high-tech specs such as NVIDIA chips and a lidar unit for automated driving, a 36.21-inch augmented reality head-up display for navigation, and an 800-volt electrical system that can charge the battery from 10% to 80% in 10.5 minutes.
• The entry-level 7X travels 605 kilometers (376 miles) on a single charge and goes from 0 to 100 km/h (62 mph) in 5.8 seconds at a sticker price of RMB 229,900 ($32,632). The dual-motor version costs RMB 269,900, has a 705 km-plus driving range, and is able to reach 100 km/h in as little as 3.8 seconds.
• Zeekr said the five-seater is great fun to drive off-road, as the standard version has 197 millimeters (7.8 inches) of ground clearance, while its air suspension can raise it another 33 mm for maximum clearance. In the case of an emergency submerging, a car driver can break one of the car’s windows with the press of a hidden button.
• With its raised ground clearance and a high roofline, the compact SUV goes for a traditional SUV shape, which the automaker said gives passengers ample interior headroom. It features an LED strip that can display customized lighting sequences with animation on the front hood, like its sibling, the 007 sedan.

Context: Designed to provide sufficient space and a gutsy performance catering to family occupants, the 7X will expand Zeekr’s portfolio into the Chinese mainstream SUV segment with a price range between RMB 200,000 and 300,000. The EV penetration rate surpassed 46% in this price segment last year, according to figures compiled by Chinese brokerage Guolian Securities.

• Zeekr delivered 121,540 vehicles across four models from January to August, including the 001 shooting brake and the 009 van, representing an 81% growth annually. It aims to deliver 230,000 EVs this year. Tesla shipped 354,693 Model Y vehicles from its Shanghai Gigafactory over the same period, including an undisclosed number for export, according to figures released by the China Passenger Car Association.

NIO on Thursday announced its foray into the Chinese ultra-competitive mainstream vehicle segment with the launch of the first model under its family-oriented lineup Onvo, featuring class-leading utility at nearly half the price of Tesla’s Model Y by leveraging its unique battery leasing program.

The cheapest Onvo L60 crossover costs only RMB 149,500 ($21,154) under the NIO’s Battery-as-a-Service (BaaS) leasing program, which allows customers to buy the EV without ownership of the battery pack, versus RMB 206,900 including the pack. Priced at RMB 100,400 cheaper than a China-made Tesla Model Y, the car aims to be a meaningful competitor not only to the world’s top-selling model, but also to popular gas-powered vehicles such as Volkswagen’s Tiguan L and the Toyota RAV4.

The management is witnessing “a surge” in order intake following the launch, as dozens of millions of Chinese netizens watched the livestream, William Li, founder, chairman, and chief executive, told reporters in the eastern Chinese city of Hefei on Friday (our translation). More than 60% of the customers choosing the Onvo L60 consider battery-swapping technology, part of NIO’s broader BaaS strategy, as the most desirable attribute, according to Li.

BaaS becomes the key

Boasting significant ownership cost savings and fast battery swapping, the Onvo L60 is the first serious contender from NIO to challenge the Tesla Model Y, and could be different from the previous Chinese SUVs trying to win over families from the global EV crown.

Stuffed with an autonomous driving stack of 30 sensors, Qualcomm’s latest processor for infotainment applications, and attractive extras such as a huge, 52-liter car refrigerator, the affordable, functional compact sports utility vehicle offers more than RMB 100,000 worth of specifications than the Model Y. That’s according to Ai Tiecheng, president of the Onvo brand, speaking at the launch event on Thursday.

Moreover, the management expects NIO’s battery swap service, with each time taking around three minutes from an ever-expanding power infrastructure network, to mitigate the range anxiety from Chinese consumers and even remove the real obstacle of insufficient chargers away from EV driving. Onvo owners can replace their empty battery packs for a charged one in over 1,000 NIO battery swap stations nationwide by December, and more than 2,500 such facilities will be available by the end of next year, said Ai.

The standard range version of the L60 went on sale on Thursday with the impressive starting price plus a leasing fee of RMB 599 per month for the use of a 60 kilowatt-hour (kWh) battery pack that offers a 555 kilometer (345 miles) driving range. The long-range version costs the same amount with a monthly rental fee of RMB 899, claiming a driving range of 730 km with an 85 kWh battery pack, versus RMB 235,900 including the pack. A typical Onvo owner would pay an additional RMB 100-200 each month in electricity costs, according to Ai’s estimates.

“Overall it costs you roughly a tank of gas per month,” Ai added (our translation). NIO said the total energy consumption of the car is 12.1 kWh per 100 km, compared with the 12.5 kWh achieved by the Model Y.

A turning point

The success of the L60 is of vital importance to whether or not 10-year-old, loss-making NIO would become a stronger and profitable player in a bruising battle where even some of the bigger rivals are already feeling the heat. Global auto giants from Detroit’s big three to Japan’s leaders have been hurt in China, and NIO is joining an expanding group of local players in bringing their better-equipped EVs at relatively affordable prices into the market.

Against the backdrop of a slowing economy, Onvo’s first model was precisely developed to meet the real demands of Chinese families who take space, safety, comfort, and affordability in the first place, rather than performance features like extra torque or horsepower, according to NIO.

NIO said the five-seater crossover could comfortably accommodate passengers in a carefully designed cabin, providing larger headroom and legroom across both rows than the Tesla Model Y. Owners could also access popular high-tech features such as an AI voice assistant from an in-house developed operating system, as well as perform autonomous driving functions in more than 700 Chinese cities and 2,700 counties, powered by advanced chips and AI algorithms.

Li said the company is now ready for a production ramp-up with parts suppliers, with a timeline of October to make 5,000 units of the car and increase the monthly output to 16,000 in January. Another 20,000-unit milestone could follow next March, Li added, again promising a “decent” gross margin for the model. NIO’s CEO earlier this month had briefed investors about the outlook of delivering 20,000 Onvo EVs sometime next year and setting a 15% margin goal for the Onvo brand in the long term. 

Having not had a distinct hit product in recent years, NIO and peer Xpeng Motors are showing signs of making a turnaround with their multi-brand strategies.

READ MORE: NIO CEO details thinking on first Onvo EV, battery swap alliance

BYD in a recent regulatory filing is seen to be planning a 35.6 million square feet global research and development center in the southern Chinese city of Shenzhen, where the auto giant is headquartered, as part of the company’s latest push to drive tech innovation.

The facility is to be located about eight kilometers (five miles) north of its headquarters in the city’s Pingshan district, covering an area of 3.3 million square meters (35.6 million square feet). It will comprise nine buildings and include parking space for 20,000 cars and 8,054 non-motorized vehicles, according to a filing published by the Shenzhen Municipal Planning and Natural Resources Bureau on Sept. 4.

The disclosure was revealed a year after China’s top electric vehicle maker announced the construction of its global R&D center, which will cost RMB 20 billion ($2.8 billion) and house 60,000 engineers, the city government said in a statement last June. The center will also include more than 50 research labs for emerging technologies such as nanoscale photonic physics and the development of high polymer materials.

BYD has been ramping up its R&D efforts to beef up its technology offerings over recent years and has pledged to invest RMB 100 billion in vehicle automation technologies over the long term. It expanded its software engineering team by more than 4,000 employees last May, according to senior vice president Stella Li, and over 80% of the 31,800 fresh graduates recruited last year have worked on R&D projects, TechNode reported.

The company announced on Friday that it now has the largest R&D team in the industry, as nearly 110,000 out of its approximately 900,000 employees are scientists and engineers, Li Yunfei, a general manager for branding and public relations, said on microblogging platform Weibo. Year-to-date sales of its passenger EVs, including all-electrics and plug-in hybrids, grew 29.4% year-on-year to 2.3 million units as of August.

READ MORE: BYD begins building a $2.9 billion industrial facility in Shenzhen

Huawei and Chinese automaker Chery on Tuesday began taking orders for the second model under their premium electric vehicle brand Luxeed, priced between RMB 268,000 and RMB 348,000 ($37,654 and $48,894) and featuring what an executive called the world’s most advanced driver assistance system. 

The Luxeed R7 five-seater sports utility vehicle has a sloping, coupe-style roofline, making it the most aerodynamic production car in the SUV segment, and boasts a roomier interior space than that of the Tesla Model S, according to the companies.

“We have conducted comprehensive evaluations from multiple aspects and found that Huawei ADS is the most powerful and safest intelligent driving system, bar none,” Richard Yu, chairman of the board of directors of Huawei’s Intelligent Automotive Solution business unit, said at a press event (our translation). Yu previously said Huawei’s employees narrowly avoided several incidents where a Tesla almost crashed into stationary objects such as trucks during a recent test drive in the US and Canada, TechNode reported.

The smartphone maker said drivers of ADS-enabled EVs traveled over 200 million km (124 million miles) in August, while its autonomous emerging braking feature has prevented more than 560,000 collisions for users as of Monday. By comparison, Tesla FSD users surpassed 1.3 billion cumulative miles in three years as of March this year.

The entry-level R7 has a driving range of 667 kilometers (415 miles) on a single charge and is capable of handling on-ramp to off-ramp driving and lane changing autonomously on national highways without the assistance of expensive lidar units. The higher-end versions of the R7 crossover travel 802 km at most per charge and feature the third version of Huawei’s Advanced Driving System (ADS), which the tech giant claimed allows cars to carry out most driving functions automatically from garage to garage nationwide.

The two companies expect the second model to help expand their product portfolio and drive sales for the new brand, following the lackluster sales performance of the S7 sedan, which reported deliveries of only 15,608 units from January to July. An earlier price cut did little to lift sales after customers complained about delivery delays and rivals such as Xiaomi and Zeekr rolled out their competitive offerings.

Meanwhile, Huawei, with another of its manufacturing partners Seres, has also introduced (in Chinese) a five-seat variant of their M9 full-size SUV priced between RMB 469,800 and RMB 569,800. Huawei said the three-row, six-passenger SUV has secured more than 130,000 reservations with non-refundable deposits since its launch in December and has been the top-selling model in the price segment of more than RMB 500,000 in China in the past five months.

Chinese electric vehicle makers BYD, Leapmotor, and Geely-owned brand Galaxy delivered record-breaking numbers in August, as demand began to pick up partly due to government stimulus measures, according to company statements issued on Sunday. Meanwhile, sales of Huawei-backed EV brands dropped sharply.

Why it matters: The figures come after many Chinese automakers have launched new and redesigned models, greatly increasing their store traffic and boosting sales, despite summer being typically a low season for car demand in China.

• Beijing’s new subsidies for consumers to trade in their old, less eco-friendly cars for new and greener ones is another driving factor, said the China Passenger Car Association (CPCA) in a post (in Chinese).
• The industry group expected sales of China’s new energy vehicles (NEVs), including all-electrics and plug-in hybrids, to improve by 36.6% year-on-year and 11.6% month-on-month to 980,000 units in August.

Details: Stellantis-backed Leapmotor in August surpassed the threshold of 30,000 units sold for the first time since its inception, as sales of its C16 sports utility vehicle continue to gain traction. More than 8,000 units of the affordable six-seater SUV, which went on sale on June 28 at a starting price of RMB 155,800 ($21,937), were handed over to customers last month.

• Geely’s mainstream EV lineup Galaxy also set a new delivery record, generating 26,510 unit sales in August, which represented a 58.7% increase from a month earlier. The strong growth was partly driven by the launch of the E5 all-electric compact crossover, featuring an in-car operating system called Flyme Auto, which is priced between RMB 109,800 and RMB 145,800 until the end of September.
• Meanwhile, sales of Huawei’s Harmony Intelligent Mobility Alliance (HIMA) dropped sharply partly because Seres, a manufacturing partner of Huawei, was reportedly refining manufacturing lines before launching upcoming models. Seres, as well as Chery, together delivered 33,699 EVs featuring Huawei’s in-car technology in August, down 25% from a month earlier and far fewer than rival Li Auto.
• Notably, Volkswagen-backed Xpeng Motors recorded its highest monthly deliveries this year, signaling a likely recovery after nearly a year of lackluster sales. The company said it delivered 14,036 units last month and secured more than 30,000 reservations with non-refundable RMB 5,000 deposits for the MONA M03 sedan in the 48 hours after its unveiling on July 27.
• BYD clocked its highest-ever sales of 370,854 passenger EVs in August with 35.3% year-on-year growth. Last month, the EV giant rolled out several new and updated models to defend its dominant position, including a facelift of its Seagull small SUV, priced from RMB 69,800, and a RMB 339,800 luxury grand tourer, the Denza Z9 GT. It also unveiled the first multi-purpose vehicle under the Dynasty lineup, the Xia, on August 30.

Context: China’s central government announced on August 16 that subsidies for trade-ins of NEVs and internal combustion engine cars have doubled or more to RMB 20,000 and RMB 15,000, respectively, a significant ramp-up compared with the previous rules issued in April.

• More than 680,000 applicants have submitted their filings for the cash incentives as of August 22 with more than half filed over the past month, according to a spokesperson from the Ministry of Commerce (in Chinese).

READ MORE: BYD, Li Auto, Leapmotor post record monthly sales, putting pressure on rivals

​US-listed shares of Xpeng Motors climbed 6.5% after the Chinese electric vehicle maker received more than 10,000 non-refundable orders for the first model under its new budget brand MONA in the first hour after it was unveiled on Tuesday. 

The company’s cheapest model, featuring an assisted driving system similar to Tesla’s Full Self-Driving (FSD) software at a price that’s half that of the popular Model 3, is expected to capture market share from top-selling offerings including the BYD Qin and Nissan’s Sylphy.

Camera-based vision: Xpeng eliminated laser sensor units from the MONA M03 in a move to accelerate the adoption of a camera-only approach to vehicle autonomy. Lidar is more often used to capture three-dimensional information for autonomous navigation on more luxurious Chinese EV models priced mostly from RMB 200,000 ($28,060) and above. 

• Instead, the entry-level M03, with a price tag of RMB 119,800 ($16,808), integrates 20 or so more affordable sensors, including a dozen ultrasonic sensors and seven exterior cameras, to enable safety-based driver assistance functions such as lane keeping assist, which enables the EV to follow the preceding vehicle. Xpeng’s Xpilot system would also maneuver the car in and out of challenging situations such as dead-end roads.
• The top-end version of the MONA M03, priced at RMB 155,800 ($21,859), will be able to function by itself on all Chinese roadways using the same hardware stack as well as more complex algorithms next spring when delivery is set to begin. This makes the car the only model available to the market that offers advanced driver assistance capabilities within the RMB 200,000 price range, said chief executive He Xiaopeng. 
• Xpeng also provided more details about its plan to make EVs with highly autonomous capabilities as part of the launch. The XNGP advanced driver assistance system (ADAS), an FSD-style technology, should require drivers to take over the wheel and steer manually less than once per 100 kilometers (62 miles) by the end of next year. The company added its cars could be completely driverless in low-speed traffic scenarios in 2026. 

READ MORE: Chinese companies take on Tesla’s Full Self-Driving with non-lidar approach, end-to-end AI 

Aiming for the youth market: Xpeng on Tuesday reported a better performance than some rivals’ offerings in terms of order intake for the MONA M03, saying the car racked up more than 10,000 pre-orders with non-refundable deposits of RMB 5,000 in roughly the first hour after it was released. 

• Positioned to be “the first car for China’s young generation,” the all-electric compact sedan boasts a large interior space, efficient energy consumption, and yet a relatively simple interior in order to attract the country’s tech-savvy but budget-conscious customers, such as fresh graduates.
• Measuring 4.8 meters in length and spanning a 2,815-millimeter-long wheelbase, the car was designed based on aerodynamic principles with sleek coupe-style body lines. This contributes to a drag coefficient of 0.194 and energy consumption of 11.5 kilowatt-hours per 100 kilometers (7.15 kWh per mile). 
• The single-motor M03 offers a decent range of up to 620 km (385 miles) on a single charge, powered by an affordable lithium iron phosphate (LFP) battery pack from BYD, and can accelerate from 0 to 100 km/h (62 mph) in 7.4 seconds. It also gives an additional 200 km of range after 15 minutes of charge. 
• Still, the car’s well-built but simple interior features may leave some drivers wanting more. An M03 buyer would have to pay an additional RMB 599 to equip their car with a conventional dashboard screen as an add-on component, for example. Xpeng stated that the car would function properly with a 15.6-inch centre screen. 

Context: Xpeng is expecting a strong sales recovery with the launch of the MONA M03 at a lower-than-expected price tag, as CEO He in April said he hoped that the car would sell better than the Xiaomi SU7 in the second half of this year, Reuters reported. The smartphone maker has been delivering more than 10,000 units per month of the $30,000 sports sedan since June. 

• He also anticipated annual sales of the car could reach 100,000 units when Xpeng revealed last year that it would acquire the EV business of ride-hailing giant Didi, including the MONA project, which stands for “Made of New AI.” He added on Tuesday that the two companies spent RMB 4 billion on the project over four years. 
• The gross margin of the M03 could be at a low single-digit level initially, however, as the Volkswagen-backed EV maker ramps up production and looks to keep its price as low as possible, Jefferies analysts wrote in a note on Aug. 20. 
• BYD’s Qin Plus was the top-selling model in the price segment of between RMB 120,000 and RMB 180,000 in China over the past six months with roughly 154,000 units sold, figures compiled by Chinese auto service platform Dongchedi showed.

Chinese action role-playing game (RPG) Black Myth: Wukong has sent shockwaves through the global gaming community and the impact is spreading to the automotive industry, as a Geely-linked firm is testing out new in-car gaming experience with the hit title, TechNode has learned.

Why it matters: Geely’s move shows how software-enabled advanced functionality, from watching video streams on the dashboard screen to letting the car change lanes autonomously, is becoming a decisive factor for vehicle buying in the world’s biggest auto market. Bringing China’s first AAA game to its cars could be a new way for Volvo’s parent to stimulate demand among younger drivers

Details: Ecarx, a Chinese auto tech startup backed by Geely’s chairman Li Shufu, has managed to deliver 2K resolution (2,560 x 1,440 pixels) at nearly 60 frames per second (FPS) when it tested the Black Myth: Wukong on its next-generation digital cockpit computing platform Makalu. That’s according to people familiar with the matter who spoke to 36Kr (in Chinese) on Tuesday.

• The instrument panel, the central information screen, as well as a head-up display (HUD) unit that projects traffic information onto the windshield, were functional while the game was streamed, according to the report. A company spokesperson confirmed to TechNode on Friday that the test is ongoing to improve the gaming performance further. 
• Ecarx announced last March that it had become the second auto firm worldwide to bring in-car entertainment experience with AMD’s laptop graphic chips. The Makalu computing platform consists of AMD’s Ryzen V2000 processors paired with the AMD Radeon RX 6000-series GPUs, providing up to 10.1 teraflops of computing power, more powerful than Qualcomm’s latest-generation Snapdragon 8295 chips.
• For comparison, Tesla has also pursued the offering of in-car games as chief executive Elon Musk reportedly claimed in 2020 that owners would be able to play the adventure title The Witcher 3 in its vehicles. AMD said in 2021 that its Ryzen computers and RDNA2 GPUs are powering the infotainment systems in the new Model S and Model X vehicles. Sony also integrates SoCs (System on Chip) from the US chipmaker into its PlayStation 5 consoles.
• The Makalu platform allows game developers to adapt their games to in-car systems by using Unreal Engine, a real-time, three-dimensional graphics creation software from game company Epic Games, which has been used by China’s Game Science for the production of Black Myth: Wukong. Vehicle occupants could therefore play triple-A high-definition video games via Ecarx’s system, the company said on its website. 
• Geely confirmed that the Lynk & Co Z10 large sedan will be the first model to feature a digital dashboard that integrates Ecarx’s latest computing platform. The first all-electric model under the Geely and Volvo Car joint brand will be launched on Sept. 5 following reservations opening on August 15 at a pre-sale price of RMB 215,800 ($30,234). 

READ MORE: Geely’s Lynk & Co 08 to use in-house car software from Meizu for the first time

Context: Global automakers such as Volkswagen and Mercedes-Benz have been working on mobile technology and artificial intelligence to shape the future of immersive in-car experiences. Meanwhile, NIO and Ji Yue, a Chinese EV brand owned by tech giant Baidu, have enabled users to play the hit racing game Asphalt in their latest models equipped with a Qualcomm semiconductor. 

• Black Myth: Wukong is China’s most successful PC-end game to date after selling around 4.5 million copies across various gaming platforms by far. Inspired by the Chinese literary classic Journey to the West, the action game hosted more than 2.2 million users online concurrently just hours after it was launched on Tuesday, TechNode reported. 

Chinese tech giant Huawei’s automotive business unit has been valued at around RMB 115 billion ($16 billion), after it struck a deal to sell a 10% stake to Changan Automobile for RMB 11.5 billion, the two companies announced on Tuesday.

Why it matters: Changan shares fell by 4.3% on Tuesday as the deal will see it own a smaller than expected stake in Huawei’s car unit. Huawei will retain its control of the business and pursue a diversified ownership structure to deepen its links with multiple automakers.

Details: Changan issued a statement (in Chinese) on Monday evening saying that its subsidiary Avatr was set to sign a deal with Huawei the next day to take a 10% stake in Shenzhen Yinwang Intelligent Technology Co., Ltd. (also called Newcool) for about RMB 11.5 billion. Avatr will nominate a director to sit on the company’s seven-person board.

• The signing ceremony, which also took place on Tuesday morning, was attended by Eric Xu, rotating chairman of Huawei, and Richard Yu, chairman of the board of directors of Huawei’s Intelligent Automotive Solution (IAS) business unit, as well as Changan’s president Zhu Huarong, according to an image posted by Avatr on microblogging platform Weibo.
• The size of the deal was significantly smaller than expected– Changan had revealed plans to acquire a stake of up to 40% stake in the joint venture in November. Meanwhile, Seres, another manufacturing partner of Huawei, said (in Chinese) on July 28 that it has been in talks with the smartphone maker to buy parts of the company.
• Huawei is also in early-stage talks with state-owned automaker FAW to sell a minor stake in its auto business, although no significant progress has yet been made. Peers Dongfeng Motor, BAIC, and JAC are among the potential investors in the business as well, with these firms all having developed or about to launch new electric vehicles that feature Huawei’s assisted driving technology.
• Huawei in 2020 co-established Avatr with Changan and Chinese battery maker CATL by providing technological know-how without direct investment in the luxury electric vehicle brand. Changan and CATL are the largest shareholders in Avatr with stakes of roughly 41% and 15%, respectively.

Context: Huawei expanded its consumer electronics business into the booming Chinese EV industry with the establishment of its IAS business unit in mid-2019, just as the US banned the supply of semiconductors and operating systems to the Chinese tech giant. It began selling EVs for partners via its retail network in early 2021 and has stressed repeatedly it has no intention to build vehicles itself, but instead aims to help automakers make better cars.

• Huawei has since faced pushback from some traditional Chinese automakers amid concerns that they could lose out on customers if the company makes its own cars. Some traditional automakers have also looked to retain more ownership of in-car technologies through partnerships with smaller suppliers.
• The situation improved last year when the redesigned Aito M7 crossover and premium M9 sports utility vehicle produced by Seres became hits. Legacy automakers have also been pressured by a ferocious price war and have struggled with their own autonomous vehicle efforts.
• Huawei estimated that around 500,000 cars equipped with its technology will be on the roads by year-end, Reuters reported in April, saying it expected its car business to break even this year compared with the RMB 6 billion losses it recorded in 2023. It is also partnering with global majors including Toyota and Audi.

READ MORE: Huawei creates separate car division, open to Changan and other outside investors

Audi’s new A5 sedan and all-electric Q6 e-tron crossover will be the company’s first models to use Huawei’s advanced driver assistance system (ADAS), enabling the cars to change lanes and overtake autonomously in Chinese cities, local media has reported.

Why it matters: The news marks the latest development in a six-year collaboration between Audi and Huawei. The companies announced in mid-2018 that they had signed an agreement to develop a level 4 autonomous driving system that could take complete control of a vehicle without human intervention in Chinese urban traffic environments.

• Audi had been planning to incorporate Huawei’s autonomous driving technologies in future China-made models scheduled for launch in 2025 and had been asked to buy small stakes in Huawei’s intelligent vehicle software and components firm, Reuters reported in December.

Details: The new A5 sedan, the first model that utilizes Audi’s last new internal-combustion architecture called Premium Platform Combustion (PPC), is expected to make its China debut at this year’s Guangzhou Auto Show in the capital of Guangdong province in November. That’s according to people familiar with the matter who spoke to Chinese media outlet iYiou.

• Sources added that the hatchback-style sedan, which replaces the A4 model in the German firm’s vehicle lineup, will also incorporate Huawei components such as electronic control units, and is intended to be launched at the Auto Shanghai Show 2025 next April. The A5 will be manufactured by Audi and FAW Group, one of its Chinese partners.
• Meanwhile, there has been speculation that the Q6 e-tron, Audi’s first car to use Volkswagen’s advanced premium platform electric (PPE) platform, will also use Huawei’s ADAS. Meanwhile, Audi is working with Chinese self-driving car startup Momenta on automated driving for a separate new series which has been under development with SAIC, the German firm’s other local manufacturing partner, the report said.

Context: Volkswagen’s Audi is not the only international car major open to teaming up with Huawei. Toyota has reportedly agreed to buy Huawei’s autonomous driving stack for its China-made electric vehicles, while a growing number of Chinese car manufacturers are expanding their collaborations with the tech giant. Huawei has estimated that around 500,000 cars equipped with its tech will be on the roads by year-end, Reuters reported.

• Huawei is gaining traction thanks to growing customer demand for intelligent driving functionalities in the world’s largest auto market. One company executive said last week that the latest Advanced Driving System (ADS) offers Chinese users better safety in a point-to-point autonomous experience than Tesla’s Full Self-Driving (FSD) system, as TechNode reported.
• Audi reported sales of more than 320,000 luxury cars in China for the first half of this year, falling slightly by 1.9% from the previous year and lower than the total sales of BMW and Mercedes-Benz. The German luxury carmaker has been planning a series of measures to defend its market share as domestic rivals such as Li Auto and NIO are on the rise, including buying EV technologies from SAIC and designing a new logo for future vehicles, reported Reuters.

Huawei on Tuesday released the latest version of its advanced driver assistance system (ADAS), saying  it would enable vehicles to carry out most driving functions automatically from garage to garage while delivering a better safety performance than Tesla’s most advanced Full Self-Driving (FSD) system.

The Chinese technology giant also launched the Stelato S9, the first model under its new luxury brand with manufacturing partner BAIC, at a lower-than-expected price range of RMB 399,800 to RMB 449,800 ($55,664-$62,625). The five-meter-long full-size luxury sedan is targeted at consumers of premium offerings such as the Mercedes-Benz S-Class and BMW’s 7 Series.

Huawei’s self-driving approach: Huawei’s latest Advanced Driving System (ADS) now allows cars to navigate complex urban environments nationwide automatically, with features such as lane switching, said Richard Yu, chairman of the board of directors of Huawei’s Intelligent Automotive Solution business unit, at a press conference in Beijing. Yu added that the ADAS will enable drivers to pass toll gates on Chinese expressways and at public parking lots without manual engagement. 

• In a demonstration video, Huawei boasts of a point-to-point autonomous experience with no human input at all, as the Stelato S9 navigates from one garage to another, autonomously piloting itself on Chinese city streets, and through a series of roundabouts and intersections. The female driver in the video is shown drinking a coffee and applying make-up behind the wheel as the car drives into and out of the garages.
• Huawei said it is transitioning to an “end-to-end” autonomous driving method, defined by Tesla’s Elon Musk as a state whereby each functionality of the ADAS is carried out by a single gigantic neural network rather than by predefined, explicitly coded rules. However, Huawei’s slides showed that the company uses separate neural networks for multiple tasks such as visual perception and path planning, in a move similar to that of rivals Li Auto and Xpeng.
• Yu said users would see an improvement in decision-making speed and more human-like behavior from the ADS 3.0 version, as well as a more agile automatic emergency braking feature in a joint effort with the company’s unnamed German supplier. Huawei claimed the system would even apply brakes automatically when a collision was imminent with a lead or rear vehicle, as well as if the system detects that drivers intend to jam the brakes but mistakenly slam on the accelerator.
• In addition, Yu spoke publicly about Tesla’s FSD after his employer tested the rival’s system in multiple cities in the US and Canada, saying the smartphone maker has a better strategy for dealing with rare or complex road scenarios, known as corner cases. “Tesla has made big progress with its FSD, but it has to raise the lower limit of the system’s performance as we encountered several incidents where a Tesla almost crashed into stationary objects such as trucks,” Yu added (our translation).

Stelato’s first model: Huawei on Tuesday stepped up its offensive against international luxury carmakers as the first Stelato model went on sale at a starting price of RMB 398,000, roughly RMB 50,000 lower than the pre-sale price and below market expectations. Yu said the S9 executive sedan was “fully ahead of” rivals’ offerings, such as the Mercedes-Benz S450L and Maybach S480, with a more efficient interior space, improved aerodynamic properties, a longer driving range, and faster acceleration.

• The first full-size luxury sedan under Huawei’s Harmony Intelligent Mobility Alliance (HIMA), a joint EV effort between Huawei and its multiple manufacturing partners, is powered by a 100 kilowatt-hour (kWh) battery pack for a driving range of up to 816 kilometers (507 miles). The battery, jointly made by CATL and Huawei, also adds 200 km on a five minute charge, when the car is used with an 800-volt capable fast charger.
• The Stelato S9 boasts a faster acceleration to 100km/h (62 mph) than existing offerings – in 3.9 seconds – as well as a drag coefficient (Cd) of 0.193, said to be the lowest among production sedans in its segment. It also integrates a lidar unit to measure and map the surrounding environment, in addition to cameras and radar, which the tech giant believes could provide extra input to its ADAS system, compared with Tesla’s pure-vision approach.
• Delivery is scheduled to begin immediately on Wednesday, demonstrating BAIC’s pledge to ”go all-in“ in its partnership with Huawei and treat the new model as its “top priority,” as chairman Zhang Jianyong told the conference. A major partner of Mercedes-Benz in China, BAIC recently established a new plant capable of making 120,000 EVs annually with Huawei and plans to expand that capacity to 300,000 EVs, Zhang confirmed on Tuesday.

READ MORE: Chinese companies take on Tesla’s Full Self-Driving with non-lidar approach, end-to-end AI

Chinese electric vehicle makers BYD and Li Auto set new records for deliveries in July, usually a low season for auto sales, following the strong momentum built in the previous month and likely extending their lead over rivals as competition in China’s EV sector remains intense.

Why it matters: The numbers come against a stormy backdrop with a growing number of Chinese automakers taking aim at each other in the press, as a cutthroat price war has been running for more than a year and a half, deepening losses of most players.

• Li Auto was put in the dock of public opinion recently for publishing weekly sales figures including those of its major rivals such as Huawei-backed Aito, Geely-affiliated Zeekr, and Xiaomi, pushing most of the rest of the industry to complain. On July 30, Ma Lin, an assistant vice president of brand and communication at NIO, quoted China’s top leadership on social media and called on Li Auto founder Li Xiang to stop the “low-level competitive behavior” that has gone on since last April (our translation).
• On the same day, Xpeng chief executive He Xiaopeng openly criticized some rivals competing over sales volumes on a weekly basis rather than leading technologies such as end-to-end neural networks, without mentioning Li Auto by name. Victor Yang, a senior vice president at Geely, also publicly endorsed their opinions on the Chinese microblogging platform Weibo. The comments came immediately after China’s top decision-making body the Politburo pledged to prevent “the vicious competition of involution” in domestic industries, Caixin reported.
• Top Chinese auto business leaders overall appear split on the issue of competition. At this year’s China Auto Chongqing Summit in June, BYD chairman Wang Chuanfu said he advocated for businesses to embrace the competitive spirit and strive for excellence, while other counterparts, such GAC’s Zeng Qinghong and Geely’s Eric Li, said disorderly competition could result in inferior quality and shrinking profits. Richard Yu, chairman of Huawei’s Intelligent Automotive Solution business unit, playfully called BYD “the king of Juan,” which in Chinese means the top player amid fierce market competition.

Details: BYD announced on Thursday (in Chinese) that it sold 340,799 EVs in July, the company’s best-ever monthly result this year and just 244 units fewer than its all-time high in December. Still, the Chinese EV giant saw a lackluster performance in its premium lineups. Fangchengbao’s sales plunged 31.3% to 1,842 units from a month earlier, while Yangwang’s sales of 439 units remained flat compared to June.

• Li Auto posted its best-ever monthly result, delivering 51,000 plug-in hybrid and fully electric vehicles in July, bringing its total delivery count for the year to almost 240,000 units. This was attributed to the strong sales of the L6, its most affordable crossover, which accounted for roughly 40% of its deliveries last month. The $34,500 extended-range hybrid crossover travels up to 1,390 kilometers (864 miles) on a full charge and tank.
• Stellantis-backed Leapmotor surpassed the threshold of 20,000 units for the second month in a row, as sales of its C16 sports utility vehicle gained momentum. The company has positioned the vehicle as a cheaper alternative to Li Auto’s L9 with a similar appearance, range, and practicality. The Hong Kong-listed EV startup said on July 22 it had more than 10,000 non-refundable orders for the six-seater crossover at a starting price of RMB 155,800 ($21,610).
• Meanwhile, sales of Geely, Great Wall Motor, NIO, and Huawei-backed EV makers were down slightly. NIO delivered 20,498 units in July, only 711 less than a month earlier, and passed the 20,000 unit milestone for a third consecutive month, though the company has been providing fewer cash incentives to buyers since July 22.
• Huawei’s Harmony Intelligent Mobility Alliance (HIMA), which for now includes Aito and Luxeed brands only, delivered 44,090 units last month, representing a 4% decline from June, in the face of competition from players such as Li Auto. The new energy vehicle (NEV) sales of Geely and Great Wall Motor, which include all-electrics and plug-in hybrids, decreased 10.5% and 7.5% month-on-month in July, respectively.
• Xiaomi, Xpeng Motors, and CATL-backed Hozon Auto each delivered more than 10,000 units last month. China’s smartphone giant is aiming for an annual delivery of 120,000 units of its first model, an all-electric sports sedan with styling similar to the Porsche Taycan and priced at less than $30,000. Xpeng’s July deliveries increased 4% from June and the company expects to improve the performance of its assisted driving technology later this year.

Context: China’s sales of new energy passenger vehicles increased 31% year-on-year but declined 5% month-on-month to roughly 722,000 units from July 1-28, according to figures released by the China Passenger Car Association. Increased promotions boosted sales in the April-June quarter, including some advance buying from customers looking to nab deep discounts, said the industry group.

NIO on July 26 gave updates on some of its most important technological developments, including the world’s first five-nanometer chip for automated driving, its in-house developed operating system for vehicles, and a voice assistant powered by its large language model.

Why it matters: NIO’s fullest disclosure yet of its technological roadmap reflects how the Chinese electric vehicle maker spends its research and development money, and how it has made a strategic bet on artificial intelligence, hoping to redesign all aspects of its vehicles including self-driving and the so-called digital cockpit.

Details: Speaking at the firm’s annual Tech Day event in Shanghai, NIO founder and chief executive William Li unveiled new details of what he called the world’s first five-nanometer (5nm) processor for autonomous driving, said to offer cameras with top-notch image signal processing (ISP) functions.

• The Shenji NX9031 system on chip (SoC) delivers high pixel throughputs of 6.25 Gigapixels per second, which NIO said makes it capable of capturing pictures with higher quality resolution in minimally lit environments, compared with flagship offerings from leading chipmakers.
• Li confirmed that the company had completed the tape-out for its self-designed autonomous driving semiconductor, which marks the end of the design phase and the beginning of production, without revealing which company has been contracted to produce it. 
• Meanwhile, NIO is on track to release its advanced driver assistance system, the Navigate on Pilot Plus (NOP+) 2.0, in the second half of the year. The Shanghai-headquartered EV maker is one of the Chinese players looking to harness the power of AI to provide cars with point-to-point navigation on Chinese highways, city streets, and parking spaces in a race led by Tesla.
• NIO also announced the launch of its proprietary operating system, the Sky OS, with which the company said it would have central control of all vehicle domains, allowing it to provide more advanced user experiences. An alternative to the open-sourced Linux OS and Blackberry’s QNX, the Sky OS forms the basis of the entire car all the way from vehicle control through to intelligent driving and infotainment, NIO said.
• This will enable a NIO vehicle to have a used battery pack exchanged for a new one fully automatically at one of the company’s nationwide recharging facilities, allowing the user to leave the car to get a coffee, according to vice president Wang Qiyan.
• The Chinese EV maker will also release its Banyan 3.0 in-car user interface through over-the-air updates in late August, featuring an enhanced version of its NOMI conversational assistant, powered by generative AI large language models (LLMs), for more human-like voice interactions. NIO owners will be able to request a chauffeur to drive them via voice commands, the company said.

Context: Only a few global chip powerhouses, such as Qualcomm and Ambarella, have announced the production of their chips using a 5nm process node with manufacturing partners such as Samsung. Tesla reportedly plans to produce its next-generation full self-driving (FSD) computer on 4/5nm processes with TSMC.

• NIO previously implied that its sophisticated chip could handle more than 1,000 trillion operations per second (TOPS), comparable to the combined total of four NVIDIA DRIVE Orin chips on 7nm, and is scheduled for mass delivery with the company’s ET9 executive sedan in early 2025.
• NIO’s research and development expenditures grew 23.9% annually to RMB 13.4 billion ($1.89 billion) last year. CEO Li told the event that the company has more than 11,000 R&D personnel, adding that its years-long efforts on in-house chip development will be “of great significance” to the industry of self-driving cars (our translation).

READ MORE: Chinese EV makers Nio, Xpeng, and Li Auto expand bets on self-produced chips: report

China- and US-based self-driving car startups Pony.ai and WeRide are ready for initial public offerings in the US, multiple media outlets have reported. The IPOs, which the firms have been eyeing for over two years, could happen as early as within the next two months.

Why it matters: Autonomous vehicle (AV) startups are generally facing pressure to lower their valuation targets as public stock markets have rapidly decelerated their interest in the space over the past few years.

• A group of 14 self-driving tech companies have seen their combined valuation decline more than 80% since their public debuts, according to an analysis by Crunchbase in 2022. Chinese car tech firm iMotion and lidar maker Robosense have lost roughly 75% and 90% in stock market valuation in the six months since their listings in Hong Kong.
• The potential listings of PonyAV.ai and WeRide, two of the highest-valued AV startups in China, will set the tone for many other companies trying to go public with their future-shaping technology, even as it takes longer than expected to become widely available.

Details: Fremont- and Guangzhou-based Pony.ai will go public as early as September, as some institutional investors have committed to purchasing shares in the looming IPO, financial media outlet Jiemian reported on Tuesday, without giving further details (in Chinese). WeRide is planning to sell its shares publicly in the US by the end of August, according to IFR, a Reuters publication.

• WeRide, with its China headquarters in the southern city of Guangzhou, last August received approval for an overseas listing from the China Securities Regulatory Commission, which requires applicants to complete the share sale within 12 months of the date of filing. Pony.ai gained a similar regulatory greenlight in April, Bloomberg reported. Pony.ai and WeRide declined to comment when contacted by TechNode on Thursday.

Context: Pony.ai reportedly suspended a public listing plan in New York at a target valuation of $12 billion in mid-2021 due to regulatory uncertainties. The Toyota-backed company said early the next year that it was valued at $8.5 billion after closing its first round of Series D financing, TechCrunch reported.

• WeRide, backed by the Renault-Nissan-Mitsubishi Alliance, filed confidentially for a US IPO aimed at raising as much as $500 million, Bloomberg reported last March. The company was valued at about $5.1 billion in a November 2022 fundraising round, according to its official website.
• Some other Chinese self-driving technology companies, including Horizon Robotics, Momenta, and Black Sesame, are rushing to sell shares publicly either in the US or Hong Kong as early investors are eager to earn money after incubating the companies for years, TechNode has learned.

Autotech startup Nullmax said on Tuesday that its latest generation of autonomous driving hardware and software package, allowing cars to navigate complex urban environments autonomously with features such as lane changing, will cost users as little as “several thousand RMB.” 

Why it matters: Shanghai and Fremont-based Nullmax is among the few players in the self-driving vehicle space claiming that cars will be able to function by themselves in urban scenarios without maps and lidar. Instead, the company said artificial intelligence models can be used to enable cars to navigate from points A to B.

• This makes its system even more cost-competitive on the Chinese market compared with Tesla’s Full Self-Driving (FSD) software, chief executive Xu Lei told reporters in Shanghai. Tesla has reportedly partnered with Baidu to leverage the latter’s lane-level navigation and standard definition mapping services, as part of its push to localize its most advanced driver-assistance software (ADAS) in the country.

Details: Xu told a press conference that his company is advocating a “pure vision” and “end-to-end” approach, as Tesla has been doing and many are following its lead, which involves deep neural networks, using cameras only to perform autonomous driving functions (our translation).

• The car could complete the highway on-ramp to off-ramp maneuver and drive through a construction zone in Chinese urban areas with affordable hardware of 7-11 cameras and a computing platform that can handle roughly 100 trillion operations per second (or TOPS), according to Nullmax.
• Xu expects some car models equipped with the technology to come to market in 2025, without saying more. Nullmax, with roughly 300 scientists and engineers, has been developing more safety-based ADAS functions for domestic automakers SAIC, and Chery, among others, TechNode has learned.
• Cars powered by Nullmax’s tech have traveled a combined 10 million kilometers (6.2 million miles). Xu added more actual driving data is required to provide cars with point-to-point navigation on city streets, while the company is training the ADAS system via simulation with AI-generated data.

Context: Chinese EV startups led by NIO, Xpeng Motors, and Li Auto have been ramping up efforts to transition from “rule-based” designs to an “end-to-end” autonomous driving method. Meanwhile, traditional car manufacturers are tapping into the power of AI by working with tech giants such as Huawei and NVIDIA, as well as startup unicorns like Horizon Robotics and Momenta.

• “What we see in the future is that the (AI-defined) AV stack will become an end-to-end model and will be trained in the cloud with massive data. More importantly, it will be validated in the cloud with simulation capabilities as well,” said Wu Xinzhou, NVIDIA’s vice president of automotive at its annual developer conference GTC in March.
• The US chipmaker in late 2021 released a system called “Omniverse Replicator” to facilitate the training of autonomous vehicles in the virtual world, Reuters reported. Chinese major EV makers, including BYD and NIO, are building their ADAS upon its DRIVE Orin computers, each of which offers 254 TOPS of performance.
• Nullmax was co-founded in 2016 by Xu Lei and Justin Song, both of whom spent time at Tesla. Xu worked on Tesla’s Autopilot computer vision team in 2015 and 2016, after leaving a senior engineering role at Qualcomm, while Song was responsible for engineering the Tesla Autopilot and car infotainment system from 2012 to 2015, according to their LinkedIn profiles.

READ MORE: Former Tesla engineer shares thoughts on end-to-end autonomous driving at WAIC 2024

Editor’s note: ‘Landing AI’ is a series of special reports focusing on the field of Artificial Intelligence curated by TechNode. By investigating the development of AI landing in China and the behind-the-scenes stories of the industry, we’re going to dive deeper into everything that’s possible under the new wave of AI.

A former Tesla engineer, now head of autonomous driving at China’s Chery, on July 4 shared her outlook on Chinese automakers’ future capacity to scale autonomous driving technology, as Tesla leads the way with its mass data and sophisticated artificial intelligence models.

Domestic companies are roughly 1.5 to 2 years behind Tesla in developing “end-to-end neural network” technology for autonomous driving. The recent release of Tesla’s Full Self-Driving (FSD) software v12 marked “a great leap forward” in applicability and user experience (our translation) according to Gu Junli, chief executive of Chery-affiliated ZDrive.ai, and a former senior staff machine learning engineer at Tesla, speaking at this year’s World Artificial Intelligence Conference (WAIC) in Shanghai.

Gu said she expects Chinese players to partially replace modules of advanced driver assistance systems (ADAS) with data-driven, rather than rule-based, neural networks by 2025. This shift comes as car companies ramp up efforts to collect fresh, multiformat data to improve models with more computing power. However, there would not be a single large neural net, where the core capability of Tesla’s latest software is built, for production cars until 2026-2027 in China, Gu added. Tesla in April announced its FSD users had traveled a cumulative 1.2 billion miles.

Recent figures from the community-run FSD Community tracker show Tesla’s remarkable progress in refining its autonomous driving features. The latest FSD v12.3.6 version can drive 372 miles (599 kilometers) between critical disengagements, compared with just a few miles it reportedly achieved two years ago, according to crowdsourced data released by the website in May. Tesla chief executive Elon Musk later forecasted on X “a major improvement in MPI (miles per intervention)” with the release of the FSD v12.5 in late June.

“This means the system has met the requirements for most users on normal daily commutes,” said Gu.

Several Chinese early movers are working to achieve a partial end-to-end pipeline for enabling their advanced driving features, hoping to gradually transition to fully end-to-end algorithms. On July 5, Li Auto revealed details about what it called “China’s first single one model” behind its Navigation on ADAS (NOA) functions. The Chinese electric vehicle startup claimed this model would use raw sensor input to generate vehicle motion plans. However, unlike Tesla’s solution of “basically photons in and controls out,” as described by Musk, it still relies on pre-defined, explicitly coded rules to produce control actions as output.

Meanwhile, other peers, such as Xpeng Motors and Huawei, have been taking partial end-to-end approaches, using separate neural networks for multiple tasks such as perception, planning, and action, while many are still struggling to improve results. According to a recent survey compiled by China’s Gaogong Industry Institute, Chinese users typically take the controls at least 1-2 times per 100 kilometers (62 miles) as their cars’ ADAS functions do not react appropriately in urban scenarios. Tesla may find issues in localizing its ADAS software due to China’s complicated traffic conditions, said Xpeng’s head of autonomous driving, TechNode reported.

Gu mentioned Chery’s plan to launch the advanced city NOA function for some premium models in the first quarter of next year, while its single deep learning network could go live in 2026. The Chinese auto major, which sold nearly 1.9 million cars last year, plans to integrate Level 2 ADAS systems into as many as 500,000 new cars sold this year, including 150,000 units for overseas markets. Based in the eastern Chinese city of Wuhu, Chery had a relatively late start in autonomous driving, with ZDrive.ai being set up early last year.  

Despite Tesla’s leading position with its training wheels finally taken off, Gu envisioned that Chinese companies would contribute to a wider and more personalized adoption of self-driving technology worldwide. State-owned Chery in April reached a joint venture deal with Spain’s EV Motors to produce cars at a former Nissan plant in Barcelona later this year, Reuters reported. Meanwhile, customers from some overseas markets could access Xpeng’s XNGP ADAS capabilities as early as next year, president Brian Gu told investors during an earnings call in May.

READ MORE: Chinese companies take on Tesla’s Full Self-Driving with non-lidar approach, end-to-end AI

As Tesla’s most advanced driver assistance software (ADAS) is becoming an immediate threat due to its impending arrival in China, major Chinese electric vehicle makers and auto tech companies are hurrying to pivot their strategies towards a more pragmatic yet challenging approach to developing similar offerings. Although Tesla’s rivals have for years been looking to cut out expensive components and master the newest artificial intelligence models, the game seems to be different this time.

Both NIO and Xpeng Motors are now embracing the so-called computer vision approach, championed by Tesla, hoping their upcoming models will achieve human-like downtown driving in cities via the use of fitted cameras and radar, rather than more expensive laser sensor units. Sinpro.ai, a supplier to NIO of ultra-high-resolution four-dimensional (4D) imaging radar, said it is prepared for delivery later this year with an annual capacity of 800,000 units. Tesla has reportedly replaced radar sensors on some models after years of attempting to remove them.

“It will be a difficult task for Tesla’s Full Self-Driving (FSD) software to deal with scenarios in China where it is common that a large number of electric scooters are usually ahead in the same lane with motor vehicles,” Li Liyun, vice president of autonomous driving at Xpeng, wrote on June 27 on Chinese Twitter-like microblogging platform Weibo. Xpeng is planning to remove lidar from its upcoming sedan, scheduled for launch later this year, local media has reported.

Xpeng and NIO are also following Tesla’s suit by transitioning to an “end-to-end” autonomous driving method after using modular-based neural networks that are heavily reliant on explicit coding. Meanwhile, more traditional automakers are turning to domestic tech companies for help, such as Huawei and DJI, to catch up with the latest AI trend. Despite big challenges pressuring the industry, some early movers have the potential to compete with the US pioneer, Liu Guanghao, partner at Shanghai-based venture capital firm Befor Capital, told TechNode.

READ MORE: China opens door wider to Tesla as local giants disrupt the EV sector with AI-defined vehicles

Vision-based approach

A break from previous strategies of using expensive chips and sensors to enable ADAS capabilities, the latest approach centers on reducing the cost of components in hopes of making more room for further price cuts. Many now have their eyes on the use of radar, ditched by Tesla in 2021 due to limitations in identifying stationary objects with low image resolution, as some parts makers now said it is coming close to lidar in terms of performance – but at a lower price tag.

“There is more overlap between lidar and our sixth-generation radar systems as we significantly improve the resolution,” Juergen Brandl, Head of Market China, Business Area Autonomous Mobility at Continental Group, told TechNode. “Radar could soon see through [objects] but lidar has some problems with the distance especially in difficult situations like fog and rain.” The German firm’s newest front-facing radar boasts a detection distance of 280 and 140 meters (174 and 87 miles) for vehicles and slow-walking pedestrians, respectively.

Advanced radar solutions like this also create three-dimensional point cloud datasets like lidars, helping automakers and developers move towards fully end-to-end models with raw data collection from multiple sensors to train their self-driving systems. “We can play a big role in the development of Level 2 plus ADAS systems in China,” Brandl said, adding that the company’s product is below the price of a lidar unit with “very good“ output in terms of point cloud data.

Some disagree however, saying the technology is still at an early stage. Production of Continental’s latest-gen radar started early this year and the company began delivering the world’s first 4D radar in 2021, which detects an object’s vertical information in addition to distance, direction, and relative velocity, generating more dense point clouds than a contentional radar. Meanwhile, major Chinese lidar makers have consistently enhanced the performance of their products and lowered the prices to just over RMB 1,000 ($137.6) per unit in recent years.

The key is whether 3D/4D radar could prove to be a more “cost competitive” option compared with lidar, said Liu. “I believe both [radar and lidar] have their special advantages and disadvantages,” Brandl said. ”I think time or the market will tell whether we need both of them or one solution only.”

Either way, prospects for early players are bright. Momenta, a self-driving car company backed by General Motors and Toyota, expects the bill of materials, or total cost of components, for the City NOA (Navigation on ADAS) function to be slashed to RMB 4,000-5,000 from the current RMB 7,000-10,000 in the next two years. The intent is to survive an unprecedented price war in China that has been ongoing for more than a year.

End-to-end AI

Chinese carmakers used to brag about their coverage of cities where their assisted driving software is said to enable cars to handle on-ramp to off-ramp driving, automatic highway lane changing, and congested streets. However, many are now pivoting their focus to provide a more human-like driving experience and a full end-to-end AI model is now seen as key to winning the battle.

Described by Tesla chief executive Elon Musk as “basically photons in and controls out,” such end-to-end neural networks play an integral role in a vehicle’s decision-making process, taking raw sensor data as input and producing control actions as output. This contrasts with the conventional approaches that see each functionality, from perception to planning and action, developed individually using rule-based designs, which are often inadequate  in addressing the vast number of scenarios that occur on the road, a team of researchers said in a recent report.

The result is that people still feel their cars pilot themselves inhumanly even when kitted out with some of the most cutting-edge ADAS on the market, partly because human driving behavior tends to be consistent rather than discrete. It is very difficult for the current AV (autonomous vehicle) stack to make coherent, long-term decisions, said Wu Xinzhou, Nvidia’s vice president of automotive at its annual developer conference GTC in March.

Recent surveys have shown automakers that customers are generally dissatisfied with existing ADAS functions. Nearly half of respondents take the controls 1-2 times per 100 kilometers (62 miles) as the city NOA functions do not react appropriately, while others make more frequent interventions, according to a recent survey compiled by China’s Gaogong Industry Institute. “When the driver has to interfere pretty often, then you cannot say this is autonomous driving,” said Brandl.

Xpeng is aiming for less than one intervention per 1,000 kilometers in major traffic areas in China, CEO He Xiaopeng announced early this year, without giving a timeframe. This was followed by a new software update for its XNGP ADAS in May enhanced by the first end-to-end AI model for production vehicles in China, according to the EV maker. NIO reshuffled its autonomous driving department recently, rolling up its perception as well as planning and control teams into a single group with a focus on new AI models, Chinese media outlet LatePost reported on June 19.

It requires huge amounts of data, for example, millions of video clips, to train AI systems, as well as deep pockets and access to AI chips. Musk told investors in April that his company will increase the number of Nvidia’s flagship AI processors it uses from 35,000 to 85,000 by the end of this year. He wrote in a post on X earlier that month that the investment in training computers, gigantic data pipelines, and video storage will be well over $10 billion cumulatively this year.

Such major investment and effort is not something all companies can handle however. “It would be so hard for most traditional automakers to do this by themselves. The best way is to pick a supplier,” Liu said.

The entry of Tesla’s FSD into China may feel like a new challenge therefore, but it may also coincide with a new era of partnerships around self-driving technologies.

Chinese automaker Dongfeng Motor will introduce the first global model from its premium electric vehicle brand Voyah as early as the third quarter of this year, in what will be a direct challenge to Tesla’s best-selling crossover the Model Y, local media has reported.

The 100% electric compact sports utility vehicle (SUV) will feature Huawei’s vision-based approach to automated driving functions on urban streets. It will also be produced in a joint plant set up with Nissan, as fierce competition has put downward pressure on the Japanese firm, a regulatory filing has shown.

Why it matters: Stellantis and Nissan’s Chinese manufacturing partner expects the model to appeal to young families and drive sales in the country’s already crowded mainstream luxury EV segment, company sources told National Business Daily (in Chinese) on Tuesday.

Details: Sources added only full-electric variants will be on offer for the Zhiyin (知音), which loosely translates as “bosom buddies” in Chinese. This is in line with a filing that was published last month by China’s top industry regulator, which shows the car has six variants with battery options using two types of cathodes – more energy-dense nickel cobalt manganese (NCM) and cheaper lithium iron phosphate (LFP).

• The five-seater SUV measures 4.7 meters in length and 1.6 meters in height with a wheelbase of 2.9 meters, according to the registration filings released by China’s Ministry of Industry and Information Technology (MIIT). This is roughly the same size as the Tesla Model Y and smaller than the first model under the Voyah brand, the Free SUV.
• The all-wheel drive version will have a dual electric motor setup with both motors providing a power output of 160 kW, while other variants will be offered with either a 215 kW or 230 kW electric motor. All the variants have no lidar components, which are normally packaged on the car roof or near the bumpers to help the vehicle sense and interpret its surroundings.
• This means the new car will utilize the so-called “pure vision” approach, a strategy championed by Tesla, relying on cameras and artificial intelligence, rather than the more visible but more expensive laser sensor units, to achieve automated driving. Voyah announced in April at this year’s Beijing Auto Show its first global model would feature Huawei’s intelligent driving software and operating system HarmonyOS.
• Dongfeng will manufacture the car at its latest joint plant with Nissan, which opened last year in the central Chinese city of Wuhan with an annual capacity of 300,000 units. Nissan has been making several electric models including the Ariya with Dongfeng at this plant to date, although it sold less than 2,000 units of its second EV-only model in the country over the past six months, figures compiled by auto service platform Dongchedi showed.

Context: Nissan’s China joint venture with state-owned Dongfeng reported sales of 723,139 cars in the country last year, representing a 21.5% plunge from a year ago. The companies last November announced plans to export vehicles from China in 2025. The Japanese carmaker was also reportedly considering a 30% cut of its annual car output in China, along with peer Honda.

• Dongfeng’s Voyah in January announced a collaboration with Huawei to develop smart EVs. The brand has three models on sale and delivered 24,869 EVs, including all-electrics and plug-in hybrids, from January to May. The automaker’s off-road vehicle lineup Mengshi also works with Huawei.

Li Auto is on track to offer its new level 3 automated driving system within the next 12 months, which would enable hands-free, eyes-off driving under certain conditions, as it transitions towards “end-to-end neural network” architecture, according to chief executive Li Xiang.

Why it matters: The comments mark an increase in competition between Tesla and Chinese automakers over the adoption of autonomous driving. The US automaker is reportedly looking to roll out the latest version of its “Full Self-driving-Driving” software in China, its second-largest market.

• The news also comes as Li Auto started calling back some former staff late last month having previously dismissed roughly 30% of its 1,100-person Autonomous Driving Group in a setback to the company’s launch of its first battery electric model, the Mega. 
• Most employees rehired by the electric vehicle maker work on the NOA (Navigation on ADAS — advanced driver assistance system) function for urban driving scenarios. Vice President Lang Xianpeng on June 7 promised surviving employees from his team that the layoff was now over, sources familiar with the matter told the 21st Century Business Herald (in Chinese).

READ MORE: China opens door wider to Tesla as local giants disrupt the EV sector with AI-defined vehicles

Details: Li Auto will roll out a new version of its NOA software in the third quarter of 2024. This version still requires the driver to be in full control but will not rely on a high-definition map. The company announced that it will launch its city NOA function to users in 100 major Chinese cities in November.

• This will be followed by a wide release of its Level 3 autonomous driving system later this year, or in early 2025, Li told the audience at the China Auto Chongqing Summit on June 8, according to short videos posted on the social media platform WeChat.
• Li said the company has been working on an end-to-end artificial intelligence approach since late 2023. The approach is using video clips from real driving situations as well as machine learning algorithms to handle the vehicles’ control. It is also exploring Visual-Language Models (VLMs), which could better learn patterns and correlate between visual data and natural language than existing offerings.
• Li Auto’s L3 conditional autonomous system will use both approaches, Li said, allowing the human driver not to monitor the vehicle constantly but to be ready to take control, according to the definitions set by the Society of Automotive Engineers (SAE). It could further evolve to a level 4 highly autonomous driving system over the next three years, Li added, without elaborating further.

Context: Automakers in China, including Li Auto, have either been applying for mapping licenses or working with map services in order to collect large amounts of highly detailed geographical data, which helps autonomous vehicles (AVs) make better decisions when performing driving tasks. However, many of them, such as Huawei and Xpeng Motors, are now looking for alternatives due to cost and complexity reasons.

• It is unknown whether Li Auto will take legal responsibility for its upcoming level 3 technology. Last year, Mercedes-Benz became the first automaker allowed to sell cars with its L3 Drive Pilot system in California and Nevada, Reuters reported. China is pushing for similar regulatory initiatives with assistance from automakers including BYD, SAIC, and NIO.
• Li Auto has been working on layoffs since May to cut 18% of the company’s workforce, which stood at roughly 5,600 as of last year, while that percentage for the Autonomous Driving Group was more than 30%, according to the 21st Century report. President Ma Donghui told investors in November the company planned to have 2,000 employees for AVs by the end of 2024. 

READ MORE: Li Auto accelerates assisted driving tech competition amid launch of first battery EV

The first electric vehicle model of NIO’s lower-priced, family-oriented sub-brand Onvo finally made its debut in Shanghai on Wednesday, the International Day of Families, after three years of development. The coupe-style sports utility vehicle was surrounded at its unveiling by not only hundreds of journalists but also Chinese parents who took their children to the event, and who the company hopes will form the very first owners of the long-anticipated car.

The Chinese EV maker on Wednesday began taking reservations for the Onvo L60 at a pre-sale price tag of RMB 219,900 ($30,471), cheaper by RMB 30,000 than the entry-level Tesla Model Y but with a longer driving range and a roomier interior space, among other advantages. The SUV, scheduled for launch in September, is expected to be priced at under RMB 200,000, and come with NIO’s Battery-as-a-Service (BaaS) program, in which customers buy a car and pay for a battery rental service, similar to the company’s scheme for NIO branded cars.

Confident in the strong competitiveness of its offering and a large customer base in China, NIO management said it has “high expectations” for Onvo, which is short for “on voyage” in English. Its Chinese name Ledao translates as “path to happiness.” Speaking to reporters after the debut on Thursday, William Li, founder, chairman, and chief executive of NIO, shared additional details about NIO’s ongoing partnerships for battery swapping with some of China’s most established car manufacturers, including Geely, Changan, and GAC. 

First Onvo EV

NIO’s first mainstream crossover, with a similar shape to its more premium siblings, competes with Tesla’s Model Y, the world’s top-selling electric SUV, from nearly every perspective. The entry-level L60 comes with a driving range of 555 kilometers (345 miles), a bit longer than that of the rear-wheel-drive Model Y (554 km), while the other two variants travel more than 730 km and 1,000 km on a single charge, respectively. 

The company also said that the L60’s outstanding wind resistance, measured by a drag coefficient of 0.229, increases its effective range, while a 900-volt electrical system reduces energy loss to heat, giving it a longer range. The total energy consumption of the car is 12.1 kilowatt-hours (kWh) per 100 kilometers, compared with the 12.5 kWh achieved by the Model Y, according to NIO. 

Li said that NIO will make a “pretty decent” gross margin from the L60, by focusing on core values that matter to Chinese families, rather than producing a vehicle with a dazzling array of unnecessary specs, a move that keeps the car’s overall costs under control. Cost savings also come from NIO and Onvo sharing research and development costs, among other synergies.

The Onvo L60 embraces the vision-based approach advocated by Tesla, which uses cameras and artificial intelligence and gets rid of lidar sensors for autonomous driving. It also uses a smaller and more affordable lithium iron phosphate (LFP) battery pack from BYD, according to Reuters.

Battery swaps

NIO also provided details about its extensive power infrastructure network, which is claimed to be the largest of its kind in China with more than 2,415 battery swapping stations as of Wednesday. It has long been a loss-making effort for the company but is now emerging as an attractive option for charging availability and cost reduction for a growing list of Chinese auto giants. 

Onvo owners will only be able to use the company’s newer swap stations (the third- and fourth-generation ones) and will share the facilities with NIO’s partners. Each of the newer swap stations, which can hold more than 20 battery packs of different sizes, can offer more than 400 swaps per day, with each pack being used roughly up to 20 times. Li mentioned the company’s plan to charge partners a service fee of roughly RMB 20-30 per swap. NIO itself completed nearly 70,000 battery swaps per day as of May 8. 

Meanwhile, NIO owners can “refuel” their vehicles at any swap shop, giving them more charging availability and a premium user experience. The company said it is on track to build 1,000 battery swap facilities on its own this year and expand the network for NIO and Onvo from the 2,316 stations available as of the end of last year, while its partners are also set to provide additional resources. 

Li added that he envisaged there being six to seven EV battery sizes at most over the long term, with partners set to use batteries of the same kind as Onvo. NIO currently uses batteries in four sizes with an energy density ranging from 70 to 150 kWh.

READ MORE: Drive I/O | Big bets on battery swaps

Tesla claimed on its Chinese app on Monday that it would be rolling out its full self-driving (FSD) system “very soon” rather than the previously stated “a bit later” (our translation). The development comes as the US automaker reportedly received tentative approval from Chinese authorities to deploy its advanced driver assistance systems (ADAS) in the world’s most competitive electric vehicle market following a surprise visit to Beijing by chief executive Elon Musk. 

Local customers have expressed mixed views about the potentially game-changing technology finally coming to their country, although many Tesla owners and loyal fans welcomed the long-overdue launch on social media. The technology may not yet work well in Chinese urban driving scenarios, despite costing RMB 64,000 ($8,838); most competitors such as Xiaomi promise to offer similar functions free of charge, Li Yang, a Model 3 owner in Shanghai, told TechNode on Tuesday.

Major Chinese EV players Huawei and Xpeng Motors, who specialize in autonomous driving, have responded in a relaxed way to what is seen by many as a major win for their US rival. 

“We are confident that Huawei’s intelligent driving system is the best in the world,” Richard Yu, Chief Executive of Huawei’s Consumer Business Group, said at a press event last week as the company introduced the redesigned Aito M5 crossover. Aware of Tesla’s release last month of the no-longer-in-beta FSD software, Yu said the company has closely studied its competitor’s technology by taking test rides in San Francisco, among other US cities. 

A long-awaited battle: Experts told TechNode that the US and China have been in a neck-and-neck race to develop “end-to-end neural network” technology, seen by experts as the biggest difference between Tesla’s FSD v12 software and earlier versions.

• Tesla chief executive Elon Musk stated publicly in November that its most updated, end-to-end FSD uses machine learning rather than rules crafted by programmers to absorb driving behaviors from large datasets. The approach reduces the carmaker’s reliance on coding, expensive sensors, and high-precision maps for exact road information.
• “It is more of a 1,000 flowers blooming type of situation here in China, while the US market is guarded by a sole player,” Paul Gong, head of China autos research at UBS, told reporters on Monday in Beijing, referencing a Chinese idiom. Both the US and China are at the forefront of the technological trend, followed by Europe and the rest of Asia, Gong added.
• Although Tesla has huge amounts of data, Chinese firms have the potential to quickly catch up by ramping up their test fleets, according to Song Chaozhong, CEO of Chinese self-driving car startup eCHIEV. Song cautioned that real, “human-like” autonomous cars are still far away, despite the technology likely being a positive driver for vehicle sales. 

A booming segment: Analysts expect Tesla’s imminent launch of the FSD software in China to not only help restore Beijing’s image among foreign investors in general, but also help China extend its lead in the adoption of driver-assist technologies worldwide, and enhance its reputation as a rising powerhouse of next generation cars. 

• China has become a source of innovative engineering for global automakers, as they are now leveraging the capabilities of Chinese companies to design global vehicle offerings, Gong said when asked his view of this year’s Beijing Auto Show. He added China’s push to open its industries to greater foreign investment and competition has been “underestimated.”
• Tech-literate Chinese EV makers are poised for explosive growth thanks to growing interest from Chinese consumers in intelligent driving features, as Elon Musk earlier predicted a “ChatGPT moment” was coming for his company. Huawei on April 24 estimated that more than 500,000 cars on Chinese roads will feature its advanced driving system by year-end. 
• Xpeng is scheduled to launch its affordable brand MONA in June with the first model expected to feature an autonomous driving system without the integration of lidar and aiming for annual sales of 100,000 units. Global shipments of vehicles with ADAS features, known as Level 2 plus autonomy, could more than double to 4.5 million units this year, with China leading the game, research firm Canalys wrote in December.

Context: Tesla’s China breakthrough comes after US Secretary of State Anthony Blinken on April 25 called on China to provide a level playing field for American businesses during his second visit to China in less than a year, Reuters reported. 

• The US automaker has reached a deal with Baidu to localize its driver assistance solution with map and geographical information from the Chinese tech giant, according to Bloomberg.

Global carmakers from Volkswagen to Toyota are introducing new models at the Beijing Auto Show 2024 with the help of Chinese tech companies in an effort to defend market share amid a major shift to electric vehicles led by local car giants. 

The biannual trade event, which on Thursday witnessed a return to pre-pandemic attendance levels after a brief pause in 2022, also represents another landmark moment for the Chinese EV sector where domestic players are once again on the offensive with an array of new models. A similar event in Shanghai a year ago reportedly prompted the industry’s legacy players to either increase their efforts or rethink their brands in order to adapt to the changes.

Below, TechNode provides a summary of some of the biggest releases from both international and Chinese automakers, including BYD, GAC, Geely, Honda, Toyota, and Volkswagen. There are also some notable updates from younger players such as Xiaomi, NIO, and Xpeng, which might give a clue as to where the most competitive EV market in the world is heading. 

BYD

China’s biggest EV maker on Thursday unveiled a higher-end variant of its Qin vehicle, the top-selling compact sedan in the country last year. The new car is scheduled for launch in the second quarter with an expected price tag of RMB 120,000 ($16,560). The Qin L measures 4.8 meters in length and spans a 2,790-millimeter-long wheelbase, placing it between the Qin Plus and the Han in terms of size. It features the company’s next-generation plug-in hybrid platform DM-i 5.0, which could suggest an improvement in range and fuel efficiency. The company also introduced the Seal 06, a plug-in hybrid EV under the Ocean lineup which is about the same size as the Qin L but loaded with more stylish design language to attract younger customers. 

GAC

Aion, the third best-selling EV brand in China last year after BYD and Tesla, showcased its first global model, replete with modern technologies and angular styling, as its state-owned parent beefs up its strategy to woo customers worldwide. GAC said its all-new Aion V, scheduled for launch in July, will maintain a driving range of over 750 kilometers (466 miles) even when the mercury dips to -30 degrees Celsius, and offers a large interior space comparable to the likes of the BMW X5. The all-electric sports utility vehicle, which incorporates traditional Chinese dragons into its design, can navigate varied urban environments worldwide with features such as lane switching by utilizing advanced artificial intelligence algorithms to process sensor data instead of high-precision maps, the company said.

Geely

Volvo’s parent showed its ambition to become a disruptive force in the global automotive industry with the debut of what it described as the world’s first production model with two sliding doors and front swivel seats. Geely has taken a radical approach to how EVs are put together, giving the 4.7 meter-long Zeekr Mix an extended wheelbase of three meters achieved through a more compact electric motor, shorter front overhangs, and repositioning of the air conditioning system, among other components. This, along with the front seats that can rotate 270 degrees, would allow kids to play or families to dine together in a 1.5 square meter interior flat space. The five-seater multi-purpose vehicle, offering a 1.5 meter width opening area for passengers, targets three-generation Chinese families, especially those with elders and pregnant mothers.

Honda

Japan’s Honda on Thursday began selling its second all-electric model with time-limited discounts in China in the company’s latest effort to boost sales. The move comes after entrenched rivals such as BYD and Tesla recently rolled out more price cuts amid slowing growth. The e:NP2 SUV has a driving range of 545 km at a price tag of RMB 159,800, providing buyers with a RMB 30,000 reduction compared to its original plan, according to Li Jin, a deputy general manager of Honda’s China joint venture with GAC. Honda also debuted the Ye, a new series of all-electrics with technologies sourced from Huawei and iFlyTek among other Chinese tech firms, as part of its plan to sell only EVs in China by 2035.

Toyota

Toyota said on Thursday it will integrate lidar sensors into its two upcoming models under the “Beyond Zero” (bZ) all-electric series, as the world’s top-selling automaker looks to provide consumers with the same level of assisted driving technology as Huawei and Xiaomi. The bZ3x and the bZ3c compact crossovers will be able to automatically change lanes, and enter and exit Chinese highways when they go on sale within the next 12 months. Toyota also announced it is exploring the uses of generative AI in collaboration with Tencent, as Chinese consumers expect their future vehicles to be more capable and personalized. This follows reports that the Japanese giant is using Huawei components to enable autonomous driving functions on its China-made EVs. 

Volkswagen 

Germany’s biggest carmaker participated in Auto Beijing 2024 with major global debuts including the ID.Code concept – which offers a glimpse into its upcoming, China-specific all-electric lineup ID.UX – as well as the Audi Q6L e-tron, the first production model based on its PPE electric platform. The coupe-styled ID.Code will be equipped for highly autonomous driving and come with a sophisticated AI assistant with contributions from local designers, as Volkswagen plans to introduce the first model under the new series later this year. In addition to partnerships with Xpeng and Horizon Robotics, the automaker confirmed it is working with Chinese tech giants including DJI, as its latest Tiguan L SUV now features an advanced driver assistance system (ADAS) sourced from the drone maker.  

Xiaomi, NIO, and Xpeng

Xiaomi was the center of attention on Thursday when the Chinese smartphone giant said it had secured 75,723 reservations with non-refundable deposits for the SU7, its first EV, with a competitive price range between RMB 215,900 and RMB 299,900. Chief executive Lei Jun expects monthly delivery to exceed 10,000 units in June and the company is set to reach a milestone with 100,000 EV deliveries by this year, which would be a record speed for any Chinese EV brand. The 55-year-old entrepreneur is an icon in the Chinese tech and auto industries, with his visits to rivals’ booths becoming one of the hottest topics at this year’s Beijing Auto show.

Xpeng Motors could take on its major frenemy with the mainstream brand MONA, short for ‘Made of New AI,’ CEO He Xiaopeng told reporters during a press conference. 

Meanwhile, one of NIO‘s new affordable brands, called ONVO, is scheduled for launch in the second quarter of this year. The luxury EV maker on Thursday launched a redesigned version of its ET7 sedan with a starting price of RMB 428,000, which is RMB 20,000 lower than its original version launched three years ago.

READ MORE: Huawei, Xiaomi, and Geely’s new EVs have details leaked on Chinese government site

Hundreds of people flocked to a Xiaomi store in the southern Chinese city of Guangzhou in the late hours of Thursday evening to be among the first to take a look at what many deem to be the electronics brand’s more affordable alternative to a Porsche, as the brand announced a lower than expected starting price for its debut electric vehicle. That’s what TechNode observed during a livestream broadcast by a Chinese electric car blogger on social media app WeChat that attracted more than 200,000 viewers within an hour. 

Xiaomi has already enjoyed a debut win after securing a record 50,000 pre-orders in just a few minutes following its SU7 EV launch event. The official livestream racked up nearly 43 million views on microblogging platform Weibo, underscoring the overwhelming interest among tech-savvy Chinese consumers in the company’s first car. Those making a reservation were asked to pay a RMB 5,000 ($692) deposit as part of the process, with the tech giant expressing its thanks to customers who did so in a brief statement on the microblogging platform Weibo (in Chinese). 

The all-electric sports sedan is selling at a lower than expected starting price of RMB 215,900 ($29,881), roughly $4,100 cheaper than the popular Tesla Model 3, while touting better performance from driving range to acceleration. The dual-motor all-wheel drive version competes with the Porsche Taycan with a top speed of 265 kilometers (165 miles) per hour at a price tag of only RMB 299,900.

Xiaomi’s car launch contrasts markedly with Apple’s surprise retreat from the EV landscape, after the iPhone maker reportedly scrapped its decade-long effort to make a car recently. “We will provide every user, including those with Apple devices, a smart and connected life experience everywhere, creating seamless integration in their homes, cars, and beyond,” Xiaomi chief executive Lei Jun said during the press conference (our translation). 

“A trump card”

Lei, the 55-year-old serial entrepreneur dubbed “China’s Steve Jobs”, tried to lure users away from traditional carmakers during the two-hour event by showcasing how Xiaomi’s ecosystem would provide universal connection and integration between different devices, including phones, cars, and gadgets at home. 

Xiaomi essentially promised potential buyers that their devices would be all tied together with a click, swipe, or a simple voice command. The car’s air conditioning will cool the interior down on a hot summer’s day once the owner tells a home speaker what temperature they want before even leaving the house, according to one example given. In another, the car’s dashboard could become a centralized command station for home accessories which will be activated as the driver approaches home. 

Although rival Huawei has touted similar efforts with its EV partners, Xiaomi claimed last November that more than 655 million devices have been connected to its IoT (Internet of Things) platform, from televisions to fitness bands, making it the biggest network of its kind worldwide. “This is a trump card from Xiaomi,” said Lei when discussing the linking of the brand’s new EV with its IoT network. 

“Better than Tesla”

Meanwhile, Lei mentioned Xiaomi’s plans to be “among the top-tier players” in autonomous driving, a field where Tesla already stands out as a pioneer globally and Huawei is establishing its name at home. The company said its EVs are already capable of traveling more than 300 km on average autonomously on Chinese highways before human drivers take over and will be able to complete most trips by themselves on urban streets across China by August. 

Xiaomi is moving towards two distinct approaches by working on both a camera-based computer vision system and another advanced driver assistance system (ADAS) that relies on more sensors including lidar. The company said it will exclusively employ Nvidia’s cutting-edge chips for both systems and bring the software development completely in-house to ensure timely over-the-air updates across all its car variants. 

“In China, Tesla vehicles will not be as good as the SU7 when it comes to intelligent driving capabilities,” said Lei, adding that customers who placed their orders before the end of this year will get the software free of charge. Tesla currently charges Chinese buyers RMB 64,000 for future access to its full self-driving (FSD) package, despite it remaining unavailable in the country. Still, it is faced with competitors from BYD to Geely which also look to offer customers highly automated features with their premium EVs.

READ MORE: Key takeaways from Xiaomi’s EV pre-launch: A top offering facing a tough test

Details of new electric vehicle models from Chinese auto and tech majors including Huawei, Xiaomi, and Geely have been leaked online via an official regulatory process. Some are expected to make their debut at the upcoming Beijing Motor Show next month, positioned to compete with models from dominant rivals such as BYD and Tesla, and potentially stirring up a new price war in the world’s biggest auto market.

The companies expect the upcoming models, now making a splash online, to become bestselling or otherwise strategically important cars for their brands. Below are highlights from the registration filings released for public review by China’s Ministry of Industry and Information Technology (MIIT) on Tuesday, giving critical details of the models ahead of their official launches. 

Stelato S9

The S9 will be the first model under the new premium Stelato brand launched in partnership between Huawei and China’s BAIC and the largest sedan model of all Huawei-enabled EVs to date. The car measures 5.1 meters in length and 1.5 meters in height with a wheelbase of nearly 3.1 meters, offering passengers a spacious and comfortable interior in an effort to draw in affluent Chinese consumers. 

The all-electric executive sedan will be available in single and dual-motor variants producing 308 and 524 horsepower respectively, and will include innovative elements such as a camera-based digital rear-view mirror system as an optional add-on, according to the filings. Analysts expect the car to be launched in June for between RMB 300,000 and RMB 500,000 ($41,730-$69,550). Shares in partner BAIC Bluepark surged 32% on the mainland Chinese stock market on the news over the week. 

Zeekr Mix

China’s Geely is raising its bet on the small but growing segment of multi-purpose vehicles with its upcoming roll-out of the Zeek Mix, an all-electric five-seater van, after the launch of its larger and more business-oriented Zeekr 009 nearly two years ago. The pictures published by MIIT show a mid-size MPV with a rounded exterior and low center of gravity as well as an optionalLIDAR unit mounted on the car’s roof for automated driving. 

It is slightly shorter than the Zeekr 007 sedan at nearly 4.7 meters in length, likely making it easier to maneuver and attractive to parents, while offering a larger interior with a 3,008-millimeter-long wheelbase. The single-motor car has a 422-horsepower electric powertrain – higher than the plug-in hybrid Denza D9 from BYD, currently a top-seller in the market, but less powerful than bigger offerings such as the Xpeng X9 and the Li Auto Mega. 

Xiaomi SU7

Xiaomi on Wednesday received Chinese government approval for a new variant of its first EV, the long-anticipated SU7, equipped with lithium iron phosphate (LFP) batteries sourced from CATL and roughly 110 kilograms heavier than the one powered by BYD’s iron-based batteries. Speculation has circulated that the new power option could be CATL’s Shenxing batteries, which have boasted of a high energy density for a longer driving range and an 800-volt electrical system for faster charging compared with existing offerings. 

China’s industry regulator had previously uncovered details about another entry-level SU7 and the more premium SU7 Pro/Max, which the company claimed could accelerate from 0 to 100 km/h (62 mph) in 2.78 seconds and would be more aerodynamic than rivals’ offerings including Tesla’s Model S. Chief executive Lei Jun said on Friday that the smartphone maker will begin deliveries immediately on March 28, when pricing of the sports sedan will be finally announced.

READ MORE: Explainer: How a new round of price cuts are reshaping China’s EV market

Huawei and Chinese automaker BAIC are aiming for monthly sales of their first joint electric vehicle model under a new brand to exceed 10,000 units, as they also set their sights on markets in Asia and Europe, Chinese media has reported. 

Why it matters: The launch of the new car brand with BAIC, named Stelato and scheduled to be unveiled in April at this year’s Beijing Motor Show, would be another test for Huawei as the smartphone maker has been looking to grow its automotive business after being hit by US sanctions. 

• Huawei’s track record in EVs has so far been mixed. It has managed to bring the struggling Aito brand back to life with manufacturing partner Seres following the successful launch of the redesigned M7 plug-in hybrid crossover in September. However, the first Luxeed EV was a setback for the tech giant and its partner Chery, as buyers were confounded by delivery delays stretching for months. 

Details: Huawei and BAIC Bluepark, a mainland-listed subsidiary of the state-owned automaker, expect monthly sales of the first model under the new Stelato brand to achieve more than 10,000 units, according to an internal memo seen by Yicai (in Chinese). 

• According to the report, the executive sedan will be priced between RMB 300,000 and RMB 500,000 ($41,730-$69,550), and is scheduled to debut next month at this year’s Beijing Motor Show; it would then officially launch in June. 
• BAIC, a main partner of Daimler in China, is planning to create a capacity of around 300,000 EVs for the Stelato project, with more high-end EV models from sports sedans to SUVs in the pipeline, the report said. 
• The automaker last month told investors it is setting up a sales and service network for overseas markets, especially Europe and Asia. An existing partner to Huawei, BAIC sold fewer than 50,000 units of its Arcfox-branded EVs from 2021 to 2023. 
• Meanwhile, Huawei is set to launch its first model as part of a separate partnership with JAC Motor by the end of this year. The two companies plan to produce 35,000 units of the multi-purpose vehicle annually in a new plant scheduled for completion in 2025, according to an environmental filing published on March 1. 

Context: More Chinese automakers, such as state-owned Changan and Dongfeng, are joining Huawei’s expanding “Harmony Intelligent Mobility Alliance” in hopes of re-creating their brand images with Huawei’s smart cockpit and automated driving technologies, and leveraging its sales network.

• Aito for now remains the biggest success story of Huawei’s push into EVs, as vehicle deliveries grew 24% year-on-year to 94,380 units in 2023. Sales further increased to more than 54,000 units during the first two months of this year when most automakers reported flagging sales.

Chinese electric vehicle maker Zeekr began selling a redesigned version of its best-selling 001 on Tuesday with a slew of performance features including an extended driving range and faster charging speed at a cost of RMB 269,000 ($37,364).

Geely Automobile’s Zeekr is strengthening its offerings with an enhanced advanced driver assistance system, or ADAS, which will enable hands-free driving functions on Chinese motorways, in the face of competition from Tesla and home rivals such as Xpeng, Huawei, and Xiaomi. 

Why it matters: The new 001 is the latest example of how Chinese automakers are stepping up a months-long price war by offering ever-increasing high-tech, luxury-style features for the same price or less, as the world’s largest EV market consolidates.

• The Zeekr 001’s first full model change in three years, with more than half of the components refreshed or replaced, has pushed costs up by RMB 50,000 per unit, but the company has kept the starting price the same. The automaker cut prices of the model by around 10% from its original price in August 2023. The Zeekr 001 accounted for 60% of the brand’s sales last year.

Range boost: At a press event at Zeekr’s headquarters in Hangzhou, company officials emphasized how the all-new 001 beats rivals with its large battery packs and high-voltage architecture, providing more range and allowing for faster charging. 

• Zeekr offers two standard versions of the revamped 001. A rear-wheel drive model is powered by a 100 kilowatt-hour (kWh) Qilin battery pack which provides a driving range of 750 kilometers (466 miles) between charges, while an all-wheel drive variant can reach 675 km on a single charge with 95 kWh Shenxing batteries. 
• Both battery types, sourced from CATL, boast an extremely fast charge. Owners can add 525 km of range in 15 minutes with the Qilin batteries which charge at a 4C rate. The latest-gen Shenxing supports vehicles to run 256 km after a five minute charge at a 5C charge rate and needs 30 minutes to go from 10% to 80% at a temperature of minus 10 degrees Celsius.

Self-driving improvement: Zeekr integrated the revamp with roof-top lidar as standard, along with high-definition cameras and ultrasonic radar sensors around its perimeter, allowing the vehicle to spot objects 200 meters away, a combination intended to avoid complex situations and possible collisions. 

• The sports wagons will feature an upgraded version of a feature called Navigation Zeekr Pilot (NZP), which would take over basically all driving functions on the motorway, such as changing lanes and overtaking cars, according to the company. It is set to be available to drivers in 60 major Chinese cities by March and 90% of the country by June. 
• Zeekr expects to start inviting drivers to beta test a more advanced version of the ADAS software by June, allowing them to navigate around complex city streets without moving the steering themselves. The EV maker has embarked on a hiring spree for advanced driving with plans to increase the unit’s headcount from 800 to 2,000 by year-end. 

Market outlook: Management expects the all-new Zeekr 001 to contribute at least half of its sales in 2024, with monthly deliveries reaching 10,000 units per month, Jason Lin, a vice president, told reporters on Tuesday. The company will commence delivery on Friday with a goal to deliver 230,000 EVs this year, nearly a 100% increase from last year, Reuters reported. 

• The battery EV competes with crossovers and sports sedans such as BYD’s Denza N7, as well as the upcoming Xiaomi SU7. Rivals from BYD to Great Wall Motor have been ramping up efforts to develop self-driving technology, a drive that has been led by Huawei and Xpeng in China. 
• Lin was bullish on the firm’s outlook, adding that the company will expand its retail network to 525 shops by year-end, adding 100 or so flagship showrooms in first- and second-tier cities and another 100 smaller ones in lower-tier cities. Zeekr currently has four models on sale, including the cheaper sedan 007, the multi-purpose van 009, and a compact crossover.

READ MORE: Geely to sell its Zeekr electric cars directly to customers

NIO Capital, a venture capital firm founded by Willliam Li, chief executive of the namesake electric vehicle maker, has raised a new China-focused fund of more than RMB 3 billion ($416.8 million), despite a global market lull and domestic economic challenges. 

Why it matters: The fundraising milestone will allow NIO Capital to further explore the “transformative potential of innovative technologies in the automotive and energy sectors,” said Ian Zhu, a managing partner at NIO Capital, in a Monday announcement. 

• The NIO-related growth-stage investor has backed some of the rising stars in the Chinese automotive world, from battery material manufacturer Ronbay to self-driving car startups Momenta and Pony.ai. It has also invested in battery giant CATL. 

Details: The deal shows the strength of NIO Capital’s ties with its limited partners, which include venture capital investment guidance funds set up by Chinese regional governments, national funds, family offices, and listed companies, according to the announcement, which did not provide further details. 

• The Shanghai-based venture firm said it will maintain its focus on areas such as auto tech and sustainable energy with the new RMB 3 billion fund, which brings the total assets under its management to about RMB 15 billion. 
• Speaking to Chinese reporters, a NIO Capital representative stressed the fund’s independence from NIO in its final investment decision-making. “We don’t require that a parts maker only supplies to NIO,” the person said as an example of its open-mindedness, 36Kr reported.

Context: The deal comes as global private investment remains soft due to interest rate hikes and economic headwinds. 

• Fundraising totals were on track to drop 28% in 2023 compared with 2022, and the number of funds could decline by almost half, US management consultancy Bain & Company said in a note dated July 17. 
• In China, roughly 5,300 private capital funds raised a total of RMB 1.35 trillion during the first nine months of 2023, down 2.1% and 20.2% from a year earlier, according to consultancy Zero2IPO (in Chinese).

The Chinese electric vehicle segment briefly lost momentum in January as a majority of automakers reported a significant sales drop on Thursday during the traditional low season, a contrast to December when big promotions and exciting discounts gave a short-term sales boost at year-end. 

Why it matters: The decline was especially marked for BYD, which accounted for nearly a third of the country’s green energy vehicle sales last year. The biggest Chinese EV maker maintained its leading position with sales of more than 201,000 units last month, although that number represented a 41% decrease compared to December. 

• Geely and Huawei-backed Aito are among the few bright spots that have bucked the trend with their new models, becoming potential challengers to BYD’s dominance. Meanwhile, international carmakers such as Volkswagen and General Motors posted decent sales with local partner SAIC, showcasing their efforts to play catch-up. 

Ups: Geely posted its best-ever month with sales of 65,826 fully electric and plug-in hybrid vehicles last month, roughly 5,400 units more than in December and nearly six times greater than what Volvo’s parent achieved a year ago. This growth was partly due to strong sales of its Galaxy and Lynk & Co brands, which sold 19,223 and 28,176 units over the month, respectively. 

• Aito also delivered a record 32,973 cars in January. That is in line with the company’s outlook and stands in stark contrast to the same month last year when it reported deliveries of only 4,475 vehicles. The brand, launched by Huawei and its manufacturing partner Seres, said on Thursday (in Chinese) that it has secured 140,000 non-refundable orders for its redesigned M7. 
• Avatr, another Huawei-backed and Changan-owned EV brand, also posted improved sales volume for January buoyed by the positive demand for its second model launched in November. However, an ongoing shortage of Huawei’s key components may add uncertainty to the outlook of the premium brand and its peers, as reported by Reuters. 

Downs: On the other hand, China’s US-listed EV trio are the ones under pressure by reporting their lowest monthly deliveries since June, as they engage in a relentless price war with larger tech and auto forces. Both Li Auto and NIO slashed prices on current lineups last month as they prepare to launch revamped models in March, potentially causing some customers to postpone purchases. 

• Sales of Aion, which has sold a large proportion of its vehicles to ride-hailing fleets, also dropped 46% month-on-month to 24,947 units in January. The EV maker, affiliated with state-owned automaker GAC, is now ramping up efforts to go abroad, as sales in the overseas markets reached the threshold of 3,000 units for the first time last month. 
• Deliveries of Great Wall Motor and Deepal, another Changan subsidiary, also decreased by 16.2% and 7.1% to 24,988 and 17,042 units respectively, from December. Meanwhile, IM Motors, a premium EV brand launched by China’s SAIC, saw a bigger drop with its January number more than halved from a month earlier to 5,305 units.

Context: China’s sales of new energy vehicles, mainly all-electrics and plug-in hybrids, increased 92% year-on-year from Jan.1-28 partly due to the year-ago low base effect marked by a wave of Covid cases after Beijing dismantled pandemic controls in December 2022. However, that number was 24% down compared with December when most automakers made a year-end sales push, figures from the China Passenger Car Association (CPCA) showed.

BYD and FAW Group are aiming to invest in DJI’s automotive business unit – one of the few Chinese companies capable of developing partially automated driving software – as part of the latest effort by traditional automakers to catch up with rivals such as Tesla, local media has reported. 

Why it matters: The news comes at a time when a growing number of automakers and suppliers are expanding their alliances in hopes of accelerating progress in and sharing the cost of making partially automated driving passenger cars. In China, the technology is being popularized by the likes of Tesla, Huawei, and Xpeng Motors. 

• It is also the latest example of Chinese auto majors taking steps to turn their business partnerships with tech companies into deeper capital relationships. Changan Automobile recently announced it will invest in a new venture set to absorb Huawei’s car business unit. 

Details: BYD and FAW, a manufacturing partner of Volkswagen and Toyota in China, recently conveyed their message to DJI Automotive, the car business unit of the namesake drone maker, 36Kr first reported on Wednesday (in Chinese). Citing people with knowledge of the matter, the report did not put a figure on the planned investment.

• BYD is planning to roll out partially automated driving functions for future affordable models priced at RMB 200,000 ($27,920) or below in 2025 with assistance from DJI, a person close to the company told TechNode on Thursday. 
• DJI is appealing to original equipment manufacturers (OEMs) for its low-cost advanced driver assistance system, as they have been struggling to lower costs and repeatedly cut prices of their cars, another source with direct knowledge of the matter told TechNode. 
• DJI, China’s biggest drone maker, stated that its technology suite, including an affordable computing chip from Texas Instruments (TI) with only a few cameras, could enable cars with automatic lane changing and on-ramp to off-ramp functionalities on Chinese expressways. 
• Some EV makers, including NIO and Li Auto, have developed similar functions based on Nvidia’s more expensive DRIVE Orin processors, with each offering 254 trillion operations per second (or TOPS), compared with the 32 TOPS offered by a top-end TDA4 chip from TI. 
• “While BYD is not the forerunner in autonomous driving, it has the ability to be a fast follower which will likely meet the needs for the majority of Chinese consumers,” Bernstein analysts wrote in a Jan. 22 note, citing economies of scale, which allows it to collect more data for software training from its large fleet, as one of the reasons.
• BYD and DJI did not respond to TechNode’s requests for comment. 

Context: Shenzhen-headquartered DJI separated its car business into an independent company in late 2022 and became open to external funding with a target valuation of $1.5 billion, Chinese media outlet Leiphone reported in August. 

• The drone giant has worked with SAIC-GM-Wuling (SGMW), a General Motors China joint venture, since 2019, with the first model integrated with its automated driving technology, the Baojun Kiwi EV, going on sale with a price tag of RMB 102,800 in September 2022. 
• BYD showcased a Yangwang luxury off-roader that integrates an unmanned aircraft on the car’s roof in a collaboration with DJI during a press event on Jan. 16. It also announced plans to release at least 10 new models this year featuring advanced driving technology.

READ MORE:  BYD’s Denza launches cheaper driver assistance system with Nvidia amid rising competition

Huawei and Dongfeng Motor, a Chinese manufacturing partner of Stellantis, are in an ongoing collaboration to develop smart electric vehicles, the companies have announced. This adds to a string of such deals by technology giant Huawei as it accelerates its entry into the auto market. 

The partnership could help Voyah, a subsidiary of state-owned automaker Dongfeng, increase sales and expand its presence in the red-hot EV market where a wave of consolidation and reshuffling is underway, according to David Zhang, a visiting professor at Huanghe Science and Technology University. 

Why it matters: The alliance is the latest example of Huawei’s multifold endeavor to expand into EVs. It has pushed two initiatives to enhance cooperation with carmakers in particular. 

• One is the so-called “Huawei Inside (HI)” business model, signifying that cars will feature Huawei’s full-stack technologies such as automated driving software and infotainment systems. Changan Automobile and Mercedes’ Chinese partner BAIC are among its partners. 
• The other is “Smart Selection,” in which Huawei not only provides technologies but also sales channels while gaining control over vehicle development. Aito has been the biggest success story under this approach, followed by the recent launch of Luxeed between Huawei and Chery. 

Details: According to Zhang, Huawei will adopt the HI approach with Dongfeng, mainly selling the carmaker components and software, and will probably not go into as much depth as it did with Seres. 

• This would allow Dongfeng, controlled by the State-owned Assets Supervision and Administration Commission, China’s state asset regulator, to maintain control of its premium EV brand, Zhang said. 
• As more EVs embrace cutting-edge technologies, an automated driving system powered by Huawei could be a big selling point for Voyah, elevating it above other brands. The tie-up could see the companies share development and marketing costs, Zhang added. 
• In a Monday announcement (in Chinese), Huawei and Dongfeng said they will move forward with the large-scale adoption of intelligent technologies with the joint development of new cars, without giving further details. 

Context: Huawei has been working on the spin-off of its automotive business unit for several months. The company in November announced plans to establish a joint venture with Changan, stating that other existing partners such as Seres have been invited to invest in the new entity. 

• Voyah posted significant growth in sales last year with deliveries of 50,552 vehicles, compared to less than 20,000 units sold a year earlier. The automaker in August launched a revamped version of its Free mainstream luxury crossover with a starting price of RMB 266,900 ($36,532) featuring Baidu’s automated driving assistance system.

Xiaomi held its most significant media event of the year in Beijing on Thursday: the debut of its first electric car. With a size comparable to the BMW 5 Series and a shape similar to the Porsche Taycan, the four-door sedan boasts some of the Chinese car market’s highest specifications, as cut-throat competition from maturing rivals rises.

The sleek, gadget-full all-electric sedan is aiming to become a top choice for China’s increasingly tech-savvy consumers, and certainly aroused widespread curiosity judging by the more than 46 million people who logged on for the three-hour-long unveiling on the country’s Twitter-like site Weibo. Yet from journalists and insiders alike, the reaction was mixed. 

From the event, TechNode has selected some of the car’s highlights. 

Main specs

The high-performance SU7 can sprint from 0 to 100 km/h (62 mph) in 2.78 seconds, as it climbs to a top speed of 265 km/h. It is claimed to be the world’s most aerodynamic production car with a drag coefficient (Cd) of 0.195. By comparison, the Taycan Turo can hit 260 km/h and Tesla’s Model S has a Cd of 0.208. It also comes just a month after rival Huawei launched the Luxeed S7 sedan at 0.203Cd. 

Xiaomi said it uses two 9,100-ton mega casting press machines to produce the front and rear underbody pieces, giving the car a torsional stiffness of 51,000 Nm/degree, nearly twice the number of the Ford F-150 Raptor and higher than any other car on the road. The technology, first adopted by Tesla, has since been embraced by Chinese EV makers from Geely-affiliated Zeekr to Huawei-backed Aito.

Vehicle autonomy

Xiaomi’s chief executive Lei Jun presented aspects of the company’s self-driving initiative for public viewing, highlighting that the premium version of the SU7 will incorporate two Nvidia Drive Orin processing chips plus a laser sensor unit on the car’s roof to carry out certain partially autonomous driving functions. Xiaomi also showed a short video of the car drawing into a tight garage space autonomously.

The Chinese tech company has set a goal for its advanced driver assistance software to be available to drivers in 100 major Chinese cities by the end of the next year, according to Lei. Huawei and Xpeng Motors are for now the leaders of this booming market, with established carmakers from BYD to Great Wall Motor trying to catch up.

Smart cabin

The SU7 will be the latest Chinese car model powered by Qualcomm’s smart cockpit computing platform SA8295, after the Zeekr 001 FR and its sibling Jiyue 01, and its infotainment system will turn on in just 1.5 seconds. It is also integrated seamlessly into the Xiaomi ecosystem with the adoption of the company’s self-developed operating system, the HyperOS, which takes only 30 minutes or so to carry out important updates, according to the company. 

CEO Lei said the SU7 would create the same smooth experience that anybody with a Mi Phone is used to, as various apps are pushed from their phones to a 16.1-inch in-car dashboard once they sit in the car. Other devices, from tablets to home appliances, also seamlessly work with the vehicle, an integration trend led by auto and tech majors such as Huawei, Geely, and NIO.

Conclusion

Xiaomi will have to pick an appropriate price tag, given it starts with a somewhat broad, unclear positioning, said You Xi, a seasoned economic and financial writer and co-founder of Chinese online media platform Communication Planet. “It remains challenging for the company to extend its brand into EVs,” You added, citing similar offerings from multiple competitors among his reasons (our translation).

The smartphone giant plans to introduce two variants of the SU7 to “contemporary elites with taste in lifestyle and technology” in China over the next few months, said Lei. Some experts have predicted the premium version of the car, with an estimated driving range of 800 kilometers (497 miles), could cost consumers at least RMB 300,000 ($41,124).

Chinese EV maker NIO on Dec. 23 unveiled a long-wheelbase executive sedan model, the ET9,  with a price range of $112,160. The model boasts its own self-driving chip, marking the first utilization of the five nanometer process technology in China’s auto industry. 

The four-door executive flagship, equipped with proprietary technologies such as a sophisticated yet lightweight chassis system and a superfast-charging battery pack, reflects NIO’s commitment to redefining the upper premium vehicle market, William Li, the company’s founder, chairman, and chief executive, told press at the annual NIO Day event on Dec. 23. Li further referred to the target segment as ”a spiritual home base“ for international luxury carmakers (our translation). 

With a pre-sale starting price of roughly RMB 800,000 ($112,160), NIO’s answer to the Porsche Panamera could serve as a low-volume halo car and is scheduled for delivery in the first quarter of 2025. Larger rivals from BYD to Geely have also launched similarly-priced offerings, indicating their aspirations to upscale and grab a slice of the luxury market.

Here are some of the key specifications of the ET9 presented by NIO at the company’s annual gathering held in the northwestern Chinese city of Xi’an. 

Design highlights: Different from old-money cars that Western brands typically offer, the NIO ET9 features a sleek and contemporary look with high ground clearance, large 23-inch wheels, and cutting-edge gadgets such as laser sensors on the roof and sides for an all-round view of the car’s surroundings. 

• The interior boasts a modern design language. The car features an almost two-meter (six-feet) console that extends through the length of the cabin and is integrated with equipment such as a 10L fridge and folding tables, offering passengers first-class-style seats and comfort.
• The grand tourer measures 5.3 meters in length and 1.6 meters in height with a wheelbase of nearly 3.3 meters, making it longer than the Bentley Flying Spur and almost as tall as a regular off-roader. The back seats can be reclined as much as 45° and come with full-body massage programs. 

Autonomous driving: The ET9 will be powered by NIO’s first self-developed system on chip (SoC), the Shenji NX9031, for partially automated driving. NIO stated it will be the first Chinese automaker to use chips with five-nanometer process technology, providing its vehicles a computing power comparable to the combined total of that created by four industry-leading processors. 

• By comparison, NIO’s existing lineups are equipped with four Nvidia DRIVE Orin chips that handle up to 1,016 TOPS on seven nanometers. The purpose-built chip is a strategic move as it could improve processing efficiency and machine learning algorithm coupling for autonomous driving, Jefferies analysts said on Monday. 
• Li declined to provide further details about the semiconductor. Tesla has reportedly turned to Taiwan’s TSMC to produce its next-generation full self-driving (FSD) computer on 4/5 nm processes. Homegrown rivals Li Auto and Xpeng Motors are also developing their own chips for vehicle intelligence. 

Large cylindrical battery: The ET9 will incorporate NIO’s in-house developed, 46105-type cylindrical lithium-ion battery cells. This implies a size of 46 millimeters in diameter and 105 mm in length with a cylindrical shape, a technology also embraced by Tesla in the hopes of increasing ranges and lowering costs. 

• NIO asserts that its new batteries have a cell-level energy density of 292 watt-hours per kilogram (Wh/kg), and a 120 kWh pack can provide a range of 255 kilometers (159 miles) after five minutes of charging at a 5C rate, facilitated by a 900-volt electrical system. By comparison, CATL’s latest Qilin battery would allow Li Auto’s Mega van to cover 500 km on a 12-minute charge at a 5C rate. 
• NIO did not reveal many further details, except for Li’s comments to investors during a Dec. 5 earnings call stating plans to outsource battery manufacturing for reduced investment and improved margins. The EV maker is reportedly in talks with Great Wall Motor-backed Svolt to set up a joint venture for making batteries in its car manufacturing base of Anhui. 

Smart chassis: NIO also launched an intelligent chassis suspension system which the company claimed would provide a refined driving experience featuring a steer-by-wire system, rear-wheel steering, and adjustable suspension altogether for the first time in a mass-produced consumer car.

•  NIO said the so-called SkyRide Intelligent Chassis System, in combination with hydraulic components, could be raised or lowered by 50 millimeters in one second to help passengers in and out of the car. In April, BYD introduced a similar offering that could automatically adjust to different road conditions and driving styles.
• In a 50-second video clip presented by NIO on Dec. 23, an ET9 prototype offered an apparently smooth ride when driving over multiple speed bumps and showcased a four-tier tower of Champagne glasses on the front body panel which remained in place despite the road surface. The flagship sedan is also highly maneuverable with a turning radius as low as 5.45 meters, a record also achieved by Xpeng with its upcoming X9 van.

Volkswagen’s software unit Cariad and Chinese auto tech startup Horizon Robotics expect to recruit 300 employees by the end of this month for a newly established joint venture called Carizon, in an effort to meet growing local demand for advanced driving technology. 

Why it matters: The hiring spree marks the German auto major’s latest effort to develop its own in-vehicle software following an announcement last year of a $2.3 billion investment deal for a 60% stake in the JV in partnership with Horizon. 

Details: The two companies have not officially provided details of the recruitment plan, but Horizon’s co-founder and technology chief Huang Chang, leading a team of more than 100 engineers, has reportedly joined the JV. 

• Su Jing, a former head of Huawei’s autonomous driving team who has been on board at Horizon since earlier this year, will lead the development of advanced driving systems based on Horizon’s next-generation computing solution Journey 6. 
• Beijing-headquartered Carizon was officially set up on Nov. 20 with a registered capital of RMB 6.8 ($940 million), with VW and Horizon taking 60% and 40% stakes in the entity respectively, according to Chinese corporate data site Tianyancha.
• The JV will focus on rolling out automated driving technology powered by Horizon’s Journey series processors and integrated into VW’s upcoming battery EVs in China, according to an announcement dated Dec. 8. 
• VW and Horizon are on track to deliver their first collaborative development effort as early as 2025, Cariad China’s chief executive Chang Qing told Chinese reporters last month. This will allow VW vehicles to function independently on Chinese highways and certain urban streets, a feature popular with Chinese consumers.

Context: VW has made a series of moves to step up the pace of its software development for the Chinese market, including a $700 million deal for a 5% stake in Chinese EV maker Xpeng Motors unveiled in July.

• The German carmaker is also forming a JV with China’s ThunderSoft to improve its infotainment systems and car cockpits, Reuters reported in April, while partnering with smartphone maker Vivo on a similar effort.
• VW sold roughly 2.3 million cars in China for the first nine months of 2023, recording a year-on-year decline of 3%, of which roughly 117,100 units were EVs, up 3.9% from a year ago. It delivered 341,100 EVs in Europe over the same period, an increase of 60.9% from last year.
• Horizon said in April it has shipped more than 3 million computing solutions for over 120 car models from prominent Chinese automakers ranging from BYD to Geely. Self-driving startups such as Pony.ai and Qcraft are developing assisted driving technology based on its processors.
• US chip powerhouse Nvidia is also ramping up its self-driving car efforts in China with plans to expand its workforce locally after recently hiring Wu Xinzhou, a former vice president at Xpeng Motors, TechNode has reported.
• Speaking to Chinese media outlet LatePost in July, Wu said that foreign companies would need to establish research and development teams of “at least several hundred people” locally to be able to compete with Chinese carmakers and self-driving car companies (our translation).

Huawei is spinning off its automotive business unit, enabling Changan Automobile and other manufacturing partners to invest, in a move aimed at turning the loss-making car division into a profitable operation amid fierce competition. 

Why it matters: The reorganization is a rare move for Huawei – a company under 100% ownership of founder Ren Zhengfei and its staff since 2003, according to its official website – as the Chinese telecommunication giant puts a date of 2025 on its target of profitability for its as-yet loss-making auto business. 

• Deemed by Citic Securities analysts as “a milestone” for the Chinese auto industry, the move is expected to help Huawei court new industry allies and gain investment to pursue intelligent vehicle technology. Established automakers such as SAIC have reportedly voiced concern about Huawei’s move into electric cars. 
• The equity structure of the new entity may be comparable to that of the United Automotive Electronic Systems, a joint venture formed by German auto supplier Bosch and several Chinese carmakers including SAIC, FAW, and Dongfeng in 1995. This would allow more automakers to benefit from collaboration, and not just Changan, analysts wrote in a Nov. 26 note. 

Details: The new joint venture will focus on areas already covered by Huawei’s Intelligent Automotive Solution (IAS) business unit, including the development of intelligent driving software, digital cockpit systems, and digital platforms, among others, according to a regulatory filing published by Shenzhen-listed Changan dated on Monday. 

• Huawei will take at least a 60% stake in the new entity but will no longer directly compete against the new company in principle. Changan and its relevant parties will acquire no more than a 40% stake in the JV. The two companies plan to discuss the details of the transaction and sign an agreement within six months, the filing said. 
• The establishment of the new entity will have no impact on the ongoing collaboration between Huawei and Chinese car manufacturer Seres, according to a Nov. 26 statement. Seres, which makes Aito-branded EVs for Huawei, added it has been asked to participate in the investment and is in discussions to jointly develop intelligent electric vehicle architecture.
• The entity will prioritize diversified ownership, said the filing. It is anticipated that various parties will engage deeply in the development of the open vehicle platform, said Richard Yu, CEO of Huawei’s consumer business group and chairman of the IAS BU, who compared the platform to “a train engine.” 

Context: Huawei, state-owned Changan and Chinese battery maker CATL announced a partnership to establish EV brand Avatr back in late 2020. The companies have sold roughly 20,000 units of the Avatr 11 battery electric crossover since delivery began last December, launching their second premium model with a starting price of RMB 300,800 ($41,240) earlier this month. 

• Huawei has also been selling EVs since mid-2021 with lesser-known Seres, formerly the Chongqing Sokon Industrial Group, followed by the launch of the first Aito-branded EV last September. Huawei said in early October that it had secured more than 50,000 non-refundable orders for the redesigned version of the M7, Aito’s second model, less than a month after its launch. That number was updated to more than 100,000 as of Monday.
• The Chinese tech giant is also partnering with domestic manufacturers Chery, BAIC, and JAC, showcasing the first model under the new Luxeed marque jointly set up with Jaguar Land Rover’s manufacturing partner Chery on Nov. 9. The Luxeed S7 sedan will be officially launched on Tuesday and the respective new models co-built with BAIC and JAC are set to hit the market in 2024. 
• Huawei reportedly invested $1 billion in its automotive business in 2021 and has since maintained its push into the Chinese intelligent EV market in an effort to diversify its revenue sources and offset the impact on its core businesses from US trade restrictions. The IAS BU recorded revenue of RMB 1 billion during the first half of this year, accounting for around 0.3% of its total revenue. 

READ MORE: Xpeng and Huawei-backed EV maker set new delivery records as demand grows for self-driving tech

A Tesla car owner who protested against the company during the Auto Shanghai show in early 2021 has been forced to apologize for damaging the US car company’s reputation by alleging that Tesla sold defective cars, Chinese media outlets reported on Wednesday. 

Why it matters: The verdict marks the latest victory for Tesla in China after it faced mounting numbers of car owner complaints over various quality issues including unintended acceleration and brake failure in the past two years. 

• Among infamous cases are the woman who climbed onto the top of one of Tesla’s electric vehicles at the Auto Shanghai show in April 2021, and an accident resulting in two deaths and three injuries in southern China last November.

Details: A local court on Nov. 9 ordered a woman surnamed Li from the northwestern city of Xi’an to apologize to Tesla and pay the company RMB 2,000 in damages in addition to bearing the cost of vehicle appraisal totaling RMB 20,000 ($2,800). The public apology should remain on social media platform Weibo for at least 15 days, the court ruled.

• The female owner alleged faulty brakes caused an incident in March 2021 when her China-made Model 3 crashed into a car coming the other way in the capital of Shaanxi province. Tesla has denied the claim. 
• Li took part in a protest at the Tesla booth during that year’s Auto Shanghai show when she wore a white T-shirt with the words “brake failure” above a Tesla logo. Another female car owner surnamed Zhang then clambered atop a Tesla to express her anger. 
• The Chinese court required a third-party appraiser to conduct systematic inspections of the vehicle after Tesla filed a lawsuit against Li for defamation. The result showed no proof that any mechanical failure had contributed to the accident. 
• A company representative confirmed the news to Caixin on Wednesday, saying the automaker was attempting to appeal for higher compensation. It is also suing Zhang for damaging its reputation without taking further steps, as she rejected a third-party investigation of her car. 

Context: Tesla in May launched a recall involving over 1.1 million EVs in China following an investigation by Chinese regulators that showed Tesla owners could hit the accelerator pedal rather than the brake by mistake when its regenerative braking system was switched on by default. Beijing said the recall was intended to reduce the chance of accidents. 

• The US carmaker shipped 771,171 vehicles from its Shanghai gigafactory during the first ten months of this year, of which 462,355 were for domestic sales, a growth of 37.9% from a year earlier, according to figures from the China Passenger Car Association. Retail sales of new energy vehicles in China, mostly battery EVs and plug-in hybrids, increased 34.2% year-on-year to more than 5.9 million units over the same period, the CPCA figures showed. 

As automakers continue their struggle amid an unrelenting price war in China, both established brands and startups are showcasing their latest products at the Auto Guangzhou 2023 show in a bid to take pole position ahead of what promises to be another year of tight competition. 

Traditionally one of the country’s largest car shows, this year’s Auto Guangzhou offers a glimpse of how intense competition in China has been, and how successful it has been at flushing out weaker foreign marques as domestic rivals fall over one another in a mad rush to crack the market. 

“Joint car manufacturers are faced with unprecedented challenges against the backdrop of the current situation,” said Wen Dali, a deputy general manager of GAC-Toyota, a joint venture between the Japanese automaker and its Chinese partner (our translation). More than 20% of the JV’s new car sales over the next three years in China are set to be new energy vehicles, mostly battery-run electric vehicles (BEVs) and plug-in hybrid EVs (PHEVs), Wen added at a press event on Friday. 

Here’s a quick roundup of some of the highlights from the Guangzhou International Automobile Exhibition, which kicked off on Friday in the capital of China’s Guangdong province. 

BYD – Sea Lion 07

The BYD Ocean family of electric cars on Friday welcomed a new sibling and its latest answer to the Tesla Model Y, the Sea Lion 07, crafted by Wolfgang Josef Egger, BYD’s design chief and a former head designer at Audi Group. 

The mid-size crossover boasts distinctive design elements with its muscular fenders, bold air inlets, and clean character lines on all four corners, while the high shoulder lines and the dual, through-type waistlines give the vehicle a sporty vibe. The features are intended to make the car look unique from miles away, Fan Jihan, a deputy director of BYD said on Friday during the show.

Slightly larger than Tesla’s Model Y at 4.8 meters in length and with a 2,900-millimeter-long wheelbase, the top-end all-electric car is expected to have a driving range of more than 700 kilometers (435 miles), compared with the 688 km claimed by the long-range version of its US rival. Scheduled for official launch later this year, it will be equipped with BYD’s latest advanced driver assistance system (ADAS), according to the company.

Geely – Zeekr 007 

This year’s Auto Guangzhou saw the debut of the long-awaited Zeekr 007, the first electric sedan under the premium marque of auto major Geely. 

The latest model from Stefan Sielaff, formerly a head of design at Bentley, the 4.9-meter-long all-electric vehicle comes with 1,711 high-intensity lamp beads powered by 75 automotive chips. This enables the car’s LED headlights to display a dazzling, customized lighting sequence with animation about 90 inches wide, showcasing some of the most advanced lighting technology by a Chinese carmaker. 

Meanwhile, the Zeekr 007 features an 800-volt battery system, which offers a driving range of up to 870 km on a full charge and can travel another 610 km on 15 minutes’ extra charge. Zeeker claims it to be the quickest accelerating road car of the same class ever made, going from 0 to 100 km/h (62 mph) in 2.84 seconds, while also being one of the earliest models to use Qualcomm’s latest 5-nanometer cockpit chip 8295. 

The company aims to begin delivering the car in January at a lower-than-expected pre-sale starting price of RMB 224,900 ($31,059). 

Xpeng Motors – X9

Xpeng on Friday was on its home court when it unveiled details of its first flagship multi-purpose vehicle (MPV) the X9, which the Guangzhou-headquartered electric vehicle maker expects will stand out from existing offerings with superior comfort and top-notch performance. 

With a competitive pre-sale starting price of RMB 388,000 ($53,544), the seven-seater has a claimed interior space of 7.7 square meters, which makes it 12% bigger than the Toyota Alphard, a worldwide top-seller in the chauffeur-driven luxury people mover category, according to chief executive He Xiaopeng. 

The family van is also said to have the best third-row seats on the market that can be adjusted for recline to a desired angle of nearly 180 degrees and folded down flat to increase cargo capacity. Meanwhile, the luggage compartment offers space for seven suitcases. 

The Xpeng X9 is claimed to be the world’s first MPV equipped with rear-wheel steering as a standard configuration, which reduces the car’s turning diameter to an industry record of 10.8 meters (35.4 feet), making it easy to maneuver. 

Li Auto – Mega

Li Auto has finally made available the details of its long-anticipated MPV, the Mega, with an exterior echoing the bullet-style look of China’s high-speed trains. The seven-seater van boasts the world’s fastest charging speed among electric vehicles of all kinds, capable of traveling up to 500 km on 12 minutes of charge powered by CATL’s next-iteration Qilin batteries. 

It has a drag coefficient (Cd) of 0.215, which the company claimed is the lowest Cd rating for an MPV, while it will consume 15.9 kilowatts (kWh) of electricity for every 100 km of travel, also among the lowest in the industry. 

The company, which has delivered more than 500,000 plug-in hybrid SUVs as of September, confirmed plans to build 300 supercharging stations in China by year-end. Pre-sales of the Mega started on Friday with a price tag of around RMB 600,000 ($82,800) and delivery scheduled for February 2024.

READ MORE: Chinese carmakers showed up big time at Auto Shanghai 2023

Images of what could be Xiaomi’s first electric vehicle model have leaked online ahead of the car’s expected launch next year. The photos from the Chinese Ministry of Industry and Information Technology show a large sedan with styling similar to the Porsche Taycan, adorned with a Xiaomi logo. 

Why it matters: Automakers are required by Chinese regulators to apply for registration before officially selling vehicles in the country, and the government ministry’s post indicates that the debut of the first Xiaomi car is approaching. 

• Xiaomi has begun trial production of its first EV at its facility on the outskirts of Beijing, with the vehicle expected to hit the market as early as February, a person with knowledge of the matter told Chinese media outlet National Business Daily on Wednesday. 
• A Xiaomi representative declined to comment when contacted by TechNode on Thursday, but in late October, chief executive Lei Jun reaffirmed the company’s plan for the car to go on sale in the first half of 2024, according to an Oct. 25 post published on the Twitter-like platform Weibo. 

Details: The Xiaomi SU7 is around five meters long and spans a 3,000-millimeter-long wheelbase, making it bigger than many mid-size sedans such as Tesla’s Model 3. It has a total mass of 2,430 kg and a curb weight of 1,980 kg, based on the registration details revealed by the MIIT on Wednesday. 

• The car features a sleek, athletic low profile with Xiaomi’s logo on the front and its name on the rear hatch, similar to the Porsche Taycan, a likeness brought to light by a Chinese auto influencer. The images also show a couple of wheel options and a choice of yellow brake calipers.
• The SU7 will be able to reach a top speed of 210 kilometers per hour on a relatively affordable, iron-based lithium-ion battery from BYD. The top speed of the premium SU7 Max will be 265 km/h, with the higher-end model equipped with a more expensive, nickel and cobalt-based battery pack from CATL. 
• An electric motor will provide a power output of 275 kW and 220 kW respectively, while the top-end version will integrate laser sensor units on the roof to enable partially autonomous driving capabilities, according to images released by MIIT.
• The five-seater sedan will be manufactured at Xiaomi’s factory in the Beijing Economic and Technological Development Zone, which has an initial annual capacity of 150,000 units, although its production application was filed in the name of a subsidiary of state-owned automaker BAIC.
• This appears to confirm speculation that BAIC, a manufacturing partner of Mercedes-Benz in China, has joined hands with Xiaomi, meaning the smartphone maker is still waiting for final approval to begin manufacture from the Chinese authorities. 

Context: Xiaomi and Huawei are among the Chinese technology giants with the potential to become major players in the EV space with advanced intelligent capabilities and a broad sales network, which remain difficult for many carmakers to replicate, Morgan Stanley analyst Tim Hsiao commented on an earnings call held by Xpeng Motors on Wednesday. 

• Huawei said on Oct. 6 that it had secured over 50,000 non-refundable orders for the revamped M7 sports utility vehicle less than a month after its launch. The number was updated to more than 90,000 as of Wednesday, local media outlet IT Home reported. 
• The telecoms giant started pre-sales of the first electric sedan under the new Luxeed brand with automaker Chery on Nov. 9, followed the next day by the launch of the Avatr 12, a premium crossover co-developed with partners Changan Automobile and CATL. 

READ MORE: Five things to know about Xiaomi’s new electric car company

Huawei on Thursday revealed its first electric sedan under the new Luxeed marque in collaboration with automaker Chery, saying it will compete with Tesla and Mercedes Benz’s premium offerings at a price comparable to the cheapest models of its international rivals.

“After some deliberation, we will make all versions of the Luxeed S7 available for purchase despite making a loss,” Richard Yu, the chief executive of Huawei’s consumer business group, told the media during a press conference in Shenzhen (our translation). This will allow more customers to try Huawei’s smart vehicle technology at an affordable price, said Yu.

The aggressive pricing strategy unveiled at the Luxeed S7’s launch marks the latest push by the Chinese technology giant to crack the world’s biggest and most competitive electric vehicle market. Huawei hopes it will be a new revenue source to offset the negative impact of US restrictions on its smartphone business. 

Here’s what we know about the newly-launched Luxeed S7 sedan:

Pricing: The sedan comes at a minimum price of RMB 258,000 ($35,381), RMB 2,000 lower than Tesla’s entry-level Model 3 in China. Pre-sale started on Thursday and the official launch is scheduled for Nov. 28.

Automated driving: The Huawei-Chery electric sedan is the first model to use the tech giant’s latest proprietary Harmony operating system. Its autonomous valet parking feature enables the car to park itself in lots and then return to a designated spot using a remote-control assisted function.

The premium versions of the Luxeed S7 will include Huawei’s laser sensor units and its Advanced Driving System (ADS) that uses deep learning networks and computer vision algorithms, including one called the General Obstacle Detection network, for navigating its surroundings. 

Huawei has claimed its partially autonomous driving technology will be accessible on major city roads across China by the end of the year, potentially ahead of rivals including Xpeng Motors. 

Main specs: Yu specifically identified Tesla’s Model S as Huawei’s major competitor, claiming that Huawei and Chery’s full-size luxury sedan outperformed its rival’s in terms of range, energy efficiency, and luxury. 

The top-end Luxeed S7 will have a driving range of more than 800 kilometers (497 miles) and be capable of driving another 400 km on 15 minutes of supercharging using Huawei’s facilities. By comparison, the dual-motor Tesla Model S has a 715 km range and can add 347 km in 15 minutes. 

The car also impresses with high energy efficiency, consuming an estimated 12.4 kWh per 100 km, compared with 13.2 kWh and 17.5 kWh achieved by the rear-drive Model 3 and the dual-motor Model S respectively. “This is far ahead of our rivals,” said Yu, using a phrase that has become a Huawei-related buzzword on the Chinese internet. 

The S7 slightly beats out the Model S with a drag coefficient of 0.203. Meanwhile, it offers a 0 to 100 km/h (62mph) acceleration of 3.3 seconds, just under the 3.1 seconds reported by the Model S performance version but faster than the Porsche Taycan 4S, according to Yu. 

Interior: The sleek, aerodynamically favorable sedan boasts of a larger cabin space than its major luxury competitors with an interior length of 1,910 mm. The Mercedes E300L and the Tesla Model S measure 1,898mm and 1,816mm in interior length respectively, according to figures cited by Huawei during the press conference. 

The S7 also comes with a sporty design concept for the inside, featuring a wide dashboard, a 12.3-inch smart screen, as well as an oval-shaped steering wheel, allowing drivers to see the whole display, rather than having to view it through the steering wheel. 

In addition, it has adopted so-called zero gravity seat technology for the front passenger seat. This allows the human body to take on a neutral spinal posture, reducing the amount of stress placed on bones and joints, while the backs of the rear seats are heated, ventilated, and 27/32° adjustable.

READ MORE: Huawei-backed Aito now has 50,000 orders for its redesigned M7 model

Xpeng Motors said on Nov. 3 that it will offer some existing owners of its P5 sedan discounts on new purchases after hundreds of customers accused it of failing to deliver promised advanced driver assistance features, which were supposed to be available across the country. 

Why it matters: The complaints, which went viral on Chinese social media last week, mounted after Xpeng on Oct. 24 unveiled plans to roll out its latest advanced driver assistance system (ADAS), the XNGP, nationwide by next year. The company said it will be applicable to existing models including the G6, G9, and P7i, without mentioning the P5. 

Details: Xpeng said in a statement issued on Nov. 3 that it will offer an RMB 20,000 ($2,747) coupon for people who have subscribed to Xpilot, its previous generation driver-assist software, along with their purchases of the premium version of the P5 sedan. The benefit could be used for a new purchase of one of Xpeng’s most popular models, including the G6, G9, P7i, or its upcoming X9 van. 

• The announcement comes after more than 700 P5 owners recently published an open letter, obtained by National Business Daily, asking the company for an explanation as to why its partially autonomous feature for urban driving has remained unavailable to them in most domestic cities, despite the company’s promises.
• Xpeng further explained that the availability of its previous-generation ADAS feature “relies heavily on” high-definition maps, which has reportedly required automakers to secure approval for using mapping data in their vehicles, partly resulting in slow progress in adoption (our translation). 
• The EV startup released the so-called City Navigation Guided Pilot (NGP) function first to Xpilot users in the city of Guangzhou last September and has since expanded the adoption to five major cities including Beijing, Shanghai, and Shenzhen. 
• “We will strive for more new features and improved user experience, despite many challenges,” Xpeng said in the statement, adding that more functions will be available to P5 owners through over-the-air updates next year, including steering assist and more music streaming apps. 

Context: Xpeng began delivery of the P5 electric sedan back in October 2021, with its premium versions featuring two lidar sensors to facilitate more reliable automated driving functions at a price range of between RMB 199,900 and RMB 223,900 ($27,453-$30,749). It sold 19,618 units of the car over the last 12 months, according to figures from the auto services portal Dongchedi.

• The automaker is shifting to a more affordable approach for autonomous driving, which will reduce its reliance on technologies such as lidar and HD maps as part of a plan to roll out its XNGP system in 50 domestic cities by December. 
• The Xpilot system, formerly Xpeng’s rival to Tesla’s Autopilot system as reported by CNBC, is unavailable for driving scenarios without the support of HD maps, according to a Q&A document published by the company last November.

Chinese electric vehicle makers Xpeng Motors and Aito on Wednesday posted record-breaking figures for monthly deliveries, as the pace of adoption of self-driving technology accelerates among local customers despite slowing growth in China’s electric vehicle segment as a whole. 

Strong orders for Huawei, Xpeng, and DJI’s city NOA (Navigation on ADAS) products mark the start of the commercialization of smart driving, Jefferies analysts wrote in an Oct. 24 note. They added that Chinese automakers are becoming more willing to “test the waters” with chips by Huawei on some of their vehicles.

Why it matters: The latest figures highlight a brutal price war that has been continuing for months in the market, and the struggle automakers are facing in having to choose between lower prices or losing market share. 

• Xpeng in June unveiled the G6, its crossover answer to Tesla’s Model Y, at a price tag 20% lower than its US rival in China. Aito’s all-new M7 sports utility vehicle now costs RMB 40,000 ($5,468) less than the version launched a year ago.

Riding the self-driving boom: Xpeng Motors handed over 20,002 electric cars to customers in October, crossing the 20,000 unit milestone, nearly a threefold increase from a year ago and  31% growth from September. 

• Aito also reported a record delivery number of 12,700 units last month. The Huawei-backed brand does not report its delivery figures consistently, but its Shanghai-listed manufacturer Seres posted sales of 40,389 EVs for the first nine months of the year. 
• The two companies appear to have taken an early lead in an emerging battlefield for partially autonomous technology among consumer carmakers. More than half of the orders of Aito’s redesigned M7 SUV were placed for versions with Huawei’s Advanced Driving System, Chinese media outlet Caixin reported on Oct. 7, citing company insiders. 
• The Max versions of Xpeng’s G6 crossover, which features the company’s XNGP assisted driving technology, accounted for 70% of total orders in the first month after the launch, chief executive He Xiaopeng said in August. Both companies said their vehicles would be able to travel autonomously most of the time in dozens of major Chinese cities by the end of the year. 

EV startups: Li Auto also accomplished a delivery milestone last month, distributing 40,422 vehicles, making its year-to-date deliveries 284,647 units, the highest among the country’s nascent EV startups. The company has upped its goal to 50,000 units for the remaining two months of the year, CEO Li Xiang said on Wednesday on the Chinese Twitter-like platform Weibo.

• NIO’s October delivery of 16,074 units represented a 59.8% growth from this time last year and a slight 2.8% increase month over month. The company has delivered 126,067 vehicles as of October this year, still far from the annual goal of 245,000 units revealed by CEO William Li in March. It is now aiming for monthly delivery of more than 20,000 units in the fourth quarter of 2023. 
• Leapmotor’s delivery of 18,202 EVs last month comes after the Zhejiang-based EV maker recently announced a deal with European major Stellantis for a $1.6 billion war chest and turned its negative gross margin into a positive for the past quarter. Rival Hozon delivered 12,085 units, representing a decrease of 32.9% year-on-year and 8.5% month-on-month. 

Established majors: BYD’s growth momentum continued to some extent in October as the company saw sales surpassing 301,000 vehicles with a mild 5.2% rise from a month earlier. Analysts expect China’s biggest EV maker to achieve its annual goal of selling 3 million cars this year, as the company on Monday launched a wagon version of its popular Song SUV and readied to sell its long-anticipated Bao 5 off-roader.

• Sales for Aion declined 19.6% from a month earlier to 41,503 units, as the GAC subsidiary ramps up production of its new models, company insiders told financial media outlet CLS. Changan-affiliated Deepal delivered 15,513 vehicles in October, a 10.7% decrease from September. 
• Zeekr delivered 13,077 vehicles last month, up 29.2% from a year ago and 8.5% from September. On Aug. 11, the two-year-old premium EV brand, set up by Volvo parent Geely, cut the price of its 001 hatchback by up to RMB 37,000 to RMB 269,000 for a limited period until the end of this year. 
• Voyah saw its deliveries grow 21% on a monthly basis in October after the Dongfeng-backed EV maker launched its redesigned Free SUV in August, with the model arriving 15% cheaper than the previous version and equipped with Baidu’s advanced driver-assist system. 

Context: Retail sales of new energy passenger vehicles, including all-electrics and plug-in hybrids, are expected to reach 750,000 units in October, up 34.6% year-on-year and 0.9% month-on-month, according to estimates from the China Passenger Car Association. The past two months, known as “Golden September, Silver October,” are traditionally peak seasons for auto sales in China.

Chinese automaker Geely on Oct. 27 unveiled its biggest bet ever on intelligent vehicles with the launch of the first Jiyue-branded model, which the company says is capable of driving itself on busy urban streets in partnership with search engine Baidu. 

The automaker stated its vehicle relies heavily on a camera-based approach to capture detailed visual information and then respond appropriately, removing expensive laser sensors from its hardware suite to keep costs down. Tesla is reportedly a rare advocate for using the so-called vision-only approach, while most other brands opt for multiple sensors to mitigate safety concerns of their self-driving technologies. 

“I believe we provide users a better self-driving experience [than existing players] in most major Chinese cities,” Luo Gang, Jiyue’s chief operating officer, told reporters during an interview, adding that the Jiyue 01 outperforms Tesla’s offerings in digital services such as its AI assistant (our translation). Tesla’s full self-driving (FSD) function is currently unavailable in China. 

The Jiyue 01, a battery sports utility vehicle, comes in two versions with a price range between RMB 249,900 and RMB 339,900 ($34,148-$46,446), slightly lower than its pre-sale price and differing based on acceleration, driving range, and number of electric motors, among other specifications. Customers are also encouraged to pay RMB 19,900, a 60% cut from its sticker price, for all the premium functions of its self-driving software. 

Here are some of the news and highlights from the launch event held in Shanghai by Jiyue, formerly known as Jidu before Geely and Baidu set up a new venture in August. 

Self-driving tech: Jiyue said its advanced driver-assistance system, the Robo Drive Max, is already available to drivers in Shanghai, Hangzhou, and Shenzhen, meaning the cars can navigate complex urban streets in the three big cities with autonomous features such as overtaking, lane changing, and on-ramp/off-ramp driving. The firm is targeting nationwide availability for the software by 2024, which would mean it matched rival Xpeng. 

• Chief executive Joe Xia claimed the car could drive itself from point to point without many user interventions by using less costly high-definition maps and training multiple neural networks such as occupancy networks in big data sets, rather than relying on lidar. Rival Xpeng is also removing two radar sensors for its upcoming MPV model but retaining lidar technology for enhanced safety, TechNode has reported.
• The five-seater Jiyue 01 is equipped with 11 cameras and 17 ultrasonic sensors and radars. The company believes it is building public confidence in autonomous car safety, as Baidu has tested its autonomous car fleets without accidents for more than 70 million kilometers (43.5 million miles). Baidu has been handling various corner cases over the past decade, which greatly improves the safety of the system, said Luo. 

Smart cabin: The Jiyue 01 also boasts the most advanced voice recognition software on the market for in-car services, which can respond intelligently in milliseconds without losing its connection, as the company deploys artificial intelligence models and moves data analytics from cloud computers to the vehicle. The system is also set to evolve and become more alert to the needs of its owners, powered by Baidu’s ChatGPT-like chatbot, Ernie Bot. 

• Notably, the automaker is bringing voice activation outside the car, saying it is the world’s first model that allows autonomous valet parking via just a spoken command without the driver sitting in the car, from as far as two kilometers away, according to an announcement. A company employee demonstrated the feature with several reporters joined by TechNode in an indoor parking lot on the sidelines of the event. 
• Xia added that the vehicle’s in-car system is powered by Qualcomm’s most advanced smart cockpit computing platform, the SA8295, which provides a processing power of over 60 trillion operations per second (TOPS), compatible with that of flagship smartphones available on the market. This would allow users to play the hit racing game Asphalt with a 35.6-inch display across the dashboard, as would NIO owners do with their handsets and a smaller screen. 

READ MORE: Baidu’s EV firm Jidu aims to take on Tesla

Xpeng Motors teased how it sees the future of electric vehicles on Tuesday with the debut of its first multi-purpose vehicle model and a new timeline for the expansion of its self-driving software, as it faces an unprecedented offensive from major rivals like Huawei in a hotly competitive battleground.

Chief executive He Xiaopeng also revealed that the company has made significant progress in bringing flying cars closer to reality, while showcasing a working prototype of its humanoid robot, in a move reminiscent of Tesla’s introduction of its Optimus bot last September. 

Here are the key highlights from Xpeng’s annual 1024 Tech Day event. 

First MPV 

Navigating within a sharp and narrow turn at low speed on the stage at Tuesday’s event, Xpeng’s X9 is claimed to be the world’s first multi-purpose vehicle model equipped with rear-wheel steering as a standard configuration. This would allow the seven-seater, three-row van to handle “just like” a regular-sized sports utility vehicle, said He (our translation). 

Xpeng’s next-generation smart cabin system, the XOS, will also be available first to the owners of the X9, which is set to be formally launched at the upcoming Guangzhou Motor Show on Nov. 17. Powered by Qualcomm’s five-nanometer 8295 processor, the in-car software will offer a split screen mode, allowing drivers and passengers to run different applications simultaneously side-by-side for efficient multitasking.

Marking Xpeng’s entry into the Chinese MPV segment, the all-electric X9 will have to compete with an increasing number of similar offerings by established makers including BYD’s Denza brand, Great Wall Motor, and Dongfeng’s Voyah marque. Huawei-backed Aito and Li Auto are also set to launch their first MPVs later this year, targeting China’s growing three-generation families with larger interior car spaces.

Self-driving availability

Xpeng has also begun its switch to a more affordable hardware suite by removing some sensors from its incoming X9 model, betting more on cameras and artificial intelligence for its XNGP advanced driver assistance system, according to He. 

The Chinese automaker has updated its self-driving technology with what it described as some of the most advanced occupancy networks in the industry, comprising a deep neural network that reconstructs barriers and vehicles and predicts occupancy in a three-dimensional space for collision avoidance. 

A similar move has allowed Tesla to remove several ultrasonic sensors from its vehicles while enabling high-definition spatial positioning, longer range visibility, and the ability to differentiate between objects with its Full Self-Driving Beta software, which was announced by the US automaker last October. 

CEO He said Xpeng will deploy its XNGP system for urban traffic roads in 50 cities by December and make the functions available to drivers across China and Europe by 2024. It is competing with Huawei, which has quickly emerged as a rising player in the industry and previously announced a nationwide roll-out of similar features by year-end, while rivals BYD and Li Auto are playing catch-up.

Flying cars and robots

Experimenting with different approaches around flying cars, Xpeng also showcased two prototype aircrafts, or electric vertical takeoff and landing vehicles (eVTOLs). One of them boasts a two-in-one design that can fold up its wings and other components into the vehicle body, although He acknowledged that there are still some safety issues to be addressed. 

The 46-year-old serial entrepreneur sees greater potential for the commercial adoption of the other prototype, which is built on a modular system allowing the separation of the flight and automobile components. This model has a spacious interior with five seats while on the road and is powered by an extended-range hybrid engine, which can also recharge its aircraft component as it drives; up in the sky, the model is capable of carrying two passengers in an all-electric mode.

Xpeng further surprised the audience on Tuesday as its humanoid robotic prototype, the PX5, made its first public appearance. The company showcased the robot’s ability to navigate different terrain and pick up hand-held objects such as pens in a video. He envisions a near future where such AI machines could help look around in its factories or even mingle with customers at showrooms, hopefully by this time next year, he added.

Chinese carmaker Guangzhou Automobile Group (GAC) is strengthening its alliance with ride-hailing platform Didi, investing up to $75 million into the latter’s autonomous driving unit. The move is expected to help GAC enhance its self-driving technological capabilities and sustain its sound growth momentum in the Chinese electric vehicle segment, according to an industry veteran. 

The deal, nearly clinched over three years ago, has recently been revived by the two companies as the impact from Beijing’s extended crackdown on Didi has waned, a person with direct knowledge of the matter told TechNode on Friday. It also comes against the backdrop of Didi’s renewed efforts to solidify its position as China’s biggest ride-hailing service with new incentives, putting smaller rivals under pressure. 

Self-driving push: Autonomous driving has proven to be among the most capital-intensive startup businesses on the current tech landscape, and the extended collaboration with Didi would allow GAC to share its costs and risks of making robocars, said Liu Guanghao, partner at Shanghai-based venture capital firm Befor Capital.

• The first robotaxi jointly developed by GAC and Didi is slated to join Didi’s ride-hailing network for commercial operation in 2025, the companies said earlier this year. GAC’s EV arm, Aion, announced a partnership with Didi back in May 2021 to develop a mass-produced car with Level 4 autonomous capabilities, indicating that the car can pilot itself without a human driver most of the time.
• Didi will also jointly test and operate autonomous vehicles for ride-hailing with OnTime, a mobility platform launched by GAC with partners in 2019, as disclosed by an anonymous source. OnTime, primarily active in the southern Chinese province of Guangdong where its parent company is headquartered, has been testing AVs with Toyota-backed Pony.ai, as well as Nissan-supported WeRide.

EV sales boost: The investment would also help GAC’s core carmaking business achieve sustained growth, especially in the Chinese commercial fleet segment, where its EV brand Aion has established a significant presence over the years, according to Liu. “Carmakers need more sales in order to survive in this highly competitive market,” he said. 

• Aion ranked second in sales among all-electric vehicles for ride-hailing, with approximately 49,000 units sold from January to October 2022, which accounted for 29% of its total sales, according to figures compiled by Shanghai-based consultancy LandRoads (in Chinese). BYD was the top-selling brand in the field, with sales of 35,000 more units during the same period, although this accounted for only 14% of its total volume. 
• GAC told investors last March that shipments of its Aion EVs for ride-hailing services only accounted for 12% of its total sales. The automaker, also a manufacturing partner for Toyota and Honda in China, reported sales of nearly 360,000 Aion EVs from January to September and is hoping to achieve 500,000 units for this year, which could almost double the number it sold in 2022. 

Context: GAC Capital, a wholly-owned subsidiary of the automaker, as well as state-owned Guangzhou Development District Investment Group, will invest the same amount of up to $149 million totally in Didi’s self-driving unit. GAC is set to inject no more than $75 million in the funding round, according to a Friday announcement (in Chinese). 

• OnTime is currently pursuing a public listing on the Hong Kong stock exchange, four years after it was launched by GAC along with a group of investors including Didi and Chinese gaming giant Tencent in mid-2019. It completed roughly 60 million rides last year and has operated in 21 domestic cities in the country’s Greater Bay area as of June. 
• GAC declined to comment when contacted by TechNode on Friday. Didi did not respond to TechNode’s request for comment. 

Aito, a Chinese electric vehicle brand backed by Huawei, has received more than 50,000 non-refundable orders for its redesigned M7 in less than a month. The orders follow the Sept. 12 public launch of the sports utility vehicle, which features Huawei’s Harmony operating system and assisted driving technologies. 

Why it matters: The latest sales figures, as revealed by a senior executive at Huawei, show tentative signs of a bounce-back for Aito from a months-long slump and could be a boost to the confidence of Huawei’s car manufacturing partners. 

• Aito has also been buoyed by Huawei’s comeback in the smartphone market with the recent launch of its Mate 60 Pro series, following the US ban on exports of advanced semiconductor technology to the Chinese technology giant. 

Details: The revamped M7 crossover has racked up more than 50,000 pre-orders with non-refundable deposits of RMB 5,000 ($685) as of Friday, Richard Yu, the chief executive of Huawei’s consumer business group, said in a post on Chinese social media app WeChat. 

• Yu described the growth momentum of the new M7 as “a miracle,” adding that more than 10,000 customers placed their orders over the past two days. He called on sales employees to ramp up delivery to meet the growing demand (our translation). 
• Accumulative orders per store averaged more than 80 following the launch on Sept. 12, according to figures posted Saturday by Sun Shaojun, founder of consumer behavior research agency CarFans. Aito said in June it operated a network of around 1,000 retail locations and service centers in 230 Chinese cities. 
• Sun added that a surge in store traffic for Huawei’s new smartphones has boosted the sales of the Aito-branded EVs, produced by Chinese manufacturer Seres, over the recent National Day holiday season. Huawei began selling EVs with its little-known partner via its retail network in 2021.
• Roughly 40-50% of the M7’s buyers are Huawei smartphone users and were coming to the stores for the Mate 60 handsets, Jefferies analysts wrote in an Oct. 5 note, citing an executive of a Chinese auto dealership. Customers compare the six-seater with Li Auto’s L7, BYD’s Tang, and the Ford Edge, analysts said.

Context: Huawei on Sept. 12 unveiled the redesigned version of the M7 SUV, featuring Huawei’s Harmony operating system at a starting price of RMB 249,800 ($34,299), which is around RMB 70,000 lower than the initial version launched a year earlier.

• The vehicle also comes with Huawei’s latest assisted driving software, ADS 2.0, which will allow it to travel by itself on busy urban streets nationwide as early as December, making it one of the most ambitious players in the Chinese self-driving car space.
• Huawei has offered future owners of the new large-sized plug-in hybrid early access to purchase its Mate 60 smartphones. The Mate 60 Pro flagship handset reportedly incorporates a self-developed 5G processor, a breakthrough for the Chinese tech giant following US sanctions in 2019. 
• Two-year-old Aito has seen sales slump during most of 2023 amid fierce competition from more established rivals such as BYD and Tesla. Seres, which produces Aito-branded EVs, recorded sales of around 33,000 units for the first eight months of this year, representing a 15.6% decline year-on-year. 
• Meanwhile, Huawei has partnered with several other domestic automakers including Changan and BAIC. It is also on track to launch the S7 with carmaker Chery in November, the first sedan under a new marque called Zhijie in Chinese that will compete against Tesla’s Model S.

Chinese premium electric vehicle brand Denza on Tuesday revealed a cheaper version of its advanced driver assistance system (ADAS) in collaboration with US chipmaker Nvidia, as the BYD affiliate ramps up efforts to compete against leading self-driving players such as Xpeng Motors and Huawei. 

Denza is also eyeing overseas expansion, having established its presence in the China market with year-to-date deliveries of nearly 80,000 EVs as of August. The company expects overseas sales to begin as early as next year, including in Australia, Southeast Asia, the Middle East, and Europe. 

Why it matters: The companies said the launch of the affordable assisted driving technology could reduce the barrier to a transition to intelligent mobility. The system facilitates Denza’s vehicles to navigate most highways in China as well as some busy urban streets in major domestic cities. 

• BYD launched the N7 crossover under the Denza marque in July, with the top-end version powered by Nvidia’s ​​DRIVE Orin processor, which offers 254 trillion operations per second (or TOPS). Now all N7 models can be equipped with Nvidia’s DRIVE Orin chips for automated driving, according to a Wednesday statement. 

Details: The new autonomous driving system will enable on-ramp to off-ramp driving, as well as automatic lane changing on Chinese highways, for Denza’s flagship N7 SUV. It has a price tag of RMB 15,000 ($2,053) and is powered by Nvidia’s DRIVE Orin processor, which can handle up to 84 TOPS. The N7 SUVs that feature the technology will have two lidar sensors removed to reduce costs. 

• The companies say that the higher-end version, priced at RMB 23,000, will allow the vehicles to function by themselves on bustling city streets for the daily commute, using a feature named City NOA (Navigate On Autopilot). Denza’s general manager Zhao Changjiang said the company would release its Highway NOA feature to N7 owners starting in December, followed by an over-the-air update of the City NOA early next year. 
• Tong Liu, vice president and general manager of China auto business at Nvidia, said that he was “impressed” by the efforts made by BYD in developing intelligent cars over the course of their three-year collaboration, calling BYD/Denza a “strong advocate” of commercializing self-driving technology (our translation). BYD’s Dynasty and Ocean lineups are also using Nvidia’s semiconductor. 

Context: Several Chinese auto and tech companies have announced ambitious plans for the adoption of assisted driving technologies for urban driving, akin to Tesla’s full self-driving (FSD) function that has yet to be made available in the country. 

• Volkswagen-backed Xpeng Motors in June launched its City Navigation Guided Pilot feature in Beijing and is on track to expand the capability in at least 50 domestic cities by the end of this year, while Great Wall Motor has set a target of covering 100 cities by 2024.
• In the meantime, Li Auto vehicles will be able to navigate on fixed routes for daily commuters in 100 major Chinese cities by year-end, following weeks of training with its collection of datasets. Rivals Nio and Geely’s Zeekr are also planning to roll out similar features later this year. 
• Huawei is by far the most ambitious company in the field in China, with its head of consumer business Richard Yu stating on Sept. 12 that Huawei’s self-driving system would be applicable nationwide for both highway and urban driving with Aito-branded EVs by December, Caixin reported. 
• BYD has made a series of moves in recent months to enhance its research and development capacity, especially for autonomous driving, including organizational restructuring and talent hiring. More than 80% of the 30,000 fresh graduates recruited by the company this year were research personnel. 

READ MORE: Baidu and Huawei take on global giants with new in-car software offerings at Auto Shanghai 2023

China’s Great Wall Motor (GWM) will bring its next-generation in-car operating system to market next year, and stick to the ambitious goal of rolling out its semi-autonomous driving function nationwide by the end of 2024, according to a press event held on Tuesday. 

The company is undertaking a targeted push to create a scalable and unified software platform for future vehicle models across multiple different brands, a concept that has become mainstream in the years since Tesla entered the market. A significant increase in the number of software updates, aimed at improving the driving experience, is expected from next year, vice president Nicole Wu told TechNode at the event, held in the northern city of Baoding, where the company is headquartered. 

China’s third biggest private automaker by sales volume, GWM had a relatively early start in autonomous driving and in-car technologies. It began testing self-driving cars with the creation of a dedicated division called Haomo.ai in 2019 and became the second Chinese automaker after Xpeng Motors to build a supercomputing center, this January. Now, the company has set up a new artificial intelligence research lab to bring generative AI tools into play in future car models. 

Here are some of the highlights of TechNode’s interview with GWM executives, including vice president Nicole Wu, senior director Jiang Haipeng, director She Shidong, and Yang Jifeng, head of the AI lab. 

Major digital cockpit progress

GWM will roll out an app store and implement it across all brands, as part of its upcoming in-car operating system, Coffee OS 3.0, scheduled for release in the first half of 2024. The store will give users access to common third-party services and infotainment apps fine-tuned for car-friendly usage, as more customers expect a smartphone-like experience in the car. 

By working with smartphone makers such as Huawei and Xiaomi, the new system will allow drivers to use a handset while operating their vehicle. She Shidong added that owners will be able to play video games and watch movies in their cars by connecting gaming consoles, augmented-reality glasses or other devices, with the car dashboard using wireless or bluetooth connections.

By making constant updates of driving and infotainment features possible, the Coffee OS 3.0 is intended to take the in-car experience to a new level. Wu envisions each new GWM model getting a major software update every two to three months. Tesla and Nio released 2.8 and 1.3 software updates per month on average respectively in China during the first half of 2022, according to figures from domestic consultancy Ways. 

Ambitious self-driving goal

GWM has maintained its goal of launching Navigate on HPilot (NOH), a function similar to Tesla’s full self-driving (FSD) technology, to drivers in 100 cities around China by 2024. The software will first be available to owners of its Blue Mountain flagship SUVs in Beijing and Shanghai by next March, according to Jiang. 

This will enable vehicles to change lanes, overtake, and make turns automatically on Chinese city streets without high-precision maps. Jiang added that a set of common middleware plays an important part in creating a platform for assisted driving software that is updateable and scalable at a reasonable cost. 

Chinese auto and tech companies have been competing for a leading position in this space at a time when Tesla’s FSD function has yet to become available in the country. Xpeng’s XNGP advanced driver assistance system is set to be available in 50 major cities by the end of this year, while Li Auto’s EVs will be capable of traveling on fixed routes by themselves after training for weeks in 100 cities. 

Bring generative AI to vehicles

GWM is also looking to greatly expand its in-car system capabilities through the integration of emerging technologies such as generative AI tools. Its first aim is to use AI to anticipate user preferences and create high-quality infotainment content in some new car models in the fourth quarter of this year.

The company’s newly established AI Lab has been exploring the use of large language models in GWM vehicles. Yang expects significant improvement with the upcoming Coffee OS 3.0, especially in voice recognition and natural language understanding, expecting that the latest operating system will be able to give detailed, relevant responses to users’ queries using AI.

Rival players are all developing ChatGPT-like virtual assistants for use in future car models. Geely is scheduled to launch its RMB 128,000 ($17,600) Galaxy L6 SUV on Saturday with a proprietary AI model that can read children’s picture books. Both GWM and Geely-affiliated Ecarx earlier partnered with Baidu to develop AI assistants based on the latter’s GPT-style large language models.

Chinese EV maker Zeekr made a splash on Sept. 1 when it launched its first high-performance, track-focused vehicle – one which it hopes will establish new benchmarks in the field and compete with established brands such as Porsche and Tesla.

The 001 FR, which Zeekr is calling the world’s best-performance electric vehicle, uses four silicon-carbine motors for sophisticated torque vectoring, producing a powerful 1,265 brake horsepower, compared with 887 hp of the Porsche 918 Spyder.

The high-performance brake, completely redesigned from the original 001, can, the company claims, accelerate from 0-100 km/h (0-62 mph) in 2.07 seconds, faster than the 2.1-second acceleration to 60 mph of the Tesla Model S Plaid. The new model promises to be an everyday supercar, with a rapid battery charge from 10% to 80% in 15 minutes.

The debut comes at a time when Chinese manufacturers are rushing to launch premium offerings with eye-catching performance specs in a quest to upscale and compete in the global luxury EV segment. 

Zeekr has not released pricing details for the 001 FR, but has said the car will be made available in limited supply of up to 99 units a month from October. This will bring it into competition with another high-end rival, as BYD begins deliveries of its RMB 1 million ($150,000) electric SUV later this month. 

“Global luxury brands have ruled the performance car segment throughout the era of internal combustion engines … but Chinese electric vehicles are now capable of competing head-to-head against European top-tier supercars,” Andy An, chairman of Geely Auto Group and CEO of Zeekr told reporters in an interview after the launch. 

TechNode also spoke to Chen Qi, vice president of Zeekr and a former Huawei executive, about the company’s approach to autonomous driving as it looks to expand overseas. Geely’s premium EV subsidiary is establishing its footprint in Europe as part of its goal to deliver 140,000 units this year while looking to sell shares publicly in the US. 

Below are highlights from a group interview after the launch, which have been translated, condensed, and edited for clarity:

On limited production of the 001 FR 

An: The Zeekr 001 FR comes with a comprehensive list of high-performance equipment among which are extremely rare parts mostly needed and reserved for professional race cars. 

For example, more than 70% of Brembo’s carbon-ceramic brake systems are provided to today’s top-tier race cars, with less than 20,000 units available for road cars annually. We are individually crafting the 001 FR to ensure the highest standards of quality are attained, which together with other factors restricts the sports car’s output capacity to less than 100 units a month. 

Our customers have reacted remarkably well: the first 99 units of the 001 FR were sold out in 15 seconds after reservations opened [on Sept. 1] and the number exceeded our annual production capacity 20 minutes after that. I think this is because the 001 FR represents the state of the art as a sports wagon, which could improve sales and help establish Zeekr’s image as a technology-driven company. 

On Zeekr’s self-driving roadmap 

Chen: Zeekr has pursued a dual strategy of initiating in-house development as well as outsourcing to catch up with rivals in self-driving technologies. We are pushing forward a new program to bring autonomous driving for urban scenarios with future models using Nvidia’s semiconductor chips.

Meanwhile, it requires a relatively long period of testing and validation for existing Zeekr models to navigate Chinese urban roads with Mobileye’s advanced driver-assist technology. Mobileye has been an early mover in creating its digital maps to enable self-driving cars and we will use its assisted driving systems mainly in the European market. 

Automakers are deploying assisted driving technology on a city-by-city basis because more effort is needed to enhance the neural network’s generalization ability in various practical driving scenarios. [Editor’s note: Transformer is a new deep neural network architecture first mentioned in a 2017 Google paper and later used by Tesla to convert location data gathered by cameras into three-dimensional space for motion planning and control. Many assisted driving software have since been written using the transformer algorithm.]

We are accelerating efforts to roll out driver assistance software, first applicable on major Chinese highways, and we will then let our cars navigate complex urban streets automatically. 

On Zeekr’s US listing plan  

An: Zeekr will venture into the capital markets. But it is not the top priority for our management at the moment. There is no update on Zeekr’s listing plan following approval from the Chinese regulator. We will keep an eye on investor sentiment before taking a chance to go public. 

Zeekr has set an annual delivery target of 650,000 units by 2025 as one of the top three luxury EV makers worldwide since its inception and remains confident under pressure. We’ve made significant progress in a comprehensive way, including building a substantial cost advantage over competitors other than Tesla, and will reach the goal with the launch of a new model later this year, followed by two all-new ones in 2024 and 2025. 

On global expansion 

An: Zeekr started exports to Europe with 500 Zeekr 001 cars last month and will begin vehicle delivery first in Sweden and the Netherlands as early as September and in several other European countries next year. We are also preparing to enter regional markets including Southeast Asia, the Middle East, and Latin America, but will keep our focus on Europe at the moment. 

We expect to see a significant contribution to sales from overseas markets in the future. Chinese electric vehicles are gaining momentum in the global auto industry and we will make use of this to go upscale and expand globally.

READ MORE: Experts bullish on Chinese automakers’ global push as SAIC seeks EU foothold

Xingji Meizu, a smartphone company controlled by Geely founder Eric Li, has decided to discontinue its chip development business for cost-saving reasons. The move is expected to result in layoffs of dozens of staff members, including some fresh graduates, local media outlet Meiren Auto reported on Tuesday.

Why it matters: Xingji Meizu is the latest company to abandon its pursuit of critical and emerging technologies in the Chinese auto and tech industries, reflecting the challenges of a faltering economy and intensifying competition. 

• The news comes just days after electric vehicle maker Nio delayed the development of its own batteries to ease cashflow constraints. Similarly, in May, rival phone maker Oppo announced the closure of its chip design unit Zeku after five years of operations.

Details: In a statement sent to financial media publication CLS on Tuesday, Xingji Meizu said the company is closing down its in-house chip design program in the face of global economic uncertainties, and will instead sharpen its focus on product innovation and user experience.

• Xingji added that it will offer compensation as required by law, along with internal job transfer opportunities, to ensure the rights and interests of employees, especially fresh graduates, without revealing further details. Geely did not respond to TechNode’s request for comment. 
• The company’s chipmaking institute employs approximately 200 people, and dozens of recent graduates are likely to be impacted by the layoffs, according to Meiren Auto. “[The news] came just three weeks into the job,” one of the new employees told the outlet. 
• Development has mostly stalled since the launch of the institute, according to a person with direct insight into the company’s operations. Chief executive Shen Ziyu told Chinese reporters in March that emerging technologies, including chipmaking, were at the center of Xingji’s strategic efforts, alongside smartphones and in-car systems.

Context: Geely’s other affiliates have reported progress in semiconductor technology. The most recent example is the Lynk & Co 08 SUV featuring an in-car operating system built upon a supercomputing platform provided by Ecarx, another auto tech firm founded by Shen Ziyu and Geely’s Eric Li. 

• Siengine, Ecarx’s joint venture with Arm China, was responsible for designing seven-nanometer chips intended for use in computers in partnership with leading global chipmaker TSMC, Shen told Reuters back in March 2021.
• Xingji Technology, a company established by Li, acquired nearly 80% shares in beleaguered smartphone maker Meizu last summer, which preceded the establishment of Xingji Meizu and the release of Meizu’s first high-end handset series in two years this March.
• Xingji Meizu is also leading the business development of Geely-owned Swedish automaker Polestar in the Chinese market, having set up a joint venture with the EV maker in June. 
• Geely founder Li first revealed his plans to enter the smartphone market back in 2021.

Chinese EV giant BYD is taking on a record 30,000 fresh graduates this year, with research personnel accounting for 80% of the total intake, in a move intended to shore up its research and development department, a company representative has confirmed. 

Why it matters: The hiring drive comes as BYD looks to retain its dominance in the Chinese electric vehicle market as rivals continue to offer a competitive challenge. The move contrasts sharply with general hiring trends as China faces soaring youth joblessness. 

• China’s unemployment rate for those aged 16 to 24 rose to a record 20.8% in May, according to the National Bureau of Statistics. One Peking University professor said she expected that number could rise to nearly 50%, according to a July 20 report by Reuters. 

Details: Around 31,800 fresh graduates have come on board at BYD since the start of 2023, more than 61% of whom have a master’s or doctorate degree, and over 80% of whom will work in R&D projects. State-owned newspaper People’s Daily was the first to report the story on July 29.

• The Chinese automaker has been hiring research employees in electronics and electricals, new energies, and semiconductors, to be mainly based in Shenzhen, Shanghai, and the northwestern city of Xi’an, according to a job post on its official website. 
• A BYD spokesperson confirmed the news when contacted by TechNode, without offering further details. 

Context: BYD has been expanding its R&D team for several years with the number of engineers hired by the company growing 31.5% year-on-year to around 40,400 in 2021. That number increased 72.6% year-on-year to nearly 70,000 as of last year. The company had around 570,000 employees in 2022, of which around 75% were production workers, financial media outlet Caixin reported.

• The Chinese EV giant, which had a relatively late start in the autonomous driving field, recently hired between 4,000 and 5,000 software engineers, Reuters reported on May 17, citing the company’s senior vice president Stella Li. It has also been running an intelligent driving research unit in Shanghai since last year, the Reuters article said, while reportedly restructuring its vehicle engineering institute. 
• The company is rushing to reach the top end of its full-year sales target of 3.6 million EV units, which would almost double last year’s total. It has sold nearly 1.3 million units this year, as of June. BYD spent RMB 20.2 billion ($2.83 billion) on R&D last year, up 90.3% from a year ago, and has developed key components in-house including EV batteries, electric powertrain systems, and vehicle control technologies. 

Denza, a luxury car subsidiary of Chinese electric vehicle maker BYD, released its first SUV model N7 on Monday, priced from RMB 301,800 ($41,705). The company said it has received more than 24,000 pre-orders since its public unveiling on April 18.

The N7 is also the first model equipped with BYD’s assisted driving technology and will be capable of navigating on complex urban roads in China early next year, general manager Zhao Chaojiang said during the press conference.

Why it matters: BYD’s latest launch shows its intention to elevate the brand and secure a foothold in the premium market. The budget-friendly automaker is hoping its sub-brand Denza will become a luxury marque, and the launch of the N7 is a crucial step towards achieving this goal.

• The N7 will also be seen as a test of the company’s aspirations and its ability to beat rivals like Tesla, Huawei, and Xpeng when it comes to autonomous driving features.

Intelligent driving: The top-end version of the N7 features a hardware suite of 33 high-precision sensors, including two 8-megapixel cameras and two lidar sensors, and is powered by Nvidia’s Drive Orin processor which offers 254 trillion operations per second (or TOPS). By comparison, Xpeng’s G6 features 31 sensors and Nvidia’s dual Orin chips.

• Denza also revealed that its advanced driver assistance system (ADAS) will cost RMB 23,000. It will allow cars to change lanes, speed up, and slow down on Chinese highways when it is updated in the last three months of this year and on city streets by next March.
• By comparison, Huawei-backed Aito and Avatr last week cut the price of their similar offerings in half to RMB 18,000. Both will roll out their assisted driving tech for urban scenarios in 45 cities by year-end, according to Richard Yu, head of Huawei’s consumer business group.

Other details: The N7 has a driving range of 702 kilometers (436 miles) and can be refueled with an additional 350 km of range in 15 minutes by BYD’s proprietary dual charging technology. For comparison, Xpeng’s G6 can travel 300 km on a 10-minute charge.

• The five-seater battery electric crossover is also among several new BYD models to adopt the company’s body control suspension system DiSus for a smooth ride on bumpy roads, with Zhao on Monday claiming the function can eliminate car sickness.
• Zhao also told Chinese reporters that around a third of the N7 reservations were from existing owners of German brands such as BMW, Mercedes, and Audi. Delivery of the vehicle is scheduled to begin later this month and the company expects monthly deliveries to reach 10,000 units as early as October.

Context: BYD and partner Daimler first unveiled the Denza brand in early 2012 two years after the set-up of a joint venture to develop EVs for Chinese consumers. Denza in late 2019 began selling the X, a seven-seater SUV with a starting price of RMB 289,800, which was discontinued two years later.

• In late 2021, BYD announced plans to restructure Denza as the company reached a deal to buy an additional 40% shares of the JV from its German partner, Reuters reported. Last August, Denza launched the D9 multi-purpose vehicle, its first model after the rebranding, with a starting price of RMB 329,800, and posted deliveries of nearly 80,000 units as of writing.
• China’s biggest EV maker has been aggressively entering the high-end market with a growing portfolio of luxury brands including Denza, Yangwang, and an upcoming sub-brand called Fang Cheng Bao. The first two models under the Yangwang brand were priced from RMB 1 million; Fang Cheng Bao will specialize in professional and personalized identities, according to the company.

Chinese EV maker Xpeng on Thursday revealed the prices of its G6 sports utility vehicles at a competitive starting price of RMB 209,900 ($28,956), more than 20% cheaper than Tesla’s Model Y in China. The automaker is under growing pressure from investors to drive up sales with the new model after the months-long slump.

Why it matters: Speaking to reporters during an interview on Thursday, Xpeng’s CEO He Xiaopeng said that the G6, which has a similar size and appearance to Tesla’s Model Y, has the potential to achieve monthly deliveries of over 10,000 units.

•  He voiced confidence in a “positive” conversion ratio of its backlog reservations to orders following Thursday’s launch, adding that the company has received more than 35,000 pre-orders for the G6 as of Wednesday after reservations opened on June 9.

Details: The long-anticipated G6 five-seater is almost the same size as the Model Y. The new model measures around 4.75 meters in length, and 1.92 meters in width, and spans a 2.89-meter-long wheelbase.

• The higher-end version is powered by dual electric motors combining an output of 358 kW and maximum torque of 660 Nm, a bit higher than the respective 357 kW and 659Nm of the Model Y.
• The G6 accelerates from 0 to 100 km/h (62 mph) in 3.9 seconds, a bit slower than the Model Y’s 3.7 seconds, yet the crossover has a maximum driving range of 755 kilometers (469 miles), compared with the Model Y’s 660 km.
• Xpeng’s CEO also boasts a faster charging time for the G6, allowing additional travel of 300km on a 10-minute charge and greater powertrain efficiency than peers’ offerings, empowered by an 800-volt silicon carbide power module.
• The charging rate could be more than twice as fast as existing offerings with a 400V charging system, according to a Thursday statement from the company. Xpeng has operated more than 1,000 proprietary charging stations as of Friday and has had plans to add 500 ultra-fast charging stations this year.
• He added that Xpeng owners will be able to access the company’s automotive driver assistance system (ADAS), called the XNGP, for urban traffic roads without the utilization of high-precision maps in 50 major domestic cities during the second half of this year.
• Additionally, Xpeng will begin offering a so-called “AI Valet Driver” function to all XNGP users from the fourth quarter of 2023, allowing its vehicles to navigate on some fixed routes like an “experienced” human driver, according to the company. Rival Li Auto shared similar plans earlier this month.
• The G6 has a price range of between RMB 209,900 and RMB 276,900 ($28,956-$38,134), with the starting price being 7% lower than its previously announced tag of RMB 225,000. By comparison, the China-made Model Y currently costs from RMB 263,900 to RMB 363,900.

Context: Xpeng reported year-to-date deliveries of 32,815 vehicles as of May, a nearly 40% reduction from the same period a year earlier.

• President Brian Gu on May 24 told investors that it expected monthly deliveries to reach 15,000 units starting September when the G6 is scheduled for mass delivery.
• Xpeng’s shares rose 6.85% in Nasdaq during before-hours trading as of writing on Friday, although its shares have fallen 80% since the beginning of 2022.

Chinese automaker GAC Group on Monday showcased an electric, unmanned flying car prototype, a product it says can move both on the ground and through the air, in a futuristic plan to take its urban mobility to another dimension.

Why it matters: The debut makes GAC the latest Chinese automaker to promise riders flying taxis, a still immature technology, after the Toyota manufacturing partner began recruiting for a number of aircraft research and development engineering roles a year ago.

Details: The prototype, dubbed Gove, is being built on a modular system in which the flight and automobile components can be separated, meaning passengers could drive away the concept once it lands.

• GAC envisions a future where passengers can easily access multi-dimensional mobility services ranging from electric air taxis to ride-hailing platforms, according to Wu Jian, president of GAC Research Institute, who spoke at the company’s annual tech day event in Guangzhou.
• The automaker did not reveal many production details about the flying car, with Wu only mentioning that passengers within the Greater Bay Area where GAC is headquartered  would prefer a driving range of at least 200 kilometers (124 miles), Chinese media outlet Caixin reported.

Context: Several Chinese automakers have been working on electric vertical take-off and landing (eVTOLs) air taxis, but none have yet received approval for commercial use from local regulators.

• Aerofugia, an affiliate of Volvo’s parent Geely, said it had filed an application for operations of its prototype test aircraft with the southwestern bureau of the Civil Aviation Administration of China last year. Aerofugia’s AE200 eVTOLs have reached the airworthiness review stage, Caixin reported on April 7.
• Xpeng Aeroht, a startup backed by Chinese electric vehicle maker Xpeng Motors, said in January that it had been granted a regulator-issued certificate to pilot test its Xpeng X2 two-person flying car which has a battery life of 25 minutes. The company plans to start selling the next generation of its flying car with a price tag of around RMB 1 million ($138,596) as early as 2025.

Update: Xpeng Aeroht said on Tuesday that it would not sell its fifth-generation flying car, the Xpeng X2, which was previously referred to in this article as the Traveler X2, but has plans to sell the next generation of its aircraft as early as 2025.

Speed is key if Continental and its auto clients are to have any hope of defending their market share in China, given the competition they face. Auto suppliers might be used to providing very specific solutions for single customers in Europe, “but in China this is not a good idea,” said Frank Petznick, Executive Vice President of the Autonomous Mobility Business Area at Continental AG.

While foreign auto executives express nervousness about the rise of their Chinese rivals, Continental’s global mobility head says he is not surprised. He says he has been “pretty aware of” of the pace of China’s progress in electric vehicle technology for a long time.

Offering products ranging from tires to dashboard displays, Continental is now growing its business in high-performance computers for automated driving, with GAC’s Hyper GT luxury coupe one of its early adopters. Speaking on April 19 on the sidelines of the Auto Shanghai show, Petznick told TechNode that companies must be lean, localized, and standardized in developing technology for the world’s biggest and most vibrant auto market.

Having lived in China for a decade before the Covid-19 outbreak, he also gave a broader perspective on the Chinese autonomous car industry and competition between global Tier-1 suppliers and local tech companies. The German auto parts giant is pushing to develop advanced electric and connected solutions not only for the China operations of multinational car majors but also for local manufacturers with global ambition.

READ MORE: Baidu and Huawei take on global giants with new in-car software offerings at Auto Shanghai 2023

Below are Petznick’s comments on the rapidly changing Chinese auto industry. The text has been condensed and edited for clarity.

China speed

The Chinese market is working completely differently from Europe, and much faster. In order to be prepared for the market, we need local companies that can put pressure on us to speed up and become more dynamic in the market. That’s why we decided to form a joint venture with Horizon Robotics two years ago. We wanted to make a Chinese joint venture that would be closer to the local market.

Global automakers underestimated China’s speed [with regard to EV transition] over the last three years, but now they are getting super nervous because they have seen what’s going on. EV companies in China have a higher demand for autonomous driving. They integrate the entire technology into their cars and can sell to local young people who just want to buy fancy cars.

A lot of the cost of ADAS [Advanced Driver Assistance System] comes from developing specific software, and what Continental can do very well is integration. We figure out what is a common part, roll out standard components in a fast and cost-competitive way, and then add specific functions to make a difference. I think this is the key [to success] in China, but many Western companies have not understood that yet.

Think local

We are working closely with our Chinese customers and developing systems in China and for China. Global automakers in China also want to use local solutions because they are afraid of being too slow and too late. The other thing is that many Chinese brands are going global very fast. It means we could also help some of our Chinese customers use a more global approach.

Every Chinese brand now has a global ambition, though new OEMs [original equipment manufacturers] are much faster at going global than traditional ones. Since the border opened [late last year], we have seen a growing number of Chinese OEMs coming to our headquarters in Frankfurt and Hanover to talk about having a global setup. In the meantime, we have the same discussions when we come here.

We have different solutions for different regions, but the software and functions are the same. We would like to help the global OEMs develop in China and help local OEMs develop in the global world. This is what we are trying to do: bridge the two.

Autonomous driving

There are some very good startup players in the US, but I believe robotaxis will become real in China before the rest of the world. There are still many difficulties in getting approval for vehicles with close to Level 3 driving capabilities. Some cities have allowed this, others have not. It’s very scattered. 

I see significantly faster development in terms of the infrastructure and the regulations needed in China. That’s why I think China could be the world’s first robotaxi-friendly country. The rest of the world could focus more on commercial trucks, which are more of a highway thing and not as complicated as robotaxis in the cities.

We are developing software basically for all levels of autonomous driving by using a lot of the expertise from our partners. The competition is very tough. You always see companies jumping forward and others catching up, but the good news is that if you can survive in this market, you can survive anywhere in the world.

Competing with tech giants

Tech companies such as Huawei and Baidu are going to be Tier-1 suppliers, while we are shifting to be more on the tech side. We need to be more agile and have a more local mindset in order to be fast enough.

We have launched a couple of products, such as a full-fledged smart camera based on processors from a Chinese partner. We are also making high-performance computers where ADAS will also be a part of it. We will be going into series production with the partners we have now. You will see these cars on the road very soon.

I don’t think we have to turn ourselves into a new Baidu. This would be going too far over to the other side. Chinese tech firms are trying to be more Tier-1 and we are trying to be more like a tech company. We are basically learning from each other. We have discussed globally that we have to become a tech player, and in the China context, we need to do that tomorrow.

Geely’s high-end car brand Lynk & Co will be the first sub-brand from Geely to incorporate an in-car operating system called Flyme Auto in its upcoming sports utility vehicle called the 08, the brands announced on March 24. Flyme is developed by Xingji Meizu, a company established by Geely’s founder after Geely acquired smartphone brand Meizu last July. 

Why it matters: Lynk’s use of the Meizu operating system is the result of Geely’s long-term effort to develop more car technology in-house. The collaboration will be a test for both brands — Geely and Meizu — with the former focusing on building its software self-sufficiency and the latter looking to revive its diminishing smartphone business by testing its system on its new owner. 

Details: The operating system, Flyme Auto, is built jointly by Meizu and Ecarx (an auto tech startup backed by Geely). It is an all-new digital cockpit and infotainment system based on the electronic architecture of Meizu.

• The news was made public by Ecarx’s chief executive Shen Ziyu who made the announcement at a corporate event on Friday in the central city of Wuhan. He also added that Lynk & Co’s 08 crossovers would be the first model to use Flyme and Ecarx’s Antora 1000 Pro supercomputer platform.
• Shen said he expected the Lynk & Co 08 crossover, scheduled for release Thursday, to be a flagship example to automakers of how Ecarx could empower the development of in-car technology ranging from autonomous driving to video streaming.

Context: Geely made its first foray into the Chinese smartphone market in late 2021, hiring talent from domestic electronics companies such as ZTE and Xiaomi, and setting up a venture called Xingji Shidai in which chairman Li holds a 55% share. Xingji Shidai acquired the majority stake in Chinese phone maker Meizu last July, TechCrunch reported.

• Many Chinese automakers have been using high-performance chips from US chip giants Nvidia and Qualcomm for their automated driving systems and car dashboards. Some models of Geely’s Lynk have also used Qualcomm. These supply routes are now threatened by US restrictions on chip exports to China. EV upstarts Nio, Xpeng Motors, and Li Auto are also developing chips in-house to ensure their supply of the key components.

Baidu will launch its first electric vehicle model using its new conversational artificial intelligence (AI) technology, with the intention of providing a ChatGPT-like experience that enables natural conversation between owners and their vehicles, an executive from the company said on Tuesday.

Why it matters: This is the latest move by the Chinese technology giant to improve its core search engine business and drive widespread adoption of AI for a range of uses.

Details: Jidu Auto, the electric vehicle arm of Baidu, will be the first company to adopt AI technology at this level of sophistication for smart EVs, chief executive Xia Yiping told reporters at a corporate event in Beijing on Tuesday.

• Xia reaffirmed the company’s plan to deliver its first production model, the Robo-01 sports utility vehicle (SUV), in the third quarter of 2023 with a “very competitive” price tag (our translation).
• Launched in October, the crossover can travel around 600 kilometers (373 miles) on a single charge, as TechNode previously reported. Pricing details have so far only been revealed for a special edition version of the vehicle, which will start from RMB 399,800 ($55,245).
• Xia said he was optimistic about the company’s sales growth in light of Tesla’s significant price cuts, adding that the sudden move reflected the US automaker’s waning competitiveness in the Chinese market.

READ MORE: Baidu’s EV firm Jidu aims to take on Tesla

Context: Baidu said on Feb. 7 that it has been pushing internal testing of its ChatGPT-like chatbot tool called ERNIE Bot, or Wenxin Yiyan, and intends for it to make a public debut next month.

• OpenAI’s ChatGPT bot has sparked a craze in the Chinese internet space, prompting dozens of Chinese tech companies, including Alibaba, NetEase, and JD.com, to announce their own AI chatbots over the past month.
• Media outlets and traditional businesses are also lining up to incorporate the latest AI technology into their services. Trip.com Group, China’s biggest travel services provider, announced today it is among the first batch of partners listed to integrate Baidu’s chatbot technology into its service platform.

READ MORE: Alibaba, Baidu, NetEase, iFlytek…Chinese companies rushing to prove they have tech similar to ChatGPT

Bosch said on Monday it is co-developing a new generation of its advanced driver assistance system (ADAS) with Chinese self-driving car company WeRide, aiming for delivery in late 2023. The system has also secured the first pilot customer, which the German auto parts maker has yet to disclose. 

Why it matters: This is the latest example of German auto firms strengthening their in-car software offerings in the face of competition from Tesla and local peers like Huawei. 

Partnership with WeRide: Delivery of Bosch’s advanced driving technology is scheduled for late 2023 to an undisclosed Chinese car manufacturer. The tech will be similar to Tesla’s Autopilot system and enable cars to operate on both Chinese motorways and busy urban streets.

• The two companies hope to secure two to three new clients by that time. Engineers are currently training and fine-tuning the automated driving algorithms running on production cars, Zheng Xinfen, a senior vice president of Bosch China, told reporters during a media event on Monday.
• Bosch revealed its investment into WeRide in May when Tony Han, chief executive of the autonomous vehicle unicorn, told Chinese media that the collaboration would be the largest of its kind in terms of order volume in China.

An indispensable market: China has been leading the world in electric vehicle adoption and in-car technology development, said Xu Daquan, executive vice president of Bosch China, citing examples such as strong demand from local customers for automated driving software.

• Xu noted that the auto parts maker has been facing urgent requests from local clients to deliver products as quickly as six months as a result of the rising consumer preference. “Accordingly, it makes sense for us to localize research and development with partners to meet the trend.”
• Xu added that German Chancellor Olaf Scholz’s recent trip to Beijing reflected the stance of German industries on business relations with China. “China is such a big market, and it’s vibrant. In that sense decoupling from China should not be a pursuit of German businesses.”

Cash-burning competition: Looking to generate revenue from intelligent and connected car services, industry players have placed their cash on future areas such as autonomous driving and digitalization.

• In April, Volkswagen opened a China subsidiary of its standalone software unit, Cariad, as the German automaker looks to develop products tailored for local customers. This was followed by a $2.3 billion investment to set up a joint venture with Chinese auto tech unicorn Horizon Robotics a few months later. Cariad recorded 978 million euros (roughly $1 billion) in losses for the first half of 2022.
• US-listed Chinese EV trio Nio, Xpeng Motors, and Li Auto favor an in-house strategy. On Friday, Nio’s CEO William Li told analysts that he expected the company’s research and development expenses to remain steady at around RMB 3 billion ($430 million) each quarter, with no significant contribution from automated driving software to its gross margin.
• Huawei has partnered with state-owned automakers BAIC and Changan in automotive software, in addition to selling EVs with automaker Seres. Meanwhile, big automakers SAIC and General Motors have turned to Chinese startup Momenta for partial automation technology.
• Volkswagen in January announced a partnership with Bosch to develop automated driving software and use them on its vehicles since 2023, Reuters reported. Speaking to analysts during an earnings call on Oct. 28, Volkswagen’s CEO Oliver Blume said the partnership with Bosch will be “more for the Western world.”

Jidu Auto, the electric vehicle arm of Chinese search engine giant Baidu, is joining a long list of Chinese companies to take on Tesla by positioning the brand in the premium segment and highlighting its strength in autonomous driving tech.

In recent media appearances, Xia Yiping, chief executive of Jidu, stated that the new automaker can compete with Tesla by leveraging the data and algorithm prowess from its parent company.

A former tech lead of in-car connectivity at Fiat Chrysler, Xia noted that he believes the race among automakers to build intelligent vehicles has only just begun in China.

On Oct. 27, Jidu showcased a special version of its first consumer car Robo-01 that it made in partnership with Chinese automaker Geely. The company plans to launch the standard version next April, which Xia told TechNode could be “very competitive” on price (our translation). He also noted a short-term target of selling at least 10,000 vehicles monthly.

Below is the highlights from a group interview at the car launch event, which have been translated, condensed, and edited for clarity:

Is it too late for Jidu to enter the Chinese EV game as a new competitor?

The EV offerings from our competitors are far less diversified, especially regarding the intelligent and connected capabilities they can offer. The competition has just begun, which I believe will be more about the deployment of semiconductors, algorithms, and computing power rather than vehicle manufacturing, as time goes on, and that’s where our capabilities lie.

We are looking to be a serious player in the medium-to-high-end EV segment, especially in the price range of RMB 250,000 ($34,370) and above, and where in-car intelligent technology has been a major selling point. Our core users are young, educated, tech-savvy, and upper-middle class, and in that sense, there is a big competitive overlap between Jidu and Tesla.

If you compare Jidu’s Robo-01 with Tesla’s Model Y, I would say our vehicle provides a roomier and more luxurious interior, as well as a longer driving range. 

Several competitors have already begun releasing advanced driver assistance systems (ADAS) for city environments. What is your advantage and how do you ensure the reliability of vehicle software?

(Note: Rival Xpeng Motors on Sept. 19 released its so-called City Navigation Guided Pilot, a feature similar to Tesla’s Full Self-Driving that allows vehicles to navigate on both highways and city streets. Huawei’s partner Arcfox closely followed with the release of its Navigation Cruise Assist (NCA) software a week later.)

Jidu’s advanced driver assistance capabilities, including those for highways and urban streets, will be fully ready once we begin vehicle delivery to customers later next year. All the variants of Robo-01 will be equipped with lidar sensors and applicable to all Jidu’s intelligent functionality.

We are developing the most advanced electrical and electronic architecture, where we must ensure the complexity of future vehicle systems and fulfill the higher demand for network bandwidth and functional safety. We run algorithms on Baidu’s supercomputers, and I think that’s one of our advantages.

Auto intelligence is not just about software engineering. You need to fully understand when it comes to where the semiconductor industry is headed and how sensors can better enable autonomous driving, among other fields. Not everyone can do that, but that’s in our DNA.

Jidu will begin delivery of Robo-01 later next year. Can you share insights on production plans, retail networks, and charging infrastructure?

Robo-01 is built based on Geely’s SEA (Sustainable Experience Architecture) platform. In early October, we aligned the production plan of Robo-01 for next year with our manufacturing partner and made reservations for many key components ahead of time.

(Note: In September 2020, Geely launched a modular, open-source vehicle platform for EVs called the Sustainable Experience Architecture (SEA), which has been used to build its own EV sub-brands like Lynk & Co, Zeekr, and Polestar.)

We plan to sell our cars via a direct sales model in the early stages so that we can maintain control over our brand image. Jidu’s first flagship store is about to open in Shanghai and we plan to enter 46 domestic cities by 2023.

When it comes to charging networks, we are building a number of charging points along with our showrooms and service centers, but we will also collaborate with public EV charge point providers to expand our footprint.

Arcfox, an electric vehicle brand launched by Chinese automaker BAIC, said it had started providing car users with its long-awaited Navigation Cruise Assist (NCA) software, a semi-autonomous driving feature developed on Friday by Huawei.

Why it matters: The introduction of Huawei’s automated driving capabilities comes nearly a year later than expected. It will test whether the Chinese telecommunications giant can provide a competitive edge for partnered EV makers.

Details: Starting Sept. 23, the NCA assistant driving feature has been available to owners of the “HI (Huawei Inside)” version of the Arcfox-branded Alpha S sports utility vehicles in Shenzhen. It will later be expanded to Beijing and Shanghai, a company spokesperson told Chinese financial media outlet Caixin, without giving a timeframe.

• The feature allows Arcfox’s cars to change lanes and speed up or slow down on highways and city streets. It also controls acceleration and braking to maintain the desired distance between the SUV and the vehicle ahead, said an official statement.
• Like other similar offerings from rivals such as Nio and Xpeng, the NCA uses a system of 34 sensors and cameras, along with high-definition maps, to realize virtually automated driving on Chinese urban streets.
• As with its competitors, the system is qualified as an advanced driver assistance technology, meaning a driver is still required to take full responsibility for driving tasks and monitor the environment at all times.

Context: Chinese automakers have slowly increased the availability and capabilities of their intelligent driving systems, which are mostly built upon a high-definition map and subject to government approvals for using geographic data, Reuters reported.

• Alibaba-backed Xpeng Motors, on Sept. 19, began testing City Navigation Guided Pilot software with selected drivers in its headquarters city of Guangzhou and is currently waiting for regulators to greenlight a wider release to other cities.
• Arcfox’s driving software was initially set to be available to car owners on major provincial highways and China’s four top-tier cities – Beijing, Shanghai, Guangzhou, and Shenzhen – by the end of 2021 and then to users from at least 20 major cities.
• In April 2021, Chinese automaker BAIC showcased the Alpha S, a premium electric sedan under the Arcfox marque equipped with Huawei’s HI system. Vehicle deliveries began in July, after more than eight months of delay.
• State-owned BAIC sold 6,723 Arcfox-branded vehicles in the first half of this year, falling far behind rivals. 
• Shenzhen-based Huawei also collaborates with automakers Changan and Seres to enter the booming EV market.

As developers struggle to overcome a long list of challenges after years of research and with billions in cash burned, the autonomous vehicle sector is entering a new adjustment phase.

While some investors are rethinking their initial optimism for the autonomous vehicle sector, industry players argue that the technology is finally closing in on some kind of mass adoption. Qcraft and Cowarobot, two Chinese self-driving car companies, talked to TechNode about the challenges and bottlenecks for vehicle autonomy at the BEYOND Expo 2022 tech conference, held online at BEYOND Metaverse. 

The text below has been condensed and edited for clarity.

Yu Qian, co-founder and CEO of Qcraft

I think the biggest issue affecting the large-scale adoption of self-driving cars is the difficulty in handling various unexpected and possibly dangerous situations – we call them corner cases. Once you operate self-driving cars in a broadly defined ODD, the number of corner cases will grow, which could be significantly complex to address.

(Note: Tech companies and car manufacturers typically define an Operational Design Domain (ODD) to indicate where their self-driving car systems can operate safely. Common ODD factors include but are not limited to the time of day, weather, and road features, according to definitions set by the Society of Automotive Engineers.)

2025 could be a very crucial year for the global self-driving space, as we’ve seen some top players quickly expand their robotaxi projects in the US. I expect a fully driverless ride-hailing service will be launched in San Francisco by 2025 or 2026, when the operator will be allowed to charge fares for rides on a large scale, and its revenue could go up relatively quickly. That could be an important milestone for the entire industry and boost the adoption of AVs here in China.

Looking ahead, I also expect more measures and support from the Chinese regulators that will help remove barriers to the deployment of robocars, such as the formal legalization of AVs on public roads and the publication of a government catalog specifically for AVs. Currently, the license plates for AV testing in China are only temporary, meaning full registration is unattainable for highly autonomous cars.

(Note: China’s Ministry of Industry and Information Technology has developed a catalog of electric vehicles. The EVs listed are allowed to go on sale and are deemed eligible for subsidies. The Chinese authorities have not issued legislation that officially authorizes self-driving cars on the country’s roads.)

He Tao, founder and CEO of Cowarobot

If you operate AVs on a very limited ODD, current algorithms in artificial intelligence can deal with traffic situations at the human performance level. And yet, you have to avoid many unexpected corner cases as well as those that are predictable but can’t be resolved at the moment, which means self-driving cars will be limited to fixed routes or a restricted area, and where the weather is stable.

In other words, Level 4 autonomy is now achievable when it comes to specialized uses such as shuttle buses, intra-city delivery, and public sanitation services. In such cases, massive data sets are essential for artificial intelligence innovations. The key question is: how to acquire a large enough amount of data over a long period, and at what cost? 

(Note: Level 4 refers to a fully autonomous system where vehicles travel from point A to point B without requiring human intervention, according to the Society of Automotive Engineers.)

Autonomous driving has been a famously capital-intensive startup business at a time when many players were used to raising outside funds to train their AV systems on huge data sets while being unable to make profits in the near term. The industry, however, is facing investor slowdown, especially in the post-pandemic era this year, as venture capitalists are seeking out those who could provide real business value with the technology.

The trend is that fundraising will be a lot more challenging for highly autonomous driving companies, or more specifically robotaxi projects, as many investors are no longer patient and will look for something more certain with a greater focus on profitability and cash flow. 

Having said that, I believe self-driving car operations for commercial use could help to improve investors’ confidence and extend their patience for the entire industry.

READ MORE: Drive I/O | Meet the Chinese self-driving car startup with Google roots

Large-scale commercial operation of highly autonomous vehicles (AVs) could become a reality “sooner than expected” in China, Baidu’s CEO Robin Li said on Thursday at the 2022 World Artificial Intelligence Conference in Shanghai.

“I think it would take a longer time to commercialize Level 3 autonomous vehicles, because there remain questions about who is liable in the case of accidents involving these vehicles,” Li said (our translation).

Level 4 vehicles, however, make it clear that the manufacturer or the owner, rather than the driver, is responsible in a crash, Li added.

Level 4 refers to a fully autonomous system where vehicles travel from point A to point B without requiring any human intervention. In Level 3, also called the semi-autonomous level, the driver is still required to take over the vehicle in emergencies, according to definitions set by the Society of Automotive Engineers (SAE).

After operating Apollo Go (Luobo Kuaipao, in Chinese), its autonomous ride-hailing service, for the last two years, Baidu said on Tuesday that it has offered more than 1 million public robotaxi rides in a dozen of major Chinese cities as of July. The search engine giant currently operates around 500 self-driving cars in China, with plans to expand that fleet to 3,000 vehicles in 30 cities by 2023.

Baidu may be a pioneer in autonomous cars, but rivals are catching up. Chinese automaker GAC Group plans to begin piloting autonomous ride-hailing vehicles along with human-operated taxis via its mobility platform OnTime in Guangzhou later this year, General Manager Feng Xingya told investors on Tuesday. The carmaker, which produces vehicles in tie-ups with Toyota and Honda in China, has been testing robotaxis with self-driving upstarts WeRide and Pony.ai.

Although excitement over self-driving vehicles has been wearing somewhat thin globally since last year as the technology gets stuck in the slow lane, China is ramping up efforts to support the sector. In August, the central government released its first national rules for commercial autonomous ride-hailing services, while Shenzhen became the first Chinese city to establish a defined legal landscape where legislators can impose a degree of liability for car crashes involving AVs.

Li called for more uniform policies with regards to driverless cars, such as a universal standard that allows companies to remove human safety drivers in more driving scenarios, as the industry continues to face multiple regulatory hurdles to mass deployment. “The window of opportunity is fleeting,” Li added. “More efforts need to be made to push forward legal reform and open the bottleneck on AVs.”   

DeepWay, a Chinese autonomous driving startup backed by Baidu, said on Tuesday that it has raised RMB 460 million (around $67.2 million) in a Series A led by Qiming Venture Partners and joined by multiple veteran investment firms.

Why it matters: DeepWay brands itself as China’s first electric vehicle startup that designs autonomous trucks from scratch for freight delivery, rather than something based on an existing truck model with minor changes, which the company claims leads to more integrated self-driving tech and reduces production costs. 

• DeepWay is also the first startup with authorization from Baidu to conduct “white-box testing,” which means it has working knowledge of the latter’s self-driving technology. By comparison, a tester rarely knows much about the internal design and implementation of an autonomous driving system during black-box testing.

Details: Jointly founded by logistics service provider Shiqiao Group and tech giant Baidu in late 2020, the two-year-old firm is now valued at RMB 3 billion by the latest fundraising round, which was led by Qiming, Chinese tech media outlet QbitAI reported, citing company insiders.

• Other investors include Lenovo Capital, an investment arm of the namesake tech titan, and Vlight Capital, an early backer of EV maker Nio, according to an announcement (in Chinese) on DeepWay’s public WeChat account.
• DeepWay said that the “body-in-white,” as automakers call a car’s basic skeleton, of its first truck model, had rolled off the assembly line at its factory in the eastern city of Yancheng in July.
• Small-scale delivery is scheduled for December. The company is targeting sales of 1,000 roborigs next year, when the vehicles will be capable of driving themselves on Chinese highways with a safety driver behind the wheel.

Context: Several autonomous truck companies have gotten off the starting grid early in the self-driving race in China, but the progress towards fully autonomous freight driving has been slow.

• Nasdaq-listed TuSimple reported a $732.7 million annual loss and $6.3 million in revenue in 2021 with the departure of a longtime executive and delayed production of commercial trucks at full driverless Level 4 operation from 2024 to 2025. The firm is reportedly in talks with Geely to sell its China division.
• JD and Meituan-backed Inceptio said in March that it raised $188 million in a Series B+ co-led by Sequoia Capital China and Legend Capital, another of Lenovo’s venture capital arms. The company claimed its self-driving trucks with a human operator traveled 2 million kilometers (1.24 million miles) on public roads in 18 months as of April.

On Monday, Chinese officials published a set of draft rules that will allow self-driving companies to offer rides and charge fees for fully autonomous vehicles (AVs). The move is part of the country’s ongoing efforts to become a global leader in artificial intelligence. The same day, Baidu announced it was to launch a fully driverless robotaxi service in two major Chinese cities.

Why it matters: The release of China’s first guidelines for commercial robotaxi services could establish a state framework for the rollout of self-driving technology and increase the number of AVs on Chinese roads.

Details: Published by the Ministry of Transport on Monday, the draft regulation said that authorities would “encourage the deployment of autonomous buses on limited access highways, as well as allow paid taxi-hailing services using self-driving cars for low-traffic, controllable scenarios” (our translation).

• The government did not outline detailed criteria for the environmental conditions under which an automated vehicle is designed to operate but said that driving routes must be selected to avoid highly populated sites such as schools and supermarkets.
• Also, the rules emphasized that robotaxi companies must deploy their automated vehicles with drivers based on different levels of vehicle automation. The rules stipulated L3 and L4 level cars need a human operator, while L5 (fully automated cars) cars need either a remote driver or an in-car safety driver. The rules also asked all cars to suspend operations in adverse weather conditions.
• In addition, the companies are obliged to record and share with the government the data logs generated by cars and drivers at least 90 seconds before and 30 seconds after any self-driving malfunctions. These logs must include in-car video footage and pictures of the surrounding environment.
• The draft will be open to public feedback until Sep. 7.

Context: China first began allowing autonomous driving road tests on designated streets in April 2018 and then expanded the testing scope to general highways in early 2021. 

• Several major cities, including Beijing, Shanghai, and Guangzhou have greenlighted self-driving car tests for passenger transport services over the past several years.
• Earlier this month, the city government of Shenzhen also passed new legislation that addresses the liability issues in accidents involving cars with self-driving capabilities.
• The central city of Wuhan and the southwestern municipality of Chongqing are the latest Chinese megacities to take a significant step towards the driverless car era, recently allowing Baidu to charge fees for rides using its driverless vehicles.

On Tuesday, the self-driving car startup Pony.ai announced that it partnered with ride-hailing company Caocao to provide robotaxi services in Beijing.

Why it matters: Autonomous driving is still a long way from commercialization. The deployment of autonomous vehicles on a familiar ride-hailing app might help Pony.ai get closer to making money from its pilot projects.  

Details: Starting from Wednesday, public passengers will have the option to choose Pony.ai’s custom-made test models of robotaxi from the Caocao app on their phones. However, the robotaxis fleet of 30 or so will be restricted to a designated area in southern Beijing. A safety driver behind the wheel is not required, but each car will have a monitor in a passenger seat.

• There is an RMB 18 ($2.67) base fare for all rides, and longer routes will cost RMB 3 per kilometer beyond three kilometers during rush hours and RMB 2.6 during off-peak hours. By comparison, local ride-hailing services usually have a base fare of RMB 14 per ride and RMB 1.8 per kilometer over three kilometers.

Context: In April, Pony.ai and Baidu received permits from the Beijing city authorities to offer driverless rides in an area of 60 square kilometers (23 square miles) in the city’s southeast Yizhuang district. The local government allowed the two companies to charge fares last November.

• Backed by big automakers Toyota and FAW, Pony.ai has been operating a robotaxi pilot project commercially on the outskirts of Guangzhou since May and has partnered with T3, an emerging rival to Didi, as well as Ontime, a ride-hailing service of Chinese automaker GAC.
• Last August, Chinese tech giant Baidu launched its proprietary autonomous ride-hailing platform called Apollo Go, also known as “Luobo Kuaipao” in Chinese. The company completed over 196,000 trips during the first three months of this year. Baidu has launched commercial operations of robotaxis in three domestic cities, including Beijing and the southwestern municipality of Chongqing.
• Rival WeRide also collaborated with GAC for a commercial launch for customers of Ontime in Guangzhou this year, while self-driving unicorn Momenta is piloting a non-commercial fleet of 60 robocars for ride-hailing with SAIC in Shanghai and the nearby city of Suzhou. ByteDance and Meituan-backed Qcraft also deploy a few dozen autonomous vehicles with T3 for free in Suzhou.

Chinese auto chip startup Horizon Robotics on Monday announced that it has secured a new round of funding from state-owned automaker FAW Group, the latest example of local automakers upping their investment in the domestic semiconductor sector to cope with a prolonged global chip shortage.

Why it matters: The investment reflects Chinese automakers’ growing anxiety about the ongoing semiconductor constraints that have crippled them for more than a year and show no signs of abating amid recent Covid-19 outbreaks in the country.

New money influx: Horizon Robotics plans to use the proceeds to speed up the development of new auto chips for artificial intelligence computing and its software development, the company said in an announcement (in Chinese) on Monday. The funding amount remains undisclosed.

• Founded by Yu Kai, a former head of Baidu’s artificial intelligence unit, the seven-year-old startup said that the company’s Journey chips, which could enable rapid processing with vehicles’ advanced driver assistance systems, have shipped more than 1 million units as of last year.
• The company added that it has formed partnerships with more than 20 car manufacturers, including SAIC and Changan, making it the country’s largest producer of automotive-grade AI chips. Its existing investors include SAIC, BYD, and GAC Capital, the venture capital unit of the namesake automaker.

Persistent chip shortages: Last year, China only made 5% of the auto chips it consumed, according to figures published by US research company IC Insights and obtained by Caixin (in Chinese). Chinese automakers’ production has been hit by the low self-sufficiency in auto chips and an ongoing chip shortage, creating more demand for building more domestic auto chip firms to fill in the growing demand. 

• GAC is among a string of automakers being hit by ongoing supply chain issues, with production cut by 160,000 vehicles, equivalent to RMB 20 billion ($2.98 billion), in the first half of this year, chairman Zeng Qinghong said on June 25 at a semiconductor conference in Guangzhou.
• GAC, Toyota’s manufacturing partner in China, expects chip shortages will continue into 2024 and is thus looking for home-produced substitutes to ensure supply. The Guangzhou-based automaker has also invested in local chip foundry CanSemi to develop microchips for future vehicle models on 12-inch wafers.
• GAC is not alone. At the same conference, Bosch China’s president Chen Yudong called for more investment to increase domestic production of semiconductors in the country, estimating that production in China has fallen by 1 million vehicles during the first six months of 2022 because of supply issues.
• Struggling to recover from a lengthy Covid lockdown affecting several of its China plants, Bosch currently meets around one-third of the total demand for its car parts in the country but expects an improvement from July when it thinks supply could meet 60% at most of the market demand.

Context: China has for years been building an independent domestic chip supply chain, reporting a 33.3% year-on-year increase in domestic output of integrated circuits (ICs) last year, according to data released by China’s National Bureau of Statistics.

• The central government recently promised to take more measures to help domestic makers expand capacity and boost innovation, China Securities Journal reported Tuesday, citing Guo Shougang, a deputy director at the Ministry of Industry and Information Technology.

Two people were killed after a Nio testing car plummeted off the third floor of a parking garage at the company’s Shanghai headquarters on Wednesday. The electric vehicle maker claimed that its vehicle was not at fault in the accident.

Why it matters: If the vehicle was not at fault, the incident should not greatly impact Nio’s vehicle sales. However, it potentially delivers another blow to the company’s reputation following a high-profile accident involving a Nio car last year.

Details: Based on preliminary investigations by the local police, there is no indication that the deaths of the two testing workers were related to an issue with the vehicle, Nio said on Thursday in an announcement published on the Chinese Twitter-like platform Weibo. It was not immediately clear what caused the crash.

• The two workers – one a Nio employee, the other a staff of a partner company – were testing Nio’s digital cockpit features and were killed Wednesday afternoon after the car drove off from the third floor of the garage and landed on the company’s campus in the city’s Jiading district. The tragic case was an “accident unrelated to the vehicle,” the company added in their statement.
• Nio shares closed slightly up 2.2% on Thursday amid a broad market rally that saw rivals Xpeng Motors and Li Auto jump 7.8% and 6.6%, respectively.

Context: Last year, Nio’s credibility took a hit when a 31-year-old Chinese entrepreneur died in a car crash while driving his Nio ES8 with the car’s driver-assistance functions activated. Nio notes in its user manual that the company’s technology currently requires active driver supervision and does not make the vehicle autonomous.

• Nio’s vehicle margin fell to 18.1% in the first quarter of this year, down from the 21.2% during the same period last year and the 20.9% recorded for the fourth quarter of 2021, due to the industry’s ongoing supply chain constraints, worsened by China’s Covid lockdowns.
• The Tencent-backed EV maker delivered 37,866 vehicles for the first five months of this year, an 11.8% increase from a year earlier. Deliveries of rivals Xpeng Motors and Li Auto more than doubled year-on-year to 53,688 and 47,379 vehicles over the same period.
• Earlier this month, the Shanghai-headquartered automaker launched the ES7, a new electric sports utility vehicle that boasts improved comfort and advanced self-driving technologies, with a starting price of RMB 468,000 ($69,825). 

US-listed Chinese electric vehicle makers Xpeng Motors, Li Auto, and Nio are undergoing significant restructuring as rising costs of raw materials and supply chain disruptions cut into profit margins. Meanwhile, EV battery makers are upping their investment to increase production capacities as China continues an accelerated shift to EVs.

Chinese EV makers are restructuring their businesses as challenges grow

Having enjoyed exponential growth over the past two years, Chinese electric vehicle startups are showing signs of contraction as supply chain constraints and rising raw material costs (partly worsened by the Covid-19 pandemic) continue to weigh on the industry. 

Facing a serious slowdown in economic growth and a resurgence of Covid-19 outbreaks, the US-listed Chinese EV trio of Nio, Li Auto, and Xpeng Motors are undertaking thorough reorganizations, laying off workers, and shifting away from non-core projects to meet their growth targets. The companies have been handling these challenges relatively well, but the outlook going forward is a bit unclear.

Xpeng Motors: Xpeng is facing a significant setback in its global ambition. Several senior executives, including vice president of overseas sales He Liyang, recently left the Guangzhou-based automaker amid a comprehensive restructuring across the company meant to streamline operations and save expenses, Chinese media LatePost reported on May 26, citing people familiar with the matter. The departures come after the EV upstart experienced lackluster sales of merely 438 vehicles in Norway in 2021, while leader Tesla took a nearly 20% market share in the country as it delivered more than 20,000 EVs over the same period, according to official figures.

In an effort to pare back losses, the Alibaba-backed EV maker is trimming its sizable staff in several major divisions, including a software team developing intelligent cockpit solutions and its data management department. As part of the change, Zhao Hengyi, a tech lead on Xpeng’s in-car voice assistant, left his position in March. The company also cut some of its plans of cultivating some fresh graduates, with dozens of them recently having their job offers rescinded.

Xpeng has been known to spend cash more quickly compared with peers. It posted a record loss of RMB 1.7 billion ($268.3 million) in the first quarter of 2022, widening from RMB 1.29 billion in the previous quarter. Analysts had warned of more losses to come from April to June due to high material costs and recent Covid lockdowns in China. The company earned a gross margin of only 12.2% during the first three months of this year, far lower than the 22.6% and 14.6% posted by rivals Li Auto and Nio, respectively.

Li Auto: A relative latecomer in a competitive industry, Li Auto is also facing a critical juncture and has scaled down some of its recruitment plans as it anticipates tough times ahead, the LatePost report said. Eight-year-old Li Auto recently lowered its delivery target for this year by 15% to 170,000 vehicles and planned to recruit 2,000 fewer people than it had initially planned, as the company worried about sales performance in the face of an economic downturn.

In anticipation of it becoming harder to get capital as investor sentiment worsens, Li Auto is also downsizing. Since March, the company has cut 20% of its full-time employees in its enterprise system development team after a large hiring spree, while dismissing some workers in its camera research and development team, formerly set up by then technology chief Wang Kai, LatePost reported.

The Meituan-backed EV maker was hit harder than rivals by the recent wave of Covid-19 lockdowns in the country, seeing its April deliveries down  62% and its second production model delayed amid the current supply chain disruption. The cuts could help the automaker reduce costs and survive a looming recession, yet investors were disappointed when the automaker forecast an even lower revenue target and warned of a worse margin for the second quarter of 2022.

Nio: Once the front-runner in the field of Chinese EV startups, Nio is making a pivot to battery-making, with plans to develop and potentially manufacture its own battery packs. The move marks a revamp of company strategy that comes as soaring material costs and supply chain bottlenecks slowing its factory output. Speaking to analysts during an earnings call on June 9, chief executive William Li said that the company now operates a team of over 400 employees on battery technologies and plans to launch an 800-volt battery pack for fast charging in 2024.

A new $32.8 million research facility is also slated for construction near its Shanghai headquarters this summer, aimed at developing lithium-ion battery cells and packs. This is in line with the EV maker’s battery strategy of both in-house development and outsourcing, a move that Li believes will benefit Nio’s overall competitiveness and profit-making capability in the long term. The company has warned that battery price hikes will continue to weigh on its margins in the second quarter.

Meanwhile, the company is reorganizing its autonomous driving team, which is at the core of its long-term ambition to become China’s top luxury car brand, following the departure of a long-time vice president of engineering in April. A team of more than 400 engineers, who work on diverse technology domains including sensors, algorithms, and system integration, has been reassigned to other departments to flatten the management structure for communication and combine functions where appropriate, Chinese media 36Kr reported.

Battery makers racing to expand capacity

Despite automakers’ short-term adjustments, the long-term prospects for China’s EV market remain robust with strong consumer demand. In response, major battery makers have kicked off a fierce expansion race in the hope of scaling up supply to meet the demand and take a larger market share. Government-backed industry group the China Passenger Car Association (CPCA) has maintained its forecast of 5.5 million passenger electric vehicle sales for this year in China despite the ongoing Covid-19 outbreaks across the country. 

Here are some of the major players’ expansion plans:

CATL is moving to become more directly involved in lithium mining in order to make its own supply of the EV battery material, thanks to soaring prices. The Chinese battery giant recently won approval to build a new lithium plant with a mining claim on nearly 1,600 acres in the central province of Jiangxi, state media CLS reported on June 1, citing government documents. The new RMB 2 billion ($297 million) facility would be capable of producing 30,000 tons of battery-grade lithium carbonate annually and is scheduled to be in production in  2023.

BYD is making a similar move and is said to be on the verge of closing deals to acquire six lithium mines in Africa, which experts estimate could allow the company to produce about 1 million tons of lithium carbonate, which translates into at least 27.78 million EVs. A BYD executive confirmed that it will supply lithium-ion batteries to Tesla “very soon” earlier this month. There has also been speculation that Nio and Xiaomi are looking at sourcing batteries from the company as well.

Gotion High-Tech is the latest Chinese battery maker to expand its local production by partnering with prominent players like Volkswagen and Great Wall Motor. The battery supplier announced (in Chinese) on May 31 that two new facilities have been put into production with a combined capacity of 30 gigawatt-hours (GWh) each year. The company is on track to double its total capacity to 100 GWh by this year and expand that number to 300 GWh in 2025.

On Tuesday, Li Auto announced the L9, a full-size, three-row sports utility vehicle, as part of its stated ambitious plan to achieve 1.6 million vehicle sales by 2025. The car’s starting price is less than half that of similar offerings from the likes of BMW and Mercedes-Benz.

Why it matters: With delivery planned to begin in August, the six-passenger L9 SUV will be the second production model from Li Auto and the Chinese EV maker appears to be confident that it might become a hit.

• Speaking to reporters on Wednesday, chief executive Li Xiang declined to reveal specifics about order volume, but said that the L9 will outsell its existing Li One, which was the top-selling large new energy SUV in China last year, according to official figures.

Details: The L9, a plug-in hybrid, is described by the company as the pinnacle of large luxury SUVs, with what it says is a spacious interior specifically for Chinese three-generation family households. The automaker said the model offers passengers more room than other luxury automaker offerings.

• The plug-in hybrid has a driving range of 215 kilometers (134 miles) on a full charge but can drive for about 1,315 miles with a full fuel tank and a full charge, a 20% increase compared with the company’s first model. It accelerates to 100 km in 5.3 seconds, according to Li Auto. 
• The model comes with many high-end tech features. It has five screens, including two 15.7-inch touch-sensitive ones in the middle of the dashboard that control the in-car entertainment system, two smaller ones around the steering wheel,  and an OLED television screen for rear-seat passengers.
• The vehicle uses a combination of 24 sensors to detect and predict road conditions, including eight 8-megapixel cameras, a long-range lidar unit, and two Nvidia Orin AI chips to enable autonomous driving.
• The L9 will only enable assisted driving on highways, once delivered; the company has not revealed when its car system will support autonomous driving in city traffic. Its rival Xpeng Motors plans to send an over-the-air update that would allow its vehicles to drive autonomously on urban roads later this year.
• The vehicle will sell for RMB 459,800 ($68,418), a price that the seven-year-old automaker claims is lower than any other similar SUV on the market. For comparison, the BMW X7 and the Mercedes-Benz GLS crossovers start at RMB 1 million and RMB 1.07 million in China, respectively.

Context: Meituan-backed Li Auto has been at the forefront of the Chinese EV field with just one model on sale, recording deliveries of 90,491 Li One vehicles in 2021, a 177.4% increase from a year earlier. The sales number is close to the sales of all three of rival Nio’s models over the same period combined.

• CEO Li Xiang has set an ambitious target of delivering 1.6 million vehicles annually by 2025, according to an internal memo obtained by Chinese media outlet Caixin in February 2021.
• Li said earlier this month that monthly delivery of the latest model could reach more than 10,000 units starting from September, although investors now reportedly expect that number to be around 5,000-6,000 units due to supply chain constraints and Covid-19 control measures.
• Earlier this month, Nio also launched a new SUV model, the ES7, with a starting price of RMB 468,000. Alibaba-backed Xpeng said in April that it will launch its second SUV model, the G9, this month.

READ MORE: Drive I/O | Nio, Xpeng, and Li Auto face more challenges after a mixed 2021

Drone maker DJI is about to see its in-car system used on a mass-produced electric vehicle for the first time through a partnership with SAIC-GM-Wuling (SGMW), General Motors’ China joint venture with SAIC Motor and Liuzhou Wuling Automobile, a small Chinese automobile company. On Thursday, the automaker announced that it will launch an EV using DJI’s automated driving technology, making it the drone maker’s first major project in the competitive sector.  

Why it matters: The launch marks a first milestone for the world’s largest maker of consumer drones in its push into the Chinese EV space and reflects the growing trend of traditional automakers partnering with tech companies to bring self-driving cars to market.

Details: The automaker said that it has worked hand-in-hand with DJI in developing intelligent vehicles since 2019,  investing “several billions of RMB” in the project and having undergone 1 million kilometers (631,371 miles) of vehicle testing, in a statement (in Chinese) published Thursday on SGMW’s WeChat account.

• The statement is sparse on details about the collaboration, but Chinese financial media outlet Caixin reported that the automaker plans to fit DJI-developed automated driving functions on Wuling Baojun Kiwi EV, a mini two-door EV launched last August.
• Full specifications, pricing details, and the launch date of the revamped model remain unclear. The original Kiwi EV is priced between RMB 77,800 and RMB 86,800 ($11,639 and $12,968) and has an estimated driving range of 305 kilometers (190 miles), according to the company.
• Company insiders told Caixin that the in-car software will allow assisted lane changing, automated driving in congested traffic, and other automated driving technologies and that the vehicle’s features will receive regular software updates.   

Context: DJI first launched its auto unit in 2016 and operated with nearly 1,000 employees as of last year, as the Shenzhen drone unicorn steps up its efforts to enter China’s booming EV market.

• SGMW’s affordable Hongguang Mini EV was the best-selling EV model in China in 2021. It recorded sales of 395,451 units last year, easily beating BYD’s Qin sedan and Tesla’s popular Model 3, which sold 187,227 and 150,890 units, respectively, according to figures from the China Passenger Car Association.
• Chinese tech giants Huawei and Baidu also continue expanding into the industry, while young EV makers Xpeng Motors and Nio have catapulted ahead of the competition by developing their own in-house autonomous driving systems.
• Huawei and its manufacturing partner Chongqing Sokon are on track to roll out their second EV model M7 by the end of this month, while the telecommunications giant has also established partnerships with state-owned automakers BAIC, Changan, and GAC, Chinese media reported on May 28, citing chief executive of consumer business at the firm, Richard Yu.
• Baidu has teamed with domestic automakers such as BYD and Dongfeng, and plans to roll out its first consumer car with partner Geely later this year, while also supplying vehicle software technology to WM Motor, an EV startup backed by the search engine giant.

Xpeng Motors released first-quarter earnings on Monday night, giving a second-quarter forecast that fell far below estimate. The company said it has made progress in ensuring the production against the backdrop of a global shortage of chip and battery supplies, but investors remained concerned that a prolonged supply crunch and China’s strict Covid-19 measures will hurt margins this quarter.  

Why it matters: Xpeng is joining a long list of Chinese tech companies facing a challenging quarter with production cuts and profits squeezed. The company expects deliveries to fall between 31,000 and 34,000 units in the three months until June, compared to the 34,561 vehicle deliveries in the first quarter of 2022.

Details: On Monday, Xpeng reported revenue of RMB 7.45 billion ($1.2 billion) in the first quarter of 2022, up 152.6% from the same quarter last year. However, net loss more than doubled year-on-year to RMB 1.7 billion. The company’s share prices fell 5.5% on Monday.

• Xpeng expects second-quarter revenue to reach up to RMB 7.5 billion, well below analysts’ average estimate of RMB 8.3 billion, according to data compiled by Bloomberg. Gross margin will also be impacted due to existing supply chain constraints, but is set to improve in September with the delivery of higher-priced new models to customers, said Dennis Lu, vice president of finance at Xpeng.
• Xpeng executives said on Monday that the company has expanded efforts to reduce the impact of supply-chain difficulties and China’s Covid-19 lockdowns. In addition, it has contracted multiple new suppliers and implemented more flexible design and manufacturing for its EVs.
• During an earnings call, Chief executive He Xiaopeng said that the company has begun to see significant progress as it aims to secure enough battery supply to meet demand in the current quarter. More “optimization” is likely to happen during the second half of 2022, thanks to a multi-sourcing strategy and the decline of battery prices, He said.
• However, the ongoing semiconductor shortage is getting worse, as the electric vehicle (EV) maker can only monitor the impact on the production of chip supply chains one week into the future. He added that the current semiconductor supply bottleneck could last into 2023 or even longer, in contrast to a previous estimate that the issue could be resolved or alleviated by the end of 2022.

Context: Earlier this month, rival EV maker Li Auto also delivered a gloomy revenue forecast for the second quarter, expecting up to RMB 7.04 billion, which is 36% lower than previous estimates, with the company citing supply chain issues related to Covid-19 lockdowns in China. Li Auto’s vehicle delivery plunged by 62% in April from the previous month to 4,167 vehicles, with Nio’s and Xpeng’s volumes nearly cut in half over the same period.  

READ MORE: Nio, Xpeng, Li Auto see dismal April deliveries as coronavirus lockdowns disrupt production

Top automakers such as Tesla and SAIC (Volkswagen’s partner in China) are slowly rolling towards a restart after weeks of shutdowns of their plants in Shanghai, China’s worst coronavirus outbreak site, in two years. Baidu and self-driving unicorn Pony.ai received permits to offer fully autonomous rides to the Beijing public in late April, the first service of its kind in the country. Domestic battery suppliers saw profits plunge in the first quarter amid rising raw material costs, thanks to a strong demand for electric vehicles (EVs) that utterly outstrips supply.

Shanghai’s Covid outbreak continues to weigh on auto production through May

As Tesla and Volkswagen’s plants in Shanghai slowly resume production, China’s auto industry is struggling to regain the momentum lost during a citywide lockdown that has dealt a significant blow to local businesses over the past two months. Government officials said on May 13 that employees from 95% of the companies on a whitelist of 666 firms prioritized for business resumption are now getting back to work, with automakers and suppliers accounting for more than a third of the total.

China’s biggest automaker SAIC said on May 13 that its joint facilities with Volkswagen and General Motors have restarted production in mid-April in a single shift rather than their usual two shifts, with each plant assembling at least 2,000 vehicles every day. As a result, Tesla shipped out another 4,000 locally-made vehicles to Europe on May 15, four days after its first shipment of 4,767 cars set sail from the Port of Shanghai – the first to do so since the start of the sweeping lockdowns in the city, Chinese media reported.

Supply chain hurdles: Disruption related to labor and supply chains continues to impact auto firms, as many workers can’t return to their workplaces due to inflexible Covid-19 control restrictions in many parts of the city. Tesla’s Shanghai facility reportedly idled most of its production lines for a few days earlier this month due to insufficient supplies, when Aptiv, one of its key parts suppliers, halted shipments of some parts due to new Covid cases at its local plant.

Auto supplier giant Bosch has only experienced a partial recovery with output at around 30%-75% of its pre-pandemic levels at several manufacturing sites, a result of worker shortage and supply chain crunch, its China president Chen Yudong said at a May 11 press conference, while also calling for the easing of Covid restrictions. 

The auto firms that have resumed operations represent only a fraction of the 20,000 parts suppliers, big and small, located in Shanghai and nearby regions, state-owned media outlet China Newsweek reported on May 11, citing several experts.

Weak Q2 guidance: Analysts expect output to slightly recover in May but believe a full recovery is still some way off, as the industry struggles with massive uncertainty caused by Covid lockdowns. Li Auto, which has a production base in the eastern city of Changzhou, was among the automakers hit hard by the lockdown, releasing poor second-quarter revenue guidance on May 11 due to a likely disruption to parts supplies.

And yet, there is still a chance to make up for lost sales in China during the rest of the year if automakers can ramp up car output, given that a growing number of consumers feel safer traveling alone than taking public transport, experts say. In April, Tesla maintained its forecast of at least 50% annual growth for vehicle deliveries this year, despite saying that production volume could take a hit of 8% in the second quarter due to a month-long production halt at its Shanghai facility. The China Passenger Car Association predicted that total passenger vehicle sales may face zero growth to remain at 20.1 million units this year, compared with 2021’s growth rate of 4.4%.

Driverless cars get a push from China’s capital

In a rare step, Beijing authorities announced on April 28 that Baidu and Pony.ai have been authorized to participate in the country’s first pilot program to provide driverless rides to the public in test vehicles. Following the move, Baidu and Pony.ai began by operating 10 and four autonomous vehicles, respectively. The vehicles operate without safety drivers on public roads in an area of 23 square miles in the city’s southeast Yizhuang district. However, each vehicle has a company employee overseeing the journey in a passenger seat, and the firms are not allowed to charge a fee for now.

Chinese self-driving car companies have faced a long and arduous reality check since a wave of early hype and hopes of scaling the technology. Now, regulators are giving the industry a boost by permitting the offering of autonomous services to the public in the country’s capital city – with no human safety driver at the wheel. Concurrently, the race to prove robotaxis are a viable business is intensifying among the top contenders.

AVs undergo reality check: Despite the milestone in Beijing, few of China’s self-driving car startups are making any money, and venture capitalists have been cooling on the companies over the past year, particularly those with little to show commercial prospects. Total investment activity for robotaxi companies fell by 22% annually to $8.4 billion in much of 2021, data compiled by startup data platform PitchBook and obtained by Reuters showed.

Major players are working hard to live up to their promises. WeRide became China’s first self-driving company by testing completely driverless cars in the southern Chinese city of Guangzhou in July 2020. In January of this year, its fleet of 300 autonomous vehicles had logged 10 million kilometers after four years of testing. For Baidu, that number is more than double, and the tech giant said that it provided more than 320,000 autonomous rides in eight domestic cities as of last year, with plans to expand the service to 65 cities by 2025. 

Chinese battery makers’ profits slump amid supply chain issues

Drops in Q1 profit: Despite being buoyed by strong demand for electric vehicles in the country, Chinese battery makers are facing a profit squeeze as the global supply chain continues to buckle under the pressure of rising costs, limited raw materials, and manufacturing disruption. On April 29, CATL reported a year-on-year profit tumble of 41% to RMB 977 million for the three months that ended in March, which came in far below expectations of a RMB 5 billion profit from multiple analysts. It was CATL’s first quarterly decline in net profit since 2020. Meanwhile, profits of the Volkswagen-backed Gotion declined 33%, while Sunwoda, a lesser-known supplier invested in by EV maker Li Auto, also saw a 26% decline in profits despite double-digit revenue growth.

Q2 easing expected: Margins for battery makers have been dragged down by surging raw material costs made worse by the Russia-Ukraine conflict and a global pandemic. An index for battery-grade lithium prices increased by 127% in the first quarter of this year, after a 280% surge in 2021, according to data provider Benchmark Mineral Intelligence. The costs of nickel and cobalt also exploded during the first three months of this year, which hit battery suppliers hard since many of them had negotiated quarterly price terms with automakers for the period up to last December.

Analysts estimate that the supply shortage of raw materials will slightly ease starting in the second quarter of 2022 as battery suppliers step up efforts to secure minerals and expand production capacity. Margins are also expected to improve as most battery makers increased the prices of their products in March by at least 15% for the second quarter, China Securities Journal reported on April 28, citing company sources. This rally in material costs has been reflected in the recent price increases for EVs, ranging from RMB 2,000 to RMB 30,000, although analysts expect that EV sales will maintain their growth momentum this year, boosted by inflated oil prices.

Qcraft, a Chinese autonomous driving startup, said at a Wednesday conference that it is partnering with ride-hailing firm T3 to bring self-driving vehicles onto the latter’s ride-share network in the eastern city of Suzhou. T3 users within the range of those vehicles’ routes will soon be able to select one for a ride.

Why it matters: The partnership is the latest example of driverless tech firms rushing to work with more consumer-facing companies as they aim to commercialize autonomous driving tech. 

Details: Starting from July, Qcraft and T3 will begin offering rides to public passengers using self-driving cars within a restricted area in Suzhou, a neighboring city of Shanghai, where the companies are already testing the vehicles.

• The initial phase of the pilot deployment is expected to allow the companies to fine-tune their robotaxi offering by collecting rider feedback and improving user experience ahead of a commercial launch, according to an announcement (in Chinese). 
• On Wednesday, Qcraft also announced plans to test its self-driving system for consumer cars beginning in the third quarter of this year, collaborating with Chinese chipmaker Horizon Robotics. Backed by leading tech companies Meituan and ByteDance, the three-year-old Qcraft is testing a fleet of more than 100 autonomous mini-buses and sedans in around 10 major Chinese cities.
• T3 has emerged as a significant rival to Didi and is backed by state auto majors FAW, Dongfeng, and Changan. The ride-hailer completes over 3 million rides every day with operations in more than 80 Chinese cities, its vice president Li Jinfeng told reporters at a press briefing on Wednesday. To compare, Didi reportedly provided 20 million trips per day in January.

Context: Other Chinese self-driving car companies are racing to launch commercial autonomous ride-share services either by themselves or with partners.

• Baidu began operating fully autonomous taxis in the suburbs of Beijing in late April, a few months after being allowed to charge customers fares for rides in the capital city. The tech giant said it had offered around 213,000 rides in eight domestic cities in the fourth quarter of 2021.
• Self-driving unicorns Pony.ai and WeRide have turned to ride-hailing service OnTime for a wider group of users, recently participating in its RMB 1 billion ($153 million) Series A, TechCrunch reported on April 27. Operating in the southern Guangdong province, OnTime was launched by state-owned automaker GAC in mid-2019 and backed by Tencent.

Huawei has lowered its forecast for its car deliveries in partnership with various automakers this year due to worsening supply chain issues impacting the country’s auto industry, according to senior executives.

Details: Speaking to analysts on Tuesday, Huawei’s rotating chairman Hu Houkun confirmed that the company has scaled back its expectations for car sales and is now seeking support and understanding from the auto industry as it “is susceptible to making mistakes” as a newcomer (our translation).

• Hu also underscored plans to launch new vehicle models with several partners this year, without revealing any further details, and reiterated Huawei’s position to partner with automakers on vehicle technology rather than making its own cars, Chinese media outlet Caixin reported on Wednesday.
• Hu made the comment a week after Huawei’s chief of consumer and auto business Richard Yu admitted for the first time that ongoing supply chain disruptions, such as microchip shortages and soaring prices, have impacted the firm’s sales targets for its auto business.
• Speaking to a Chinese auto journalist on April 18, Yu talked about the Aito M5, the first premium electric SUV co-launched by Huawei and its partner Sokon in December, saying that sales of between 100,000 and 200,000 vehicles for the year would likely be a best case scenario.
• That is a significant cut from its original goal of selling 300,000 Aito-branded vehicles annually, which was announced during a company meeting early this year, Caixin reported (in Chinese).

Context: Sales of the Aito M5 appear to have run into a brick wall, with just over 5,000 vehicles sold during the first quarter of 2022. The luxury crossover, powered by Huawei’s HarmonyOS operating system, was launched at a price of RMB 250,000 ($39,053), but the base model cost will be increased by RMB 10,000 starting from May 5. The companies behind the model blamed soaring raw material costs for the price hike.

• Meanwhile, Sokon has ended production of the Seres SF5, its first EV model jointly developed with Huawei, after months of lackluster sales. The Seres SF5 sold around 8,000 units in 2021, while EV startups Nio, Xpeng Motors, and Li Auto each recorded deliveries of nearly 100,000 cars.
• The Chinese telecommunication giant also has plans to launch new EV models with state-owned automakers Changan and GAC this year, although another partner BAIC has not yet delivered new Arcfox-branded vehicle editions beyond the initial standard model, which were set to be equipped with Huawei’s advanced driver assistance system and released last April.

Nickel prices climbed to an all-time high and could further increase the cost of electric vehicles (EV) and force automakers to cut earnings forecasts. Ride-hailing giant Didi became the latest Chinese tech company to enter consumer EV space; it plans to deliver an entry-level sedan next year. Shares of Nio closed flat in the company’s Hong Kong trading debut. Its listing follows the steps of Xpeng Motors and Li Auto. All hope to attract more investors in China amid growing financial market tensions between China and the US.

Soaring nickel prices cast shadow over Chinese EV players

As the price of nickel jumped to an all-time high since early March, auto industry insiders expressed concerns that an escalating Russia–Ukraine conflict could disrupt supplies of the metal, a key component of EV batteries. While watchers have differing views about the impact on EV adoption, most expect battery prices to remain high and to weigh on the margins of Chinese EV makers for the rest of the year.

Nickel craze: Nickel markets had a wild ride early this month. On March 8, the price of three-month nickel on the London Metal Exchange (LME) more than doubled in a short period, reaching an all-time high of $101,365. The unusual surge prompted LME to halt trading for seven days, set new price limits, and adjust prices. When it reopened, the price dropped back down to around $80,000, yet still about 300% higher than the $20,000 price in late February. 

• China’s nickel producer Tsingshan Holding Group was caught on the wrong side of the market, having built the biggest short position in the metal and betting the price would fall since last year. Tsingshan on March 14 said it reached a deal with its banks to backstop its short position after struggling to pay margin calls on its position during the nickel price surge a week earlier. The company faced an $8 billion paper loss based on nickel’s price of $48,002 on March 14, the Wall Street Journal reported.
• The unprecedented price surge was partly due to concerns from Russia. Western nations imposed sanctions on Russia after it started a war with Ukraine in late February. Russia is the world’s biggest exporter of nickel, prompting buyers to worry that Russian nickel suppliers could be hit by sanctions and transport disruptions.
• The already tight supply of high-purity nickel, fueled by the rising sales of EVs, contributed to the highs in the metal’s price. Nickel is increasingly used in EV batteries as it ensures high-energy density that allows the vehicle to travel further.

Higher cost for EVs: Nickel’s price surge is magnifying the current supply chain woes that have dramatically pushed up automakers’ production costs. The global semiconductor shortage and a boom in the prices of other metals have been the principal factors. 

• The input cost of an EV equipped with a 60 kilowatt hour (kWh) battery pack will increase RMB 9,000 ($1,418) due to nickel’s price growing from about $20,000 early this year to the recent price point of around $50,000, according to estimates from China International Capital Corporation (CICC). Nickel’s price will probably stay high over the short term, partly thanks to low inventories in the country, but the high price may be hard to maintain long-term, CICC wrote in a March 9 report.
• Many experts anticipate an accelerated shift towards lithium phosphate (LFP) batteries from the current mainstream types that use nickel and cobalt as core materials. Nickel-free LFP batteries generally provide a lower driving range and cost less to produce than its counterparts, and yet are now also under price pressure thanks to rising lithium prices, the Wall Street Journal reported.
• Average prices of lithium-ion battery packs are expected to slightly grow to $135 per kWh this year from $132 a year ago, ending nearly a decade of price declines, Bloomberg New Energy Finance estimated in a report published on Nov. 30, 2021.

Impact on EVs: Predictions vary among experts of how nickel’s price hikes could affect the EV supply chain and affordability for customers.

• Morgan Stanley automotive analyst Adam Jonas, one of the leading voices warning investors of massive earnings drops for automakers, expects at least a $1,000 increase this year in the input cost of an average EV in the US. If sanctions against Russia are extended to nickel, it’s “probably time for investors to take auto company earnings forecasts down,” Jonas wrote in a March 7 note.
• Other experts say the overall impact will be limited. The high price of nickel is likely to prove a temporary phenomenon since the metal has long seen high output, Cui Dongshu, secretary general of the China Passenger Car Association (CPCA), told reporters during an online conference on March 8. There are currently few signs of risks to output from other major nickel producers such as Indonesia, Cui added.

Didi’s first consumer EV could hit the roads in 2023

News: China’s red-hot EV market just added another competitor as struggling ride-hailing platform Didi reportedly plans to develop its first consumer car in-house. The compact EV could begin mass delivery as early as next June, according to a local media report on March 15. With an estimated price tag of RMB 150,000 ($23,580), the new model will be an entry-level compact sedan competing with existing offerings such as BYD’s popular Qin EV, the report said. The company is said to have more than 1,700 staff dedicated to the project at its Beijing headquarters. In addition, it is considering a deal to buy Zhijun Auto, a little-known EV manufacturer with a plant in central Jiangxi province.

Insights: The launch of a consumer car might create a new revenue stream for Didi as its core business falters. The project can also cover the high cost of developing autonomous driving technology, an initiative the company has undertaken since 2016. The move would also see the Chinese mobility giant lining itself up to compete with big auto names such as BYD, which is also its manufacturing partner.

Didi had a rocky start in its first attempt to produce an EV with BYD. The D1 was a purpose-built electric crossover for ride-hailing services developed by the two companies. It entered into production in late 2020, six months later than expected, the report said. 

Didi’s ride-hailing volume reportedly declined to 20 million trips per day in January, a 20% plunge from daily figures in the first quarter of 2021. Over the same period, the company’s ride-hailing market share in China has shrunk from nearly 90% to 70% due to Beijing’s ongoing cybersecurity review of the company that began last July. 

Nio shares debut in Hong Kong secondary listing

News: Chinese EV maker Nio made a weak debut in Hong Kong on March 10, closing down 0.69%. The listing took place after a long and winding journey. Already listed on the New York Stock Exchange, Nio has followed in the steps of rivals Xpeng Motors and Li Auto by tapping into Hong Kong’s capital markets. However, Nio did not sell new shares or raise money, and it chose to list by introduction. Xpeng and Li Auto, on the other hand, raised HK$14 billion and HK$11.8 billion, respectively, by selling shares in Hong Kong in the summer of 2021.  

Insights: Nio explained the move by saying it hopes to attract more investors by enabling more listing locations and flexible trading hours. A Singapore listing may be another possibility. The Hong Kong locale does bring the Shanghai-based EV maker closer to mainland investors and provides the automaker insurance against the risk of delisting in the US. But Nio said it had “a sufficient pool of working capital,” according to financial media Caixin (our translation), and did not have an urgent need to raise additional funds. 

Plagued by a shortage of semiconductor chips and batteries, among other supply-chain headaches, Nio has posted lackluster monthly sales volumes for several months. Sales of Nio’s existing three models have been slow. Its first sedan, the ET7, is scheduled for delivery later this month. The company hopes to catch up: It plans to begin delivering its second sedan, the ET5, in September and to launch a sports utility vehicle (SUV), its fourth, by year-end.

Note: This article was first published on TechNode China (in Chinese).

Ever since Huawei announced its push into the Chinese electric vehicle (EV) space last year, the industry has been watching the telecom giant’s moves. 

Huawei had some modest successes in the past year, first partnering with BAIC and Changan on their self-driving technologies. It also provided the powertrain system to a little-known Chinese automaker Seres, and its SF5 model debuted last April.  

Now it looks like the tech giant has pinned its hopes on a new car model released in partnership with Seres. Last December, the two companies released Aito M5, the first EV model equipped with HarmonyOS, Huawei’s alternative to Google’s Android operating system. (Huawei developed Harmony after Washington banned Google from working with Huawei in 2019.) 

On Feb. 18, TechNode China had a chance to test drive the Aito M5 in the southwestern city of Chongqing, home of the Seres’ factory. So how did Huawei do in EV tech? Here are our takeaways. 

Seamless connectivity for existing Huawei users

Aito M5 is the first luxury EV model manufactured by Seres. The hybrid sports utility vehicle claims to reach 1,242 km (772 miles) on a single charge and tank, with a price range from RMB 249,800 to RMB 319,800 ($39,518 to $50,592). By comparison, Chinese EV maker Li Auto’s plug-in hybrid crossover Li One, the best-selling medium-to-large size SUV in China last year, features a maximum range of 1,080 km and is priced from RMB 338,000.

The in-car version of the HarmonyOS shares a similar design language with Huawei’s smartphones along with some of the most frequently-used features. For example, we could activate most of the car’s functions by voice control. The car dashboard also has a shortcut bar for fast access to the most used features.

Aito M5 came with many apps, including a navigation map app, streaming services such as Tencent-backed Ximalaya FM, and Alibaba’s Youku. You can use Youku to watch videos or relax with music or audiobooks while driving when stuck in traffic. An alert system will also notify users of significant changes in road traffic.

Huawei’s ability to integrate its ecosystem with the car differentiates Huawei from other EV players. Huawei devices, smartphones, tablets, smartwatches can seamlessly work with the vehicle. Phone calls and messages could be synced on Huawei’s devices, including the car’s dashboard. That will probably become one of the biggest competitive advantages for rival EV players. 

Fast and accurate voice assistant

Huawei also brought a powerful in-car voice assistant called Xiaoyi to the car. The assistant is powered by Huawei’s in-house cloud infrastructure. During the test drive, the assistant provided accurate responses promptly. It recognized voice commands from riders in the front passenger seat and from the rear seats, opening windows and unlocking the doors for the respective speaker, for example. Huawei said Xiaoyi can control all the features in the vehicle.

Riders can even issue multiple commands to Xiaoyi without repeating the wake word (“Xiaoyixiaoyi” in Chinese). The assistant will continue to listen for another request after it completed the previous ones.

Huawei’s virtual assistant also serves as a voice guide. For example, Xiaoyi suggested turning on the in-car air purifying function when the car drove into a tunnel and encountered bad air quality. It also searched for a charging station and navigation when the vehicle battery ran low. 

Speaking to a virtual voice assistant for those control functions within the car is well-developed in the industry. Major rivals such as Nio and Xpeng have similar offerings. Nio owners could start a conversation with a voice assistant using the three-syllable phrase “Hi, Nomi,” while Huawei’s wake word “Xiaoyixiaoyi” has four syllables. Alibaba-backed Xpeng in late 2020 said each of vehicle owners used its voice assistants effectively 25 times per day on average, compared with 13 times from part of Ford models, Chinese financial media Caixin reported.

Integrating home and auto 

The Aito M5 helps Huawei build a connection between an EV and its wide range of digital and smart home devices. That connection is taking shape as Huawei and its auto partner have introduced dashboard-based smart home management tools for users to integrate their homes into the vehicle.

Being able to sync all their Huawei devices means users can read and send text messages directly by voice command in the car, then continue listening to music and podcasts at home exactly where they left off from the in-car system. However, the integration may not work as seamlessly for non-Huawei users. 

Challenging road ahead

The Aito M5 showcases in-car technologies that Huawei offers: a dashboard that performs many of the same functions as Huawei smartphones and a network that allows remote connectivity to a plethora of its home appliances.

And yet, the Chinese telecom giant and its obscure manufacturing partner will need to build a reputation for building quality cars. The Aito M5 is entering a Chinese EV market crowded with established players, competing heads on with similarly-priced rivals, such as Tesla’s Model Y and Li Auto’s popular crossover Li One. 

Chinese electric vehicle (EV) sales achieved a strong momentum over the past two years, reporting robust figures in January. They are expected to reach 5.5 million units this year. Tesla ended 2021 with a solid profit performance driven by both strong consumer demand in China and Europe, and cost improvement from expanded production in its Shanghai factory. Battery maker CATL retained its competitive lead, dominating the global EV market last year, followed by a group of smaller domestic competitors. BYD’s chip unit is racing the clock to complete an initial public offering in the mainland stock market, thanks to explosive growth in EV sales amid a worldwide chip shortage.

January EV sales signal a strong 2022

News: China’s electric vehicle market remains buoyant despite the seasonal holiday slowdown and the looming impact of the recent subsidy reductions. January retail sales of new energy vehicles (NEVs), including all-electrics, plug-in hybrids, and hydrogen cars, totaled 347,000 units and a 132% yearly increase, according to figures published by the China Passenger Car Association (CPCA). However, this figure is a 27% decline from last December, as China auto sales in January and February tend to be affected by the Lunar New Year holiday (roughly the first two weeks of February this year) when consumers often delay purchases and automakers halt production, the industry group said.  

Insights: The market was relatively flat during the first half of January due to a last-minute push by automakers to get their cars delivered in December. Yet sales recovered fairly quickly during the last two weeks of the month, said Cui Dongshu, secretary general of the CPCA. Cui remained positive about the impact of Beijing’s 30% subsidy cut on EVs, with CPCA affirming its previous forecast of 5.5 million vehicle passenger EV sales in China this year. Although multiple automakers have raised prices for their EVs just enough to offset the subsidy cut, Cui expects overall EV prices to maintain relatively stable, as automakers have been taking various measures such as diversifying sourcing of parts to reduce costs.

News link: TechNode 

Tesla posts second profitable year as Shanghai factory reaches full capacity

News: Riding a wave of growing customer interest for green energy vehicles, Tesla on Jan. 26 posted a profit for the second year in a row. It ended 2021 with a net profit of $5.5 billion, a more than sixfold yearly increase. Annual deliveries also surged 87% in the year, marking the fastest pace of growth since 2019, thanks to strong sales in China and Europe. The US EV giant expects to achieve 50% annual growth in vehicle deliveries “over a multiyear horizon,” while warning that the ongoing global chip shortage could dent its production output “across all factories” this year.

Insights: Rising demand in China has been a key driver for Tesla’s growth. The total sales of Chinese-made vehicles reached 484,130 units last year, accounting for over half of its global deliveries, China Passenger Car Association (CPCA) data shows. The company’s Shanghai factory also plays a prominent role for its global expansion, becoming a “main export hub” with a shipment of around 163,000 vehicles last year to EU, Japan, among other regions, said Tesla’s financial chief Zachary Kirkhorn during its fourth-quarter earnings call.

Now, as EVs continue their current growth trajectory, Tesla has planned to invest RMB 1.2 billion ($188 million) to increase the production staff of the Gigafactory Shanghai by a quarter to about 19,000, Bloomberg reported in November citing sources. The Shanghai plant, which began deliveries in late 2019, was designed to produce up to 500,000 vehicles annually and has been regularly running at a capacity of 450,000 units per year.

News link: TechCrunch 

Battery giant CATL’s dominance unabated in China’s EV boom

News: CATL’s dominance of the EV battery market has continued unabated. It retained its top spot as the world’s biggest battery vendor last year, thanks to an accelerated shift of consumers embracing EVs in China. The Chinese battery giant supplied 96.7 gigawatt-hours (GWh) equivalents of EV batteries in 2021, representing a 167% yearly increase. It commands a 32.6% global market share, according to data compiled by market tracker SNE Research. South Korea’s LG Energy Solution came in second with 60.2 GWh, while Chinese auto major BYD ran a distant fourth with 26.3 GWh. Smaller Chinese players Gotion High-Tech, CALB, AESC, and SVOLT all rank lower in the world’s top 10 battery makers and form a combined market share of around 8%.

Insights: This has been the fifth year CATL retained its position as the world’s biggest battery maker, buoyed by a rebound in EV demand in its home market in 2021. A total of 150 GWh of battery capacity were deployed into newly sold NEVs in China last year. That number is expected to grow by over 50% year on year to 230 GWh in 2022, according to a Jan.12 report published by Chinese brokerage Huaan Securities.

The battery maker is also quickly expanding its manufacturing capacity to meet a surging demand. In December, it kicked off production at its largest plant to date in Fuding, a city in the eastern Fujian province, with a designed capacity of 120 GWh per year. 

News link: TechNode

BYD’s chip unit to list on Shenzhen stock market

News: The chip unit of Chinese automaker BYD is racing to go public with an offering that could raise as much as RMB 2 billion ($314.4 million), after getting a green light from the Shenzhen Stock Exchange. The listing is expected in the next few months and it would become the first auto chipmaker to list in China. BYD Semiconductor became an independent subsidiary of the Chinese EV giant in April 2020 and mainly develops less advanced chips such as microcontrollers (MCUs) used for controlling simple functions in cars. The company has become China’s biggest MCU manufacturer with nearly two decades of chip-making experience, Chinese media Caixin reported last month, citing analysis from market research firm Omdia.

Insights: The imminent listing comes at a time when the Chinese EV industry has seen a strong rebound in demand, despite significant disruption due to the global chip shortage over the past year. BYD Semiconductor estimated its net profit will jump by up to 574% yearly to RMB 395 million in 2021. Revenues are projected to reach an upper limit of RMB 3.2 billion, an 122% increase from 2020. However, the company is still a tiny player in the global automotive MCU sector, which is dominated by Japan’s Renasas and six other chip powerhouses with a combined market share of 98%, according to figures from information services company IHS Markit.

And yet, investors have high expectations for the subsidiary. It has already raised RMB 2.8 billion from a list of big names including Xiaomi’s industry investment fund, Sequoia Capital China, and CICC Capital prior to the IPO filing. BYD’s stake will fall from 72% to 65% after the listing is completed.

News link: TechNode 

Trunk Tech, a Chinese autonomous truck technology startup backed by EV maker Nio, raised an undisclosed amount in its Series B, the company announced Wednesday. The round was led by state-owned automaker BAIC, which is also partnering with ride-hailing giant Didi to get a fleet of self-driving taxis on public roads by 2025.

Why it matters: Trunk Tech is one of the most promising startups in the Chinese self-driving car space. The Beijing-based company has been backed by a list of prominent investors, and is among several players to test autonomous vehicle systems for hauling freight at domestic harbors.

Details: New investors in this latest fundraising round include private equity firm Pre-IPO Capital Ltd and Zhengzhou municipal investment fund, according to a statement published Wednesday (in Chinese).

• Trunk Tech, which was founded in 2017, said it has delivered more than 100 driverless trucks to clients including China National Offshore Oil Corporation in trial projects at domestic ports in Tianjin and Ningbo.
• The startup said it is operating a testing fleet made of dozens of autonomous trucks, which haul cargo for partners such as JD Logistics, Deppon Logistics, and Alibaba-backed STO Express, with its fleet having now driven 1.5 million kilometers (932,057 miles).

Context: Chinese automobile and tech companies have been racing to develop and commercialize their own self-driving tech which they claim will increase road safety and improve fuel efficiency for traditional trucks.

• TuSimple, which was founded in 2015, was the world’s first self-driving truck company to go public on the US stock market, in April 2021. Its fleet had logged more than 6.3 million miles on public roads as of Dec. 31, 2021, according to its fourth-quarter earnings report. The company plans to start mass producing driverless trucks for clients in 2024.
• Inceptio, a robotruck startup backed by tech giants JD and Meituan, unveiled plans early last year to start mass producing autonomous trucks with automakers Dongfeng and Sinotruk as early as the end of 2021. The company reaffirmed the goal in September, adding its vehicles had passed summer endurance tests.
• Fabu, a company founded by former Didi executive He Xiaofei, nabbed RMB 100 million ($15.4 million) in a Series B+ in August, after testing automated trucks in Zhejiang province’s Ningbo-Zhoushan Port for more than a year.
• Trunk Tech raised an undisclosed Series A from investors including NIO Capital and logistics giant GLP’s investment unit Hidden Hill Capital, in October 2018. Other existing investors include German auto supplier Bosch and artificial intelligence company iFlytek.

Baidu’s electric vehicle (EV) project Jidu Auto announced on Wednesday that it has raised nearly $400 million in Series A as the Chinese search engine giant accelerates the development of EVs with self-driving capabilities.

Why it matters: Jidu will use the proceeds on research and development as the company aims to unveil a concept car in April later this year and release its first production model in 2023, according to the announcement.

Details: Baidu and its manufacturing partner Geely both raised their stakes in Jidu by jointly investing almost $400 million in the venture. The two companies didn’t reveal the sharing ratio. 

Context: Baidu and Geely linked up last January with a deal that would allow the tech giant to make its own consumer EVs with autonomous driving capabilities.

• The result was the establishment of the RMB 2 billion ($316 million) joint venture Jidu Auto two months later, with Baidu and Geely holding 55% and 45% of the total shares, respectively, according to business research platform Tianyancha (in Chinese).
• Baidu has operated an autonomous driving unit testing its vehicles in China and the US since 2015. It began commercial autonomous ride-hailing services for passengers in Beijing in November last year, and plans to expand the service to 65 domestic cities by 2025.  
• Early last year, Geely also launched its own premium EV brand Zeekr, which raised $500 million a few months later from a group of investors, including China’s biggest battery maker CATL and streaming giant Bilibili.
• Geely reportedly plans to acquire Alibaba-backed smartphone maker Meizu as part of its long-term electric mobility vision.

Chinese automaker BYD said on Wednesday it is partnering with US autonomous driving startup Nuro to make electric robocars for goods delivery services.  

Why it matters: The partnership is the latest example of Chinese automakers working with overseas tech companies to build autonomous vehicles.

Details: BYD is currently working with Nuro to design and develop the latter’s next-generation autonomous delivery robots, which will be equipped with components provided by the automaker such as electric motors and lithium-iron-phosphate blade batteries, according to a Thursday announcement. 

• Softbank-backed Nuro plans to begin mass production next year of the low-speed vehicles, which come with two spacious cargo areas and an external airbag for pedestrian safety, as it scales up production at its currently under construction manufacturing facility in Nevada, according to TechCrunch. 
• The two companies expect the collaboration to help scale Nuro’s last-mile delivery services to “millions of people” in the US.

Context: Nuro was co-founded in 2016 by Zhu Jiajun and Dave Ferguson, two former engineers at Google’s self-driving car project. The company announced in December 2020 that it had received first-of-its-kind approval by US regulators to operate and charge for its driverless delivery services, TechCrunch reported.

• Warren Buffet-backed BYD formed an alliance with Japanese automaker Toyota in mid-2019 to jointly develop electric vehicles, which they expected to hit the Chinese market under the Toyota brand name by 2025.
• Another Chinese auto major Geely said earlier this month that its premium EV brand Zeekr is working with Intel’s autonomous driving unit Mobileye to develop electric and autonomous passenger vehicles, aiming for release as early as 2024 in China, according to TechCrunch.

China’s electric vehicle (EV) sales soared in 2021, bucking the national trend of slowing auto sales. Local automakers have shown strong competitiveness against overseas counterparts. However, industry players may face new challenges: a looming price war among competitors will likely reduce profits, a UBS Securities analyst said on Tuesday.

Why it matters: There might be greater supply than demand in the Chinese EV market this year, since consumption could be reduced by slowing economic growth amid the recharged pandemic, Paul Gong, head of China auto research at UBS, told reporters on Tuesday.

• An easing chip shortage may also help EV makers return to normal auto production this year, Gong said. He warned that an intense “price war” would push the prospect of profitability further away for automakers in the short term.

Details: Still, the rise of domestic EV makers will be “the way of the future” in China, as local players have generally “achieved greater progress” in the development of products and technology than foreign auto majors, according to Gong (our translation).

• UBS projects cautious optimism in its outlook for the industry over the long term. It expects that, compared to its 2021 projection of 3 million vehicles, China’s EV sales will increase by 35% to more than 4 million vehicles this year. Sales could grow to 7.05 million units in 2025, according to UBS. 
• Sales of new energy vehicles, which include all-electrics and plug-in hybrids, increased by nearly 160% year-over-year to 3.52 million units in 2021, according to a statement published by China’s Ministry of Industry and Information Technology on Wednesday.

Read more: Drive I/O | Auto China 2021: A banner year for Nio, Xpeng, and Li Auto

Context: The number of passenger electric vehicles sold in China surged 169% year on year to nearly 2.99 million units in 2021, according to figures published Tuesday by the China Passenger Car Association (CPCA). That figure beat the estimated 2.4 million units the industry group made in June.

• Tesla China sold a record 70,847 locally-made vehicles in December and saw its total 2021 sales reach 320,743, taking the third spot in the list of China’s top-selling EV makers. BYD dominated the market with sales of 584,020 vehicles, followed by SAIC-GM-Wuling with 431,130 cars, CPCA figures showed (in Chinese).
• US-listed Chinese EV trio Nio, Xpeng, and Li Auto are among the top 10 sellers, each achieving deliveries of nearly 100,000 vehicles. German auto giant Volkswagen sold around 130,000 passenger EVs, more than doubling its 2020 total, according to CPCA.
• CPCA raised its forecasts for China’s NEV sales, including passenger and commercial vehicles, by over 10% to 6 million units in 2022 from the previous year. The association added that China will maintain leadership in the global EV race.

Zvision Technologies, a Chinese startup that makes lidar sensors for self-driving cars, announced a new investment from three Chinese automakers on Monday, including Xpeng Motors. The company becomes the latest startup to tap growing investor interest in the self-driving car space.

Why it matters: The investment is another sign of the increasing interest in lidar sensors, seen as a crucial building block for future vehicles by most auto and tech firms. Lidar is a key component for self-driving cars and uses laser light to sense surroundings. 

Details: Zvision has raised “hundreds of millions of yuan” in a pre-Series C led by Xpeng Motors, according to a Monday announcement (in Chinese). Shang Qi Capital, a private equity firm owned by Chinese automaker SAIC, participated in the round. 

• State-owned automaker Dongfeng Motor and existing backer Intel Capital also joined the round. A company spokeswoman declined to disclose an exact valuation when contacted by TechNode on Tuesday.
• Zvision plans to use the funds to accelerate the development and mass-production of its automotive-grade lidar sensors, including improving its production line and supply chain. The company has yet to show a timeline. 

Context: In September, Xpeng had begun delivering the world’s first Lidar-equipped production vehicle, the P5, which the company boasts can distinguish objects within a range of up to 150 meters and can run autonomously under a driver’s supervision on Chinese roads, the South China Morning Post reported. 

• Xpeng sources lidar sensors for the P5 from Livox, an affiliate of Chinese drone maker DJI, which previously encountered technical issues when attempting to meet the reliability requirements for automobiles, TechNode reported last June.
• In November, the EV maker unveiled its second sports utility vehicle model, the G9. It will come fitted with two lidar sensors provided by Robosense, a lidar startup backed by SAIC and BYD. The G9 is scheduled for delivery in the third quarter of 2022.
• Xpeng chief executive He Xiaopeng told investors last May that the company was testing lidar technology from multiple suppliers.
• Rival EV maker Nio has heavily backed Innovusion, a startup formed by two Baidu veterans in 2016, and used the firm’s lidar technology for its first sedan model ET7 as well as the upcoming ET5.

Huawei burrowed further into the auto industry with the launch of the first vehicle with its homegrown operating system. The Chinese government cut purchase subsidies on new energy vehicles (NEVs) by 30% this year, while scrapping ownership limits on foreign automakers’ investments in the auto industry. Chinese electric vehicle (EV) makers Nio, Xpeng, and Li Auto celebrated record annual deliveries of nearly 100,000 cars in 2021. Alibaba’s head of autonomous driving lab quit the company after more than four years. Didi, soon to delist, shows a few signs of approaching break-even with its first post-IPO earnings report.

Huawei intensifies auto plans with launch of first vehicle with ‘seamless’ Harmony

News: Huawei on Dec. 23 unveiled the first EV model equipped with its HarmonyOS operating system with manufacturing partner Seres. Huawei boasts that this in-car software system offers users a seamless experience of smartphone and car features across devices. Priced from RMB 250,000 ($39,063), the Aito M5 sports utility vehicle runs on electricity or fuel and has a 1,242-km driving range, which compares with the 1,080 km offered by Li Auto’s popular plug-in hybrid crossover Li One. Huawei said that it will showcase the vehicle in 180 Huawei shops across 42 cities and deliveries should start around Feb. 20.

Insights: As US chip sanctions crippled its smartphone core business, Huawei is trying to diversify its operations by breaking into the Chinese automobile sector. The Chinese telecommunications giant last April started selling Seres vehicles through its sales network, but they did not sell well. From April through November, Seres achieved sales of only 7,080 SF5 EVs, which were equipped with Huawei powertrain system and in-car software, according to figures published by China Passenger Car Association. Huawei has also partnered with state-owned automakers BAIC and Changan to equip vehicles with its autonomous driving hardware and software. Yet some industry insiders are doubtful that the tech giant will eventually make its own cars.

News link: TechNode 

Beijing sticks to plan to end EV subsidies in 2023

News: Chinese authorities on Dec. 31 unveiled long-awaited details about its national subsidy program for new energy vehicles (NEVs), such as all-electrics and plug-in hybrids. For 2022, beginning Jan. 1, subsidies to EV buyers will be cut 30% compared to 2021. According to a document released by the Ministry of Finance, the grants for EVs delivering driving ranges of at least 400 km (248 miles) will be cut by RMB 5,400 on an annual basis to RMB 18,000 ($2,824). Meanwhile, the subsidies this year for all-electrics with a driving range of 300 km to 400 km will be lowered to RMB 13,000, while those for plug-in hybrids will be cut to RMB 6,800. Beijing also reaffirmed its plan to eliminate subsidies entirely at the end of this year. Subsidies for purchases of new energy vehicles (NEVs) were already trimmed by 10% and 20% during 2020 and 2021, respectively. 

Context: In reaction, several overseas automakers have raised prices for their EVs in China to offset the subsidy cuts. The prices of Tesla’s popular China-made Model 3 and Volkswagen’s ID series EVs have risen by RMB 10,000 and RMB 5,400, respectively. Newer local EV makers are taking a more active approach to reduce the impact of the subsidy cut. Nio on Jan. 1 announced moves to make up the difference between sticker prices and reduced subsidies of its vehicles for customers who had paid a deposit before the end of 2021 and who will get their vehicles delivered by Mar. 31. Cui Dongshu, secretary general of China Passenger Car Association (CPCA), forecasts that the trimmed government incentive program could still give a great boost to the EV adoption in the country, noting that the manufacturing cost of EVs and batteries are falling significantly. Cui estimated China’s NEV sales could more than double to around 6 million vehicles in 2022 from the previous year and therefore maintain leadership in the world EV race.

News link: Reuters 

China lifts restrictions on foreign auto ownership

News: China now allows overseas automakers to operate wholly-owned ventures in the country’s passenger vehicle sector. As of Jan. 1, 2022, foreign firms are no longer limited to 50% ownership in their joint venture auto operations. The law had been in effect since 1994. In addition, foreign automakers can now set up more than two joint ventures that make the same type of vehicles.  The new ownership rules were detailed in a Dec. 27 release from the Ministry of Commerce and the National Development and Reform Commission, China’s top economic planner.

Insights: The move has been perceived as a positive signal that would create a level playing field for domestic and foreign carmakers, Cui Dongshu, secretary-general of the China Passenger Car Association, told state broadcaster CGTN. Nonetheless, Cui said there would be no significant impact on the market from removing the limits since they were expected. German auto major BMW is expected to become the first internal-combustion vehicle maker to take advantage of the new JV rules. It plans to up its stake to 75% from 25% in its JV with Chinese partner Brilliance Automotive by the end of 2022. The Chinese government since 2018 has gradually ramped up efforts to fully liberalize the domestic auto industry, starting by scrapping limits on foreign ownership of EV makers as it aims to be a global leader in the sector. Tesla became the first foreign auto brand to enjoy the relaxed EV regulations when it set up its wholly-owned venture in Shanghai in May  2018.  

News link: Global Times 

China’s EV trio post record deliveries numbers in 2021

News: The US-listed Chinese EV trio of Li Auto, Nio, and Xpeng launched the new year by publishing record delivery numbers for 2021. Each noted that they had delivered nearly 100,000 vehicles in 2021, despite global chip shortages. All had doubled their deliveries from 2020. Xpeng Motors had stood out among its peers, delivering a record 98,155 vehicles last year, up 263% from its 2020 delivery count. It surpassed Nio, whose annual deliveries totaled 91,429 electric crossovers. Nio was hit by supply chain issues and changes to its manufacturing lines during the second half of last year. Meanwhile, Li Auto saw 2021 deliveries surge 178% year on year to 90,491 vehicles.

Context: Chinese automakers have been riding the wave of growing popularity of EVs in the country, boosted by a years-long national subsidy program and special license plates to EV buyers, among other policy measures. Nio, Xpeng, and Li Auto, all once struggling to stay afloat and beset by lackluster sales, are the poster children of the revolution. The trio has laid out ambitious plans to expand their sales and service networks as they vie to grab market share from internal-combustion vehicle segments. Analysts surveyed by Seeking Alpha expected Nio’s annual revenue to increase by 74% this year, Forbes reported, while Citigroup forecast that Xpeng’s deliveries could almost double to 175,000 units in 2022.

News link: South China Morning Post 

Alibaba’s head of autonomous driving quits

News: Alibaba has parted ways with Wang Gang, a renowned computer scientist who has served as head of the tech giant’s autonomous driving lab under its Damo Academy research division for three years, Chinese media reported on Jan. 5, citing people familiar with the matter. A former tenured professor at Nanyang Technological University, Wang joined Alibaba in early 2017 as the chief scientist for the company’s artificial intelligence lab and was tasked with improving speech recognition capabilities for its first smart speaker device, the AliGenie X1, launched later that year. Wang has begun working on a startup developing robot vacuum cleaners and has raised an unknown amount of funds, the sources added.

Insights: The move is noteworthy in many ways. For one, Chinese industry giants had hoovered up research talents and poured resources into exploring the potential of artificial intelligence (AI) over recent years. The rush is over given a slower-than-expected process of implementing AI in industries, as many top scientists give up the high salaries in the industry for academia, while others start up their own businesses. Wang’s departure comes after Li Lei, the director of ByteDance’s AI Lab, left the company to join the University of California Santa Barbara as a professor last August, following the resignation of ByteDance Vice President Ma Wei-Ying a year earlier, SCMP reported. Chinese tech powerhouses also struggle with executive turnover and layoffs, as Beijing’s regulatory clampdowns continue to weigh on the sector.

News link: TechNode 

Didi’s first earnings report after IPO: $4.7 billion loss

News: On Dec. 30, Didi reported its first earnings as a public company. It wasn’t pretty: The company lost RMB 30.4 billion ($4.7 billion) on RMB 42.7 billion ($6.6 billion) in revenue during the September quarter of 2021. To compare, the company reported a profit of RMB 665 million on revenue of RMB 43.4 billion in the same quarter of 2020. Didi’s largest source of revenue is still its domestic ride-hailing business, which yielded RMB 39 billion, down 12.9% from the previous quarter. The company posted an 8% quarter-over-quarter decline to 2.36 billion in ride volume over the period.

Context: Still the largest ride-hailing service in China by ride volume and revenue, Didi has been at the forefront in Beijing’s wide crackdown on local tech companies. Did’s business has taken a hit from a suspension order that has kept its services off Chinese app stores since July. Having been listed in the US for less than six months, the Chinese mobility giant on Dec. 3 announced plans to take its shares off the New York stock market and instead pursue a listing in Hong Kong. Beijing has yet to announce the results of its cybersecurity investigation into Didi, and the company’s shares have fallen more than 60% from its IPO price.

News link: TechNode 

Just a year ago, Nio, Xpeng, and Li Auto faced a cloudy future. All three had burned through hundreds of millions of investors’ dollars and were beset by lackluster sales. Most observers thought they had yet to hit bottom. Not anymore.

Despite the lingering impact of the pandemic on China’s automotive industry, 2021 has been a fantastic year for Tesla’s major Chinese challengers. The three companies all reached their 100,000-vehicle production milestones, racked up big war chests from new investors, and recently set records for their vehicle deliveries. Their cars are going mainstream in major domestic cities, according to Xpeng President Brian Gu, as internal-combustion vehicles and legacy automakers are increasingly being regarded as outdated.

The Chinese trio, all listed in the US, not yet profitable, but all poised for stronger growth in the coming year, have become the poster children for the country’s EV revolution. Despite a 20% cut in subsidies this year, the world’s biggest EV market in September witnessed an unexpected growth rebound, as the NEV (new energy vehicle) penetration rate surpassed 20% of all new car sales for the first time. 

We round up the most significant milestones in the three companies’ turbulent histories this year and forecast what’s next for them in the coming year.

Nio — Mounting a comeback

With deliveries beating those of BMW and Audi EV at a price tag comparable to those of German auto giants, Nio is literally the first Chinese automaker to have gained a foothold in the country’s premium vehicle segment. Formerly referred to by Deutsche Bank analysts as number one among the promising local EV makers, Nio was overtaken by its peers, as measured by deliveries, due to its relatively slower pace of growth this year.

Once maintaining a leadership position in the non-Tesla piece of the Chinese premium EV segment, Nio found itself in a bittersweet position over the past few months as rivals’ sales grew at a stunning speed. Li Auto and Xpeng in July recorded deliveries of 8,589 and 8,040 vehicles, respectively. Those numbers surpassed Nio’s monthly output for the first time ever. Nio produced only 7,931 for the month.

Then Nio’s monthly deliveries decreased to an even lower level of 3,667 vehicles in October. That number was less than half of both Li Auto’s and Xpeng’s for the month. The company blamed the drop on the restructuring of manufacturing lines in preparation for introducing new models. The most recent sales figure of 10,878 vehicles in November marked a strong rebound for Nio, despite an ongoing industry-wide chip shortage. Moreover, that figure lagged behind those of the other two US-listed EV makers by several thousand units.

More notably, Nio faced one of its worst public relations crises in China in August, when a 31-year-old driver was killed using Nio’s driver-assistance feature with his ES8 electric crossover. The incident not only put further dents into an already tough outlook for the regulatory environment and public confidence in China’s autonomous vehicle space: It also stoked criticism of Nio for overstating the capability of its technology and fragmented its once incredibly loyal fanbase. Details about the accident still have not been released.

Nonetheless, the Tencent-backed EV maker is ramping up efforts to regain its leading position in the market. It’s currently on track to deliver its first premium sedan model ET7, equipped with a Lidar sensor and Nvidia’s supercomputer, in March 2022. It also just launched a lower-priced new sedan model, ET5, as it aims to lift its sales in the country. At the same time, it is rushing to launch a mass-market EV sub-brand next year, targeting the most competitive and yet the biggest segment in China’s auto market.

Xpeng — Taking first place

Once chugging away in Nio’s tracks , Xpeng has raced ahead as China shifts from gasoline power to electric transportation. It is emerging as the new leader in the competitive mid- to high-end Chinese auto segments. The Alibaba-backed EV maker delivered a record-breaking 15,613 electric vehicles in November, bringing its annual deliveries to more than 82,155 vehicles. That figure surpassed Nio’s 80,940 deliveries in the year to date.

Xpeng’s strong performance comes at a time when the country has seen a major rebound in EV demand, signaling a tipping point for mass adoption. Sales of NEVs, comprising all-electrics and plug-in hybrids, are expected to more than double to 3.4 million units annually this year and could further increase by 47% to 5 million units in 2022, according to estimates made by the China Association of Automobile Manufacturers (CAAM) earlier this month.

To ride the wave of the EV recovery momentum, Xpeng has aggressively expanded its product lineup with the release of a premium sports utility vehicle (SUV) model and an affordable family sedan. The company boasts that both will offer the most advanced automated driving capabilities in China.

G9, Xpeng’s first luxury electric crossover, will be equipped with an 800V supercharging platform, which could boost driving range to 200 kilometers (124 miles) with only a five-minute charge. It also has advanced driver assistance software that will allow vehicles to cruise autonomously in gnarly urban traffic conditions. Aiming for a price range between RMB 300,000 and 400,000 ($47,100 and $62,800) according to Jefferies analysts, Xpeng’s G9 model is scheduled for delivery in the third quarter of 2022. It will then compete head-to-head against Tesla’s Model Y and Nio’s ES6, among other top-line EVs.

Meanwhile, Xpeng’s second sedan model, P5, is expected to be a hit. It is equipped with two Lidar sensors, offering urban automated driving capabilities, and is priced competitively, beginning at just RMB 157,900. With P5 deliveries started in October, President Brian Gu expects the company to continue to experience rapid growth in the coming months. Gu projected a monthly delivery target of 15,000 vehicles for the last two months of 2021 during an earnings call last month.

Analysts are bullish on Xpeng’s growth prospects, expecting its monthly sales momentum of 15,000 vehicles will continue in 2022, Chinese media reported in late November, citing Daiwa Securities Meanwhile, Citigroup analysts forecast that Xpeng’s deliveries could nearly double to 175,000 units in 2022.

Li Auto — Radically readjusting

Considered by many as taking a conservative yet non-mainstream approach in betting on the transitional extended-range technology, Li Auto also had a vintage year in 2021. In the year to date, the company has delivered nearly 80,000 Li One electric crossovers, its first and the only model currently on sale. That number is almost as much as the combined deliveries of Nio’s three SUV models.

Backed by Chinese food delivery giant Meituan, Li Auto pursues greater operational efficiencies than its peers. The strategy paid off, with the automaker reporting an impressive gross margin of 23.3% in the third quarter of this year, compared with Nio’s 20.3% and Xpeng’s 14.4%.

Also, each of Li Auto’s stores makes more money on average than those of Nio and Xpeng. The company in November sold nearly 80 vehicles per showroom, more than double Nio’s figure for the same period. Li Auto planned to expand its sales network to 200 stores by this year’s end. In contrast, both Nio and Xpeng said they will each operate more than 350 outlets by that time.

However, Li Auto’s competitiveness in self-driving technologies has lagged far behind rivals’. For example, earlier this month, it shipped an over-the-air update that includes an automated parking feature—the same feature Xpeng offered its customers three years ago. The company’s vehicles are also unable to cruise Chinese highways on their own while being supervised by active drivers. That assisted driving function, similar to Tesla’s Navigate on Autopilot, is already available to Nio and Xpeng customers.

To catch up with rivals and prolong its upward trajectory, Li Auto will shift its strategies radically in the coming years. Executives said the company would more than triple the annual research and development budget to RMB 3 billion ($500 million) this year. That number will be further increased to RMB 6 billion per year by 2024, financial chief Li Tie pledged to investors during an earnings call in February.

And yet, the EV maker claims that it will maintain a healthy gross profit rate while working on a significant expansion of its product lineup and production footprint over the long term. CEO Shen Yanan last month reaffirmed the plan to launch a full-size extended-range electric SUV next year, followed by the release of its first fully electric vehicle model in 2023. Its second manufacturing plant launched construction in Beijing in October. When production begins there in 2023, Li Auto hopes to double its total annual capacity to 200,000 vehicles.

Nio on Saturday revealed its second fully electric premium sedan model ET5, featuring an automated driving system, a fresh design, and a lower price point, to reach a larger customer base.

Why it matters: Speaking to reporters on Dec. 19, chief executive William Li said he expects the Nio ET5, which is priced 25% cheaper than the brand’s first sedan model ET7, will help the company attract more younger and female buyers.

Details: The new ET5 sports sedan comes with the same hardware package as the ET7, including a dozen ultrasonic sensors, 11 cameras, a Lidar unit, and Nvidia’s Orin autonomous driving processors, which allow the vehicle to detect its surroundings using supercomputing power.

• The new sedan comes in a rainbow of nine different hues, including a soft pink, the first time Nio has used such a shade for its models.
• It also features an in-car information and entertainment system equipped with custom augmented reality (AR) glasses that can project information onto the windshield for drivers.
• Priced from RMB 328,000 ($51,430), the ET5 sedan is scheduled for delivery in the third quarter of 2022. The price could be as low as RMB 258,000 if a customer pays a monthly rental fee of RMB 980 for the use of batteries.
• Nio also announced Saturday that it will begin delivering the ET7 on Mar. 28, 2022. ET7 and ET5 can achieve more than 1,000 kilometers (621 miles) on a single full charge with a new 150kWh battery pack.

Context: With three existing models, the seven-year-old Nio had so far delivered 80,940 vehicles to customers this year, a 120% yearly growth rate. Nio’s peers Xpeng Motors and Li Auto delivered 82,155 and 76,404 vehicles respectively during the same period.

• With a similar price range, Nio’s ET5 will compete against Tesla’s popular Model 3, Xpeng’s P7, and Geely’s premium EV model Zeekr 001.

As tech and auto giants continue to toil on their autonomous vehicle projects, according to one Geely executive, such technology will come to flying cars more quickly than road vehicles.

Despite regulatory barriers and ethical issues, self-flying cars are likely to be ready before autonomous road vehicles, Guo Liang, chief executive of Geely-owned Aerofugia said at the Beyond Expo event held in Macau on Dec. 3. The movements of unmanned aerial vehicles are supervised by aviation authorities, which makes them more coordinated and safer than ground vehicles, according to Guo.

The idea of building an on-demand ride-hailing service for the skies is gradually getting off the ground in China, as eVTOLs, or electric vertical takeoff and landing vehicles, are transitioning from a pie-in-the-sky concept to a maturing technology and promising investment opportunity.

Geely is among a number of companies – from auto majors to venture-backed startups – racing to grab a foothold in this nascent market. In September 2020, the company took a big step into aviation by forming Aerofugia through a merger with Chinese drone developer AOSSCI.

Prior to that, China’s biggest private automaker had invested in German flying taxi startup Volocopter in September 2019. Volocopter in April said it had formed a joint venture with Geely’s Aerofugia, with a plan to bring 150 eVTOL aircraft to China as early as 2024, Reuters reported.

Electric vehicle maker Xpeng Motors also aims to stretch itself beyond just selling cars, in October leading a $500 million Series A investment round in aviation startup HT Aero at a pre-money valuation of $1 billion. Morgan Stanley estimated the global urban air mobility (UAM) market for flying cars and taxis could be worth $1.5 trillion by 2040.

Unlike conventional helicopters or airplanes, electric-powered vertical-lift machines can take off and land without a runway while delivering low-carbon air travel without producing emissions, providing commuters with a new way to speed above crowded streets.

Guo called for a joint effort from industry players to scale up urban air mobility operations with regards to technical standards and route management, among other industry-wide aspects.

A clear regulatory framework and established infrastructure on the ground also need to be in place for the vehicles, noted Jiang Yutao, vice president of Chinese flying taxi startup EHang. Also speaking as part of Beyond, Jiang added that eVTOLs could help the country meet its climate goals.

Nasdaq-listed Ehang expects to obtain approval from the Civil Aviation Administration of China for its autonomous aircraft EH216 “in the next few months,” Hu Huazhi, chief executive told investors during its third-quarter earnings call last week.

Such moves make Guo’s prediction more grounded than it may first appear, with a number of companies seemingly believing that automated flying cars are very much on the horizon in the next decade.

Tesla and General Motors Wuling are the two undisputed leaders of the pack in China’s $49 billion electric vehicle (EV) market, together holding nearly a 20% share this year. But more than a dozen legacy and infant automakers are in hot pursuit. All emerging from rough patches, three US-listed domestic makers—Nio, Xpeng, and Li Auto—now comprise the second tier of contenders. Riding on high-growth trajectories, the trio are tipped to be Tesla’s most formidable domestic challengers.

Yet with a flood of new money supercharging the industry, third-tier EV makers are coming on as powerful forces as well. Reporting deliveries in significant numbers and backed by a growing list of reputable investors, several pose a real-time threat to the US-listed trio, and speculation is building that some are preparing for listings in Hong Kong.

The third-tier upstarts have been buoyed by strong growth in domestic electric passenger car sales this year. Sales of 321,000 EV units in the first ten months of 2021 represented a 141% year-on-year increase from the same period in 2020, when the overall auto sales slumped 14% from the year before, data from the China Passenger Car Association shows.

Here is our roundup of the four most competitive upstarts emerging in China’s EV space.

WM Motor – First mover making a comeback

Along with Nio, Xpeng, and Li Auto, WM Motors was once one of Deutsche Bank analysts’ “Fab Four” of likely candidates to grab the non-Tesla piece of China’s EV market. 
 
Founded in 2015 by Freeman Shen, a former top Volvo executive, WM Motor was one of the earliest EV startups to deliver production vehicles to Chinese customers, reporting a quite respectable delivery number of around 22,000 cars back in 2019. That was a few thousand more than Nio’s numbers that year, and far eclipsing Xpeng’s, which delivered just over 5,000 vehicles. Trailing far behind, Li Auto churned out its first model Li One later that year.
 
While WM Motor took an early lead in entering initial production, it was quickly overtaken as its sales growth remained virtually flat. Meanwhile, rival Xpeng jacked up deliveries almost five-fold in 2020. Now WM Motor’s delivery numbers of 34,068 vehicles for the first ten months of this year are only half of those of Nio’s and Xpeng’s.

How did WM Motor lose its first-mover advantage in a fast-growing market? There is a consensus that the automaker presents itself as a rather faceless brand (in Chinese): Its cars are functional but middle-of-the-road. Meanwhile, peer Nio is increasingly perceived by customers as a high-quality premium brand with top-of-the-range services. WM Motor has also lagged behind Xpeng in the autonomous vehicle space. Then in late 2020, it was plagued by a recall affecting over 1,000 of its vehicles following several reports of fires within a single month in late 2020.
 
Nonetheless, many venture capitalists are still anticipating great things for WM Motor. The Baidu-backed EV maker in October said that it was near wrapping up its $500 million Series D funding round led by PCCW, a Hong Kong telecom company owned by the family of local business magnate Li Ka-shing. WM Motor is aiming to launch its fourth production model and first sedan, the M7, by next year. The model will face off against the likes of Tesla’s Model 3, Nio’s ET7, and Xpeng’s P5.

Hozon – Closing in on Xpeng and Li Auto

Surpassing Nio and Li Auto in monthly vehicle deliveries for the first time in October, the lesser known Hozon may soon be a rising force to be reckoned with in the Chinese EV market.
 
With three affordable entry-level cars in its portfolio, the Zhejiang-based automaker handed over 8,107 vehicles to Chinese customers in October, marking a stunning growth of 294% compared to its deliveries in October 2020. Deliveries for the first ten months of this year totaled nearly 50,000 vehicles, closing in on the numbers of Xpeng and Li Auto. Each delivered more than 60,000 units during the same period.
 
A strong sales recovery in China’s EV market as a whole is a key factor fostering the rise of the likes of Hozon, said Cui Dongshu, secretary general of the China Passenger Car Association, during an online conference last month. China witnessed strong growth in electric passenger car sales, recording a 141% year-on-year increase in October to 321,000 units, when the overall auto sales slumped 14% from a year earlier, data from the industry body showed.
 
A wave of local but big state companies have noted the uptick in EV sales this year and are rushing to back growing EV startups. The government of Yichun city in central Jiangxi province is Hozon’s largest shareholder, taking a 51.31% stake in the company, The Economic Observer reported (in Chinese). And Hozon in October said it closed an RMB 4 billion ($626 million) Series D1 led by Qihoo 360, representing a major endorsement by China’s biggest cybersecurity firm.
 
This was followed by an undisclosed amount of investment by CATL, the first publicly known investment in a young EV maker by the battery giant, Yicai reported in November. (CATL also has invested in Zeekr, a premium EV subsidiary of Geely.) Eyeing a capital raise of $1 billion from an initial public offering in Hong Kong next year, Hozon aims to achieve annual sales of 70,000 vehicles this year and increase that number more than sevenfold to 500,000 in five years.

Leapmotor – Budget buyers’ choice

China’s fast-growing EV market has drawn an array of unusual competitors from television makers to real estate firms. Among them is Dahua, China’s second-biggest surveillance equipment maker. Formed in 2015 by Zhu Jiangming, Dahua’s co-founder and former technology chief, Leapmotor is the newest Chinese EV unicorn, having raised over RMB 11.5 billion ($1.8 billion) amid the flood of new money pouring into China’s EV space.
 
In its most recent funding round, announced in July, the company raised RMB 4.5 billion from heavyweights including state-backed CICC Capital and investment entities led by the municipal government of the eastern city of Hangzhou, where its parent Dahua is headquartered. This was quickly followed with a plan to build a new assembly plant with a production capacity of 200,000 cars annually in Hangzhou, Chinese media reported. The plant is scheduled for completion in 2023.
 
As with Hozon, the current three Leapmotor models are all budget-minded mainstream vehicles, priced between RMB 60,000 and RMB 200,000 ($9,390 to $31,300). And yet, Leapmotor’s budget mini-electric car, T03, has really gained traction in the market. With a starting price less than $10,000, the four-seater mini-electric car claims a range of 403 kilometers (250 miles) on a single charge and offers assisted driving functions such as lane departure warning and automatic emergency braking.
 
With T03 accounting for over 90% of the company’s deliveries this year, Leapmotor has declared a wildly ambitious annual target of more than 800,000 deliveries by 2025. That would account for nearly 60% of all EV sales in the country, according to the China Association of Automobile Manufacturers (CAAM). The Hangzhou-based EV maker is reportedly weighing a Hong Kong listing of more than $1 billion as soon as next year.

Human Horizon – Crazy rich SUVs

Born in late 2017, Shanghai-based Human Horizons is unique among a large pool of EV startups in China: It has never raised any outside investor money. That’s in sharp contrast to the likes of Nio and Xpeng which used to struggle to secure funding for their cash-burning businesses.

Founder Ding Lei also has an unusual background. Ding started his career as a quality engineer for the joint venture set up by Volkswagen and SAIC in Shanghai in 1988, then became a vice president of the state-owned automaker in 2007. Yet his most notable experience occurred in 2013, when he became a deputy head of the city’s Pudong New Area for a two-year period. In 2017, he founded both Human Horizon and an investment firm called East Coast Capital.

The company’s premium EV brand Hiphi attracted many eyeballs by releasing what is believed to be the most expensive made-in-China EV model ever: Hiphi X. The limited edition electric sports utility vehicle costs RMB 800,000 (around $125,000). With a driving range of 550 km (342 miles) on a single charge, the luxury vehicle boasts a stand-out performance and an opulent interior to “a degree at which the [Tesla] Model X looks quite conventional,” as one reviewer put it. 

Human Horizon’s efforts with the Hiphi X were successful. In September it delivered 641 Hiphi X units, becoming the first locally-made car to top sales in China’s premium EV segment, defined as autos priced above RMB 500,000 ($78,450). Its sales beat both Porsche’s electric supercar, Taycan, and Audi’s sports sedan, E-tron, according to CPCA figures. The company last month announced it would launch a second premium SUV model, the GT-Hiphi Z, next April and start delivery within the year. No price details have been released.

China’s tech giant Baidu officially launched an autonomous ride-share service in the capital city Beijing, after receiving the country’s first permit for commercial robotaxi services.

Why it matters: This is the first time that the Chinese government has allowed companies to legally charge Uber-like fees to the public for their robotaxi services, a major milestone for Chinese self-driving car development. 

Details: The service, known as Apollo Go (or Luobo Kuaipao in Chinese), ferries passengers around a 60 square kilometer (around 23 square miles) area in the Beijing Economic and Technological Development Zone in the south of the city, Baidu said on Thursday.

• A 55-year-old female resident surnamed Yuan took the first commercial trip on the platform and paid RMB 1.34 ($0.2) for her 3km ride (with a 95% discount), according to an announcement released Thursday (in Chinese).
• Qualified users can locate one of 67 autonomous cars in the vicinity and hail a ride by themselves by using the Apollo Go App. Baidu is currently operating the fleet from 7 a.m. to 10 p.m. each day in the area.

Context: Baidu, as well as self-driving unicorn Pony.ai, obtained approval from the head office of the Beijing High-level Automated Driving Demonstration Area to start charging for rides using autonomous vehicles (AVs) in the zone, China Daily reported on Thursday.

• Pony.ai said in an announcement (in Chinese) that it will gradually transition its free trial service, which began in April, into a commercial one in the future, without revealing further details.
• Baidu in May launched a fully driverless, paid robotaxi pilot project using 10 AVs in the Shougang Industrial Park on the outskirts of Beijing, and plans to expand the fleet to more than 100 vehicles during the Beijing Winter Olympics next February.
• The search engine firm claimed its robotaxi project offered 115,000 rides during the third quarter of this year, and its testing vehicles had logged 10 million miles as of September. Google’s self-driving unit Waymo announced in January 2020 that its vehicles had driven 20 million miles on public roads, Quartz reported.

Chinese self-driving startup WeRide is partnering with Guangzhou Automobile Group (GAC) to bring autonomous vehicles (AV) onto ride-hailing platform Ontime. It is part of a joint push toward the commercial deployment of robotaxi services, the two companies said on Thursday.

Why it matters: WeRide’s expanded partnership with automaker GAC is the latest example of the startup branching out to work with more companies as it develops self-driving vehicles and related services.  

Details: WeRide is working with GAC to integrate its autonomous driving system into the latter’s Aion S electric sedans and make them available for customers of Ontime, a ride-hailing subsidiary of the auto major, according to a joint statement issued Thursday (in Chinese).  

• The companies said they have been testing AVs equipped with WeRide’s latest sensor hardware on public roads in restricted areas of the southern city of Guangzhou. The goal is to offer a robotaxi pilot service to Ontime users in 2022.
• WeRide and GAC will work together to develop purpose-built new L4 self-driving vehicles. The current robotaxi model, based on the Aion S, is still a retrofitted version, a WeRide spokesperson said when contacted by TechNode on Thursday.

Context: Guangzhou-headquartered WeRide has been working since 2018 with GAC, which is Toyota’s and Honda’s Chinese manufacturing partner, to retrofit its software and sensors into GAC’s vehicles such as the Trumpchi GE3 crossover.

• The four-year-old startup launched its geo-fenced robotaxi pilot service to the public using a fleet of Nissan cars on the outskirts of Guangzhou in November 2019. Its fleet of 300 AVs has driven more than 7 million kilometers (nearly 4.4 million miles) as of writing. For comparison, Google’s self-driving subsidiary Waymo said early last year that its vehicles had logged more than 20 million miles since its inception in 2009, Reuters reported.
• GAC first launched its ride-hailing service Ontime in partnership with Tencent in Guangzhou in June 2019, later expanding to Shenzhen. It currently handles more than 300,000 rides a month.
• Rival T3, backed by three local automakers, is racing for market share and has completed over 1.2 million trips monthly. Didi remains the dominant player, however, with monthly orders of 20 million in August, Chinese media reported, citing people familiar with the matter.

Huawei ramped up its involvement in the Chinese electric vehicle (EV) space on Monday, offering in Shanghai the first view of the Avatr 11 sports utility vehicle (SUV) with partners Changan Automobile and CATL. The first model in the Avatr brand, the car features a full suite of Huawei’s autonomous driving technology. Huawei partnered with carmaker Changan and battery supplier CATL a year ago to form the Avatr premium luxury brand of EVs.

Why it matters: The Avatr 11 electric SUV will be the second mass-produced car to get Huawei Hi, a complete automotive hardware and software suite that includes the company’s operating system Harmony OS as well as computing platforms for autonomous driving.

• The three companies expect the new models to take a significant share in the Chinese premium EV segment and to fulfill their ambitions to establish a “world-class, high-end” Chinese car brand, Changan Chairman Zhu Huarong said on Monday at a press conference in Shanghai.

Details: The first EV model produced with Changan and CATL features a supercomputer developed by Huawei and running at 400 trillion operations per second (TOPS). That compares with Tesla’s 144 TOPS for its two-chip full self-driving computer.

• The electric crossover will have a driving range of more than 700 kilometers (435 miles) in a single charge with batteries supplied by CATL. It will also boast a high-volt, fast-charging electrical system that will support a maximum of 200 kW of charging power. That is a bit lower than the charging power of Tesla vehicles which, offer up to 250 kW.
• Changan plans to build the Avatr 11 SUV in a manufacturing plant in the southwestern municipality of Chongqing. Annual capacity is projected to be 350,000 vehicles, with delivery to begin in the third quarter of 2022. Prices have yet to be released but Chinese media, citing Changan sources, have reported a figure around RMB 300,000 ($47,000).

Context: Changan, CATL, and Huawei announced their smart EV tie-up back in November 2020. That was followed by the establishment of a joint venture with the state-owned automaker as the biggest shareholder in Avatr.

• Changan, Ford’s Chinese manufacturing partner, in August announced goals to have 35% of its annual car sales, or 1.05 million out of 3 million vehicles, be EVs by 2025. Rivals Geely and SAIC have previously launched their own premium EV brands, called Zeekr and IM, respectively, as Chinese companies step up their competition with Tesla. 
• Huawei and state-owned automaker BAIC in April co-launched the Arcfox-branded Alpha S, the first mass-produced vehicle model equipped with Huawei’s self-driving technology and priced from RMB 388,900. That’s more than 50% higher than the price for Tesla’s China-made Model 3, SCMP reported.

READ MORE: Huawei begins selling EVs in stores, may offset sinking phone sales: CEO

Self-driving startup QCraft will equip the next generation of its autonomous driving system with Nvidia’s Drive Orin processing chip. The chip can be deployed to both self-driving prototypes and mass-produced vehicles.

Why it matters: Nvidia claims the Drive Orin system-on-a-chip (SoC), unveiled in late 2019 and scheduled for shipping in 2022, is by far the “world’s highest-performance, most-advanced” processor for use in autonomous vehicles (AVs). Its use will allow QCraft to develop driverless vehicles for road testing and partially automated cars for the consumer market.

Details: Nvidia’s Drive Orin chipsets will underpin the hardware suite that will be fitted as standard for QCraft’s next-generation self-driving car fleet, the two companies announced as they unveiled the deal at an event on Tuesday.

• Capable of performing 254 trillion operations per second (TOPS), the Orin supercomputer will be the brain of QCraft’s vehicles, providing the necessary power to process inputs from multiple sensors as well as GPS and mapping data, according to a statement released Tuesday.
• QCraft will begin testing its AVs with the Nvidia-assisted computing platforms by the end of next month, given that it already has experience in adopting the chipmaker’s technology, chief technology officer Hou Cong told reporters during an online conference on Wednesday.
• The self-driving startup, backed by tech giants Bytedance and Meituan, on Wednesday also unveiled its third-generation hardware system, including nine cameras, five lidar units, and four millimeter-wave radar units, claiming a lower production cost and improved sensing capabilities.
• The technology costs of installing an array of sensors in vehicles will be lowered to between RMB 100,000 and RMB 150,000 ($15,620 to $23,430) in the next two years, chief executive Yu Qian told Chinese media last month. A testing AV reportedly costs at least $130,000 in sensors and computers, according to John Krafcik, former CEO of Google’s self-driving subsidiary Waymo.

Context: Nvidia has also signed a series of deals with Chinese electric vehicle upstarts (including Nio, Li Auto, and WM Motor), supplying their upcoming vehicle models with the chipmaker’s SoCs.

• QCraft has been operating a self-driving test fleet of nearly 100 vehicles for public transit and ride-hailing pilot services in nine domestic cities, including Shenzhen and Wuhan. It is also developing assisted driving technologies for several automakers, chief scientist Da Fang told TechNode in September.
• In August, it raised $100 million in a Series A+ led by YF Capital (the investment firm founded by Jack Ma) and participated in by Longzhu Capital, a venture capital fund of Chinese on-demand service giant Meituan.

Geely unveiled on Monday an electric semi-truck model. The company said to deliver the model with highly autonomous driving functions in 2024. Such a commercial vehicle from the Chinese automaker could pose a threat to Tesla and other manufacturers.

Why it matters: The launch will allow Geely to expand its presence in the global commercial market, poised for double-digit growth in the coming years, and to compete against Tesla’s long-awaited Semi truck model and similar offerings produced by Daimler, Nikola, among others.

Details: Called the Homtruck, the semi-truck will be available in several power options, including fully electric and plug-in hybrid, and feature a modern sleeper cab interior, which Geely said will give drivers extra comfort.

• Geely did not reveal many specifics about the first premium truck model in its new commercial vehicle group, to be branded Farizon Auto, but said in a Monday statement that it has begun taking advance orders with an initial deposit of RMB 2,000 ($313).
• Scheduled for delivery in early 2024, the Homtruck will have highly autonomous features that allow drivers to cruise around hands-free on certain roads, the company said, intending to achieve full autonomy in 2030.
• Geely, the parent company of Volvo, intends to expand its presence globally with the launch of the semi-truck, eyeing markets such as Europe, Korea, Japan, and North America, Mike Fan, chief executive of Farizon Auto told CNBC on Monday.
• Farizon Auto is targeting an annual sales of 250,000 clean energy vehicles in 2025. It expects that number to more than double by 2030 with a 20% share in the market, Fan said during a press conference in Shanghai on Monday.

Context: The Chinese electric vehicle (EV) industry has been on a rebound after a market slump that began in late 2019 and lasted an entire year. Carmakers sold 357,000 EVs in China in September, accounting for more than 20% of monthly new car sales for the first time.

• Geely announced on Oct. 31 that it is aiming for 40% of its annual sales, by 2025 to be new energy vehicles; that share would total 1.55 million out of the 3.65 million vehicles and would include battery electric, plug-in hybrid, and hydrogen cars. The company also promised to invest RMB 150 billion into research and development by 2025, SCMP reported on Nov. 1.
• Tesla unveiled its electric truck model Semi in November 2017 but has since delayed volume production until 2022. The Semi truck’s first customer, Pepsico, is expected to accept delivery of 100 pre-ordered vehicles in the last quarter of this year, CNBC reported on Monday, citing PepsiCo CEO Ramon Laguarta.

Among a rash of Chinese tech behemoths venturing into the car manufacturing business over the past year, newcomer Xiaomi may pose the most serious competitive threat to other carmakers. With a strong brand name and a dominant position in the country’s consumer electronics market, a Xiaomi car has the potential to turn the automobile and mobility industry upside down in the coming decade.

Now the Chinese smartphone giant is saying a Xiaomi car will be coming in the next few years. Speaking at Xiaomi’s annual investor day on Oct. 19, Chief Executive Lei Jun said the company is aiming to mass produce its first electric vehicle model for consumers in the first half of 2024. The company reportedly (in Chinese) has a master plan for its auto business, eyeing a total sales target of 900,000 vehicles within three years of production. That number would be almost equivalent to China’s total EV sales in 2019.

Bottom line: The auto industry has been uncharted territory for Xiaomi, but boss and founder Lei, 51, is setting wildly ambitious goals for the car-making project, which he calls “my last major entrepreneurial project.” A latecomer to the transition from fossil-fueled vehicles to the future of electric and autonomous mobility, Xiaomi is largely unprepared (in Chinese), even compared to newbie entrants such as Baidu and Huawei. To catch up, the company is pouring billions of RMB into the sector and placing some big bets on multiple startups as it attempts to build a complete auto supply chain.

Here’s a look at some of the company’s biggest deals in the auto industry. 

Chipmaking: Black Sesame

Black Sesame Technologies, a five-year-old Chinese auto-chip startup, announced on Sept. 22 that it had raised “hundreds of millions of dollars” from Xiaomi and other investors. Black Sesame became the smartphone maker’s first big bet in auto-chip designing. One of Xiaomi’s investor affiliates, Hubei Xiaomi Changjiang Industrial Investment Fund, led a Series C investment announced the same day as the strategic investment, which valued Black Sesame at over $2 billion.

Already backed by renowned investors including auto major SAIC, Black Sesame is one of the three domestic companies with the potential to develop high-performance central processors for next-generation electric and connected vehicles, an investment manager who declined to be named told TechNode in September. The other two domestic chip powerhouses are considered to be Huawei and Horizon Robotics. China’s consistent pursuit of self-sufficiency in chip manufacturing may be what made Black Sesame an attractive deal for Xiaomi, this person added.

Xiaomi and Black Sesame have yet to share details about any potential collaboration. And yet the Chinese chipmaking upstart in April unveiled its powerful A1000 Pro chipset. The chipset claims to have a processing speed of 196 trillion operations per second (TOPs), which would outperform Tesla’s full self-driving (FSD) computer running at 144 TOPs. Black Sesame also said its four-chip full autonomous driving system will be capable of a range of applications such as highway and urban driving. The system is scheduled for release with mass production vehicles by the end of 2022.

Autonomous driving: Deepmotion

Following the announcement of its own electric vehicles in March, the only acquisition that Xiaomi has publicly made known to date was its $77.37 million buyout in August of Deepmotion, a Chinese self-driving startup with Microsoft roots. Acquiring the team of a well-known but struggling software startup is expected to help Xiaomi to absorb the talent inside of the company and finally discover a path to develop its branded consumer vehicles with autonomous driving capabilities.

Founded by four computer scientists from Microsoft Research Asia, the biggest overseas research arm of the US tech company, Deepmotion in mid-2017 began working on high-definition 3D maps and localization functions for autonomous vehicles (AVs). However, the company never obtained the required license for surveying and mapping from the central government. It later pivoted to develop AI algorithms and software that enable the use of HD maps and camera sensors for vehicles to navigate the roads.

Thus, the hints are strong: Xiaomi will probably adopt a very conventional approach to self-driving technology by using multiple sensors to help AVs navigate, competing head-to-head against players such as Nio and Xpeng Motors in this space.

In separate moves, the Chinese smartphone maker earlier this year invested in Geometrical Pal, another startup that develops software solutions for radar sensors used in AVs, while also backing Zongmu Technology, a company with a specialty in software development for self-parking functions.

Lidar: Hesai

Xiaomi in June made news again by co-leading the $300 million investment in a top Chinese lidar supplier, making its first bet on what has been heralded as a crucial component enabling self-driving cars to perceive the world. The company, called Hesai, has long been among the highest funded lidar companies worldwide; its products are used in most Chinese self-driving cars. At least 10 out of the top 15 robotaxi developers worldwide are reportedly (in Chinese) among its clients, including Baidu and Didi.

China’s highest-valued lidar startup, Hesai used to develop mechanical spinning lidar sensors for self-driving prototype vehicles. They were usually perched on car roofs with a set of rotating laser sensors housed in motorized turntables to provide 360-degree vision. Such bulky rotating sensors are too unreliable and expensive for mass production vehicles. Hesai in 2019 therefore launched a more compact, solid-state lidar unit which it claims could spot small, dark objects at a range beyond 300 meters.

Chinese automakers and their lidar partners have been working to include lidar, still an immature technology compared with cameras and radar, in their future production vehicles for accommodating high levels of automation. Hesai said in a June statement that the $300 million war chest would be used to accelerate mass delivery of its solid-state lidar units to multiple auto clients without elaborating further. Xiaomi did not reveal details of a possible deal with the company. 

READ MORE: Lidar is hard—but it’s coming soon 

Batteries: Svolt

Partnering with battery makers has become a critical piece of automakers’ plans to secure enough battery supplies as they produce millions of EVs in the next few years, and Xiaomi is no exception. The smartphone giant has actually invested in four Chinese companies across the battery supply chain. In its most recent bet, the company joined a group of investors to pump RMB 10.3 billion ($1.6 billion) into Svolt, a battery maker formed by automaker Great Wall Motor.

The Series B, led by Bank of China Group Investment with participation by IDG Capital and others, has reportedly (in Chinese) pushed Svolt’s valuation to about RMB 36 billion—some 38% higher than just six months ago. A distant rival to the likes of CATL and BYD, three-year-old Svolt is stepping up efforts to jostle for market share with plans to increase its production capacity to over 200 gigawatt-hours (GWh) by 2025. That would be one-third of the capacity of market leader CATL.

As China’s EV sales continue to grow at an astonishing pace, Xiaomi, like many other automakers, is rushing to build a sustainable supply chain to make sure its future models won’t be held up by a battery crunch. Previously, the consumer electronics company had poured RMB 375 million into Ganfeng LiEnergy, the battery-making unit of lithium producer Ganfeng Lithium, according to a statement (in Chinese) released on Jul. 31. Another Chinese battery maker, CALB, also raised an undisclosed amount of funding from Xiaomi and others last December, reported Shanghai Securities News (in Chinese).

Conclusion

In a matter of months, Xiaomi has rapidly acquired capabilities, ranging from software development to chip manufacturing, which could facilitate the company’s ambitious plan to build a complete supply chain under its control and finally make EVs on its own.

However, the consumer electronics giant, still new to auto making, faces the formidable challenges of pulling together these partners from various sectors, managing an entire auto supply chain, and navigating persistent global supply disruptions. Furthermore, Xiaomi has yet to reveal where it intends to manufacture its EVs, triggering speculations about possible contract manufacturing with carmakers such as Great Wall Motor, while peers Baidu and Huawei moved quickly to partner with Geely and BAIC, respectively.

Previously an investor in both Nio and Xpeng Motors, Xiaomi now finds itself competing against these established EV makers. Nio and Xpeng earlier this year hit notable milestones, each delivering more than 100,000 vehicles to customers. Even farther ahead are the dominant EV market players, Tesla and GM’s Wuling. These automakers are all well prepared to defend their territories from attacks by upstarts like Xiaomi. The race will stretch long into the future.

Self-driving startup QCraft has begun operating an autonomous shuttle bus pilot project in an eastern Chinese city, the latest example of local entrants racing to make driverless transport commonplace in the country.

Two-year-old QCraft, backed by tech giants Bytedance and Meituan, currently operates a fleet of around 70 self-driving mini-buses in several major cities, including Shenzhen and Wuhan, the largest fleet of its kind in China.

Now, the company is expanding its footprint with the launch of a robobus project in the downtown area of Wuxi, a city in the eastern Jiangsu province, which started taking local residents on rides on Oct. 23.

Initially deploying five robobuses for three routes totaling 15 kilometers (9.3 miles), QCraft said its pilot service covers a range of about 10 square kilometers in the busiest portions of the city and connects major shopping centers and subway stations to residential properties.

On Oct. 23, TechNode got a look at the driverless future with a ride on one of QCraft’s Wuxi buses. The electric mini-bus model, called Longzhou One, is equipped with an extensive self-driving sensor suite including five Lidar units, four front cameras, and two millimeter-wave radar units, making it capable of seeing objects from long distances of up to 250 meters.

The self-driving buses still have a driver to reassure passengers and comply with government rules. The driver took control of the vehicle once during a 15-minute ride when a large bus zoomed past it in the overtaking lane.

The buses typically travel between 30-50 km/h (19-31 mph) and are currently programmed on fixed routes. Each bus carries a maximum of nine passengers and has a driving range of up to 200 km (124 miles) on a single charge. 

Passengers can access real-time transit information from the company’s app on their phones. The buses operate from 9 am to 6 pm on weekdays, which the company said will meet the needs of nearby residents with their daily commute.

After receiving $100 million from reputable investors, including Meituan and Jack Ma’s YF Capital, QCraft is on track to expand its test fleet to more than 100 vehicles by year-end.

READ MORE: Drive I/O | Meet the Chinese self-driving car startup with Google roots

Xpeng Motors will launch a pilot program for autonomous ride-hailing services in China in the second half of next year, Xpeng’s executives said at an annual tech day event on Oct. 24. The program is part of the company’s latest efforts to offer full-scenario autonomous driving capabilities by the middle of 2023. 

Why it matters: Xpeng expects the move to accelerate the development of its advanced assisted driving technology for mass-produced vehicle models. The company claims its upcoming advanced assisted driving technology will cover most traffic conditions. 

Details: Xpeng will operate a fleet of vehicles equipped with its advanced assisted driving software in Chinese urban environments in the second half of 2022, Wu Xinzhou, vice president of autonomous driving in Xpeng, told reporters during a media interview on Tuesday.

• Xpeng hopes to use the pilot scheme to study so-called “corner cases,” meaning traffic scenarios that do not happen very often and find possible solutions, according to comments made by Chief Executive He Xiaopeng at the Oct. 24 event. 
• Full details of the project are yet to be announced, but Wu said that the company will deploy its basic mass-produced vehicles rather than “retrofit vehicles with expensive sensors and semiconductors.” 
• “Xpeng will become the first carmaker in China that explores mobility solutions enabled by autonomous driving,” He said, adding that Xpeng has intended to focus on developing consumer cars with autonomous driving capabilities rather than become a mobility service company.
• On Oct. 24, the company announced plans to launch Xpilot 4.0, Xpeng’s advanced driver assistance system (ADAS), in the first half of 2023. Xpeng said the system will offer drivers unlimited, full-scenario driving capabilities.
• The Xpilot 4.0 will go beyond the current 3.0 version, which handles only Chinese highways and some expressway-style urban streets. It will also be superior to the upcoming 3.5 version, which features automated driving capabilities on urban roads, scheduled for release by next June.

Context: Compared to robotaxi companies, electric vehicle makers such as Xpeng have chosen different approaches in their quest to achieve fully autonomous driving technology. EV makers are gradually working their technology up from assistant driving to semi-autonomous driving, hoping to arrive at fully autonomous driving.

• In contrast, robotaxi companies such as Waymo believe there is no clear path from semi-autonomy to full autonomy. They chose to start their work at a high driving automation level. Baidu, Pony.ai, and WeRide are the early robotaxi players in China.
• Waymo has openly dismissed EV maker’s step-by-step approach. “It is a misconception that you can just keep developing a driver assistance system until one day you can magically leap to a fully autonomous driving system,” Bloomberg reported in January citing former Waymo CEO John Krafcik. A Waymo’s testing vehicle reportedly costs at least $130,000 in sensors and computers, according to Krafcik.
• Xpeng seems to disagree with robotaxi’s dismissal. “We will be ready to have a similar performance to any robotaxi company in China,” Wu told TechNode on Tuesday. “[The robotaxi companies] have to work very hard to find a path to a mass-production vehicle. If they don’t do that, two years from now, they will find the technology is already available in mass production, and their value will become much less than today’s,” Wu told TechCrunch in an interview in April.

Baidu announced on Tuesday that its highway driver-assistance system will be available to customers for the first time via electric vehicle maker WM Motor. The search engine giant is rushing to lead the race in popularizing partially automated features on consumer cars in China.

Why it matters: Advanced driver assistance systems (ADAS) technology is increasingly considered a major stepping stone to fully autonomous vehicles. Major Chinese auto and tech companies are looking to seize the growing market potential.

Details: The new WM Motor W6 sports utility model will have 29 autonomous driving sensors and Baidu’s Apollo Navigation Pilot (ANP) software. The vehicle will have semi-autonomous driving capabilities, such as automated lane changes on highways, according to an announcement sent to TechNode on Tuesday.

• Backed by Baidu since 2017, WM Motor announced it began delivering an earlier version of the W6 fitted with Baidu’s robotic valet parking feature at this year’s Auto Shanghai show.
• Baidu has also been working with automakers, including Geely and GAC, aiming to supply its Apollo autonomous driving system to 1 million vehicles within five years, Reuters reported in April, citing Li Zhenyu, a senior vice president at Baidu.
• No official release date for the updated WM Motor W6 was announced.

Context: Market research firm BlueWeave Consulting estimated that the global ADAS industry recorded $25 billion in revenue in 2020, and that number is expected to nearly triple by 2027, according to a Financial Times report.

• In April, Huawei and its manufacturing partner BAIC co-launched the first consumer EV equipped with Huawei’s autonomous driving technology and are on track to begin delivery in the fourth quarter of this year.
• Shanghai-based WM Motor delivered 13,378 vehicles in the third quarter of this year, representing a 137.5% increase from the same period last year, according to a statement (in Chinese).

QCraft, a Chinese startup co-founded by four former engineers of Google’s self-driving project, is developing a driverless vehicle expected to launch by the end of this year.

Called “Longzhou Space,” the autonomous shuttle will be “a hybrid between robotaxis and robobuses,” Da Fang, co-founder and chief scientist of QCraft, said on Sept. 17 on the sidelines of TechNode’s Emerge 2021 conference in Beijing. The vehicle is one of the two-year-old company’s efforts to expand its autonomous commercial fleet, which now numbers about 70 robobuses operating in six cities.

“It’s going to provide city bus services but, when the demand is not high, it also can fulfill the (function of) ride-hailing,” Da said. He added the latest product highlights the company’s thinking on the future of shared mobility, in which autonomous vehicles (AVs) will be seamlessly shareable among people and can be adapted to carry freight and for other purposes.

Da declined to reveal the name of its manufacturing partner. He told TechNode that the company is also developing assisted driving technologies for vehicles larger than minibuses and sedans in separate collaborations with automakers, without revealing further details.

Since its founding in 2019, QCraft has quickly become a rock star self-driving car company in China. Its co-founders include Yu Qian, a former team leader at Google Maps, as well as Da Fang, Hou Cong, and Wang Kun. All four are former software engineers at Waymo, Google’s self-driving unit that is widely considered a technical leader in autonomous driving.

Now employing about 200 in China and the US, the startup in August announced it had closed a $100 million Series A+ from new investors including YF Capital, a private equity firm founded by Jack Ma, and Longzhu Capital, food delivery giant Meituan’s industrial fund. This followed another major funding round of “dozens of millions of dollars” reportedly from TikTok parent ByteDance and other investors earlier this year.

TechNode took the opportunity at last month’s Emerge 2021 conference to interview Da, a former Waymo engineer in motion planning, one of the most challenging areas for autonomous driving. Da obtained a PhD in computer science at Columbia University where he focused on computer graphics and animation, developing simulation methods used in the modeling of liquids.

The following conversation has been edited for clarity and brevity.

TechNode: Behavior prediction is one of the hardest problems in autonomous driving. Companies including Waymo are training their driverless cars using simulated software to handle various unpredictable situations. How does that work?

Da: Behavior prediction is basically trying to model the world, including the agents such as the other vehicles and people, and predict how they are going to behave. The difficulty is that the future is not really certain. If you imagine a pedestrian standing on the side of the road and he is moving towards the middle of the road, a human driver probably knows how to react to it and to brake to let the pedestrian pass first. But there are uncertainties in the humans’ actions. The pedestrian may stop in the middle of the road or accelerate and speed through the road and reach the other side quickly. So your reactions need to change accordingly.

There are other difficulties as well. For example, negotiations. Sometimes prediction is not just about predicting what the other people or vehicles will move, but also about negotiating with them. If you imagine two cars merging into one lane and they approached the point at roughly the same time, one of them has to proceed first and the other has to brake a bit later. So this will definitely involve some negotiations in scenarios like this. Motion prediction is not only about predicting what other vehicles will do or will not do, but also about understanding how our actions will affect those predictions.

TechNode: There has been a significant debate over whether AVs should leverage multiple sensors or purely rely on cameras to navigate the environment. What is your take on that?

Da: Our view is that these different sensors are very complementary to each other. There’s just no reason to not use them at this stage of AV development. We know that right now, the most commonly used sensors are cameras, lidar, and millimeter wave radar. Lidar is really good at measuring distance. You can get lidar points that specifically, accurately pinpoint an object in a 3D space and know how far they are from us. That’s what cameras can’t do. With millimeter wave radar, you get speed measurement as well, but at a lower resolution.

READ MORE: DRIVE I/O | Lidar is hard—but it’s coming soon

There are many advantages to cameras in terms of high resolution. You can recognize textures. You can recognize small objects, like traffic cones and faraway pedestrians. That’s something you cannot do with lidar. But you will get a lot of negative impact in adverse weather like rains, snows, and frogs with both lidar sensors and cameras and that’s where radar sensors really shine. All of these sensors have their strengths and weaknesses. None of them by itself is going to be enough for dealing with all the scenarios. Basically, we have to use all of them together in order to build a really safe vehicle.

TechNode: But which one is better for AVs to detect and react to stable objects such as parked vehicles? That’s one of the major technical issues behind the recent Tesla and Nio crashes.

Da: We know that’s a very challenging problem for radar, mostly because of the low resolution. Radar sensors have reflections of these objects, but they have a difficult time telling them apart from backgrounds like the ground or buildings on the side of the road. Lidar will be much better because of a higher resolution. You can recognize objects directly apart from the background, even though it’s just a stationary point. With cameras, you can do the same, because you have much richer information in both resolution and color. I think right now lidar is proving to be the most important sensor of the three for highly autonomous driving.

TechNode: You and your founding team members worked at Waymo for a few years before setting up QCraft. What have you learned from that?

Da: One of the things that Waymo has done really well that we are trying to replicate here is that engineers do not just try to solve the problems, but try very hard to solve the problems in the right way. So, for example, the computation of the headway.

When you’re controlling the vehicle to follow the vehicle in front in the same lane at a comfortable distance, there’s this headway distance that you want to figure out. If an engineer is tasked with computing this optimal headway distance, how would he proceed? An average engineer will probably say, let’s put up a few driving logs, watch what human drivers have been doing, measure the distances, and maybe do some averages. They will get some numbers like 10 meters, 20 meters, 30 meters, depending on different scenarios, and then just use the numbers.

Obviously, that’s a really bad solution because it doesn’t generalize. Let’s say 20 meters may be good for a reasonably high-speed road, but it’s not good for expressways and urban areas as well. A better engineer would realize that it depends on the driving speed and the circumstances, such as the width of the road, but, most importantly, the speed of the two vehicles.

But that’s not good enough, still. If we have a really good engineer, who’s trying to always go one step further, he will think about when we are driving the car ourselves, why we will pick a different headway distance at different speeds. The answer is probably at a high vehicle speed, if we don’t leave enough room in front of our vehicle when the car in front of us brakes, we will not have enough reaction time.

That’s going straight to this concept called RSS, which stands for “responsibility-sensitive safety.” If we have a really good engineer who’s trying to find the right solution, they will basically discover RSS by themselves by solving this problem. That’s a really important thing that we would value. (Editor’s note: RSS is a mathematical model for AV safety framework developed by Intel’s self-driving division Mobileye.)

Creating robotic ride-hail vehicles capable of transporting people is proving a hard nut to crack, so Chinese self-driving companies are exploring other commercially viable paths to achieve mass adoption of the technology.

Robotaxi developers are “evolving into platforms” that enable a variety of vehicles for different user scenarios, Tu T. Le, managing director of consultancy Sino Auto Insights, said Friday during TechNode’s Emerge 2021 conference in Beijing.

“It is very difficult for autonomous vehicle companies to be profitable with the one single-use case of ride-hailing,” Le said. He added that applications in which autonomous vehicles (AVs) operate at low speeds with cheaper sensors and higher utilization rates could be “closer to commercialization en masse.”

Le’s comment highlighted the growing uncertainties surrounding the future of driverless ride-hailing services, or robotaxis. As a result AV players are shifting their attention to more practical applications.

A commercial robotaxi is the “holy grail” of autonomy, because it would address the most complicated traffic scenarios by offering timely and comfortable rides to customers in dense urban areas, according to Da Fang, co-founder and chief scientist of driverless technology startup QCraft.

The sector has been in a trough of disillusionment over the past few years, as Google’s self-driving unit Waymo and General Motors’ affiliate Cruise, both pioneers in AV research, suffered several setbacks in their much-hyped quests to launch a driverless ride-hailing service.

Waymo’s valuation in late 2019 was cut 40% to $105 billion by Morgan Stanley, followed by the departure of its CEO John Krafcik and several other executives earlier this year. The company’s vehicles had traveled autonomously for more than 20 million miles (32 million kilometers) on public roads as of early 2020, but Waymo’s fully driverless robotaxi pilot is still only available in certain areas of Phoenix, Arizona, after a couple of years of rigorous testing.

As capital is being reallocated into research and development of more promising applications,  autonomous bus service looks to be one of the more appealing bets. Backed by Chinese big tech firms Meituan and Bytedance, QCraft is one of the early movers in this sector. It has piloted a fleet of around 70 self-driving buses for public passenger transit in five domestic cities including Shenzhen and Wuhan since July 2020.

“There are several scenarios where we think AVs have actually become technologically viable and robobus is one of them,” Da said, adding that autonomous shuttles encounter situations similar to those faced by robotaxis in urban areas, but lower driving speeds and fixed routes reduce crash risks.

Chinese makers of the underlying technology, whether big tech companies or rising startups, are now training their sights on commercial vehicles, including trucks and vans for freight delivery. Toyota-backed Pony.ai has reportedly been testing its self-driving trucking technology in Guangzhou since December, while WeRide, Nissan’s bet in China, earlier this month announced moves to test driverless vans in everyday delivery scenarios.

Then came the news on Sept. 17 of Baidu’s entry into the logistics industry with the debut of Xingtu, the first heavy-duty truck model built upon Baidu’s “Apollo” autonomous driving system. Baidu, viewed as China’s answer to Google, had previously announced a seemingly ambitious target of deploying 3,000 robotaxis in 30 cities over the next three years. But that figure would mean  passengers in each city, on average, would have access to fewer than 100 robotaxis. 

“A pilot program with multiple model types of AVs also makes sense from a technology point of view. You simply need to collect data as much as possible by testing different types of vehicles in various kinds of scenarios,” said Da.

Deeproute.ai, a Chinese self-driving car startup, announced Tuesday that it had raised more than $300 million in a Series B led by Alibaba.

Why it matters: The investment is perhaps Alibaba’s most significant move in autonomous driving.

• The announcement didn’t disclose the size of Alibaba’s investment nor Deeproute’s valuation. Chinese media Che Dongxi reported (in Chinese) that Alibaba poured $200 million into the startup and said it’s valued at more than $1 billion. That would make Deeproute, a two-year-old startup, the newest self-driving unicorn, following Pony.ai and WeRide. 
• A Deeproute spokesperson declined to comment on the report.

Details: Alibaba led the Series B. Other investors include Jeneration Capital, a Hong Kong-based venture capital firm, and an investment fund of Chinese automaker Geely, according to the Tuesday announcement.

• Deeproute plans to increase its proprietary fleet of self-driving test vehicles from more than 30 to 100 by the end of this year. An autonomous ride-hailing service already opened to the public in Shenzhen, Deeproute’s regional headquarters, on July 19.
• The company is testing some sedans and trucks for Chinese auto giants Dongfeng Motor Group and Geely in the central city of Wuhan and the eastern city of Hangzhou, among other cities. 
• The company also plans to launch a robotaxi pilot project in California later this year. It received a permit from the state in June to provide autonomous rides. 

Context: Deeproute develops software for self-driving cars and operates several pilot programs to transport people and goods. In September 2019, Deeproute closed a $50 million pre-Series A, led by Fosun RZ Capital, Chinese conglomerate Fosun Group’s investment affiliate. The company secured an undisclosed amount in Series A a year later.

• Alibaba previously had invested in AutoX, another self-driving startup based in Shenzhen. The e-commerce giant’s investment program, Alibaba Entrepreneurs Fund, participated in AutoX’s $100 million Series A in September 2019.

Correction: An earlier version of this article incorrectly stated the number of Deeproute’s proprietary test vehicles as 20, not more than 30.

Chinese autonomous driving startup WeRide is testing a self-driving cargo van that can carry out delivery services. WeRide is partnering with carmaker Jiangling Motor Corporation (JMC) and courier firm ZTO Express.

Why it matters: Since the coronavirus pandemic, Chinese companies are seeing accelerated adoption of autonomous vehicles (AVs) for contactless delivery.

Details: WeRide on Thursday announced that it has been working with JMC, a Chinese manufacturing partner of US automaker Ford, to test a self-driving electric van designed for cargo delivery since the second half of last year. Courier company ZTO will purchase an undisclosed number of the vans to test.

• ZTO, backed by Alibaba, will help find driving routes for the pilot project, ZTO Vice President Jin Renqun said during an online press conference on Thursday. The courier will also test the van in real-life delivery scenarios. 
• The company said in a Thursday statement that the van can drive in “all-weather” conditions and on urban roads and highways. The van will also be equipped with Level 4 autonomous capabilities, which means the car can pilot itself without a human driver most of the time. 

Context: Guangzhou-based WeRide began testing self-driving minibuses in its headquarters city in January. It also completed a $310 million Series B, led by Yutong Group, a Chinese electric bus maker.

• Chinese authorities are laying regulatory grounds to help companies test more self-driving automobiles. Regulators in July released a new regulation (in Chinese) that allows AVs to be tested on selected highways and city roads.
• Tech giants and automakers have invested heavily in Chinese driverless startups. Inceptio, a robotruck startup last month secured $270 million from investors that included JD Logistics, the delivery arm of online retailer JD. Robot delivery firm Neolix recently closed a fresh funding round, led by Softbank Ventures Asia and CICC Capital.

READ MORE: The Chinese startup bringing robotaxis to the masses

Nio is enveloped in a public relations nightmare after Chinese traffic authorities last month disclosed the first known fatality involving one of the company’s vehicles using its partially automated driving system. 

Called Nio Pilot, the advanced driver assistance system (ADAS) has been a major selling point for the maker of luxury electric vehicles (EVs). Now it stands accused of overselling the capabilities of the technology. There could be more consequences to come as Nio is in advanced plans to enter the competitive mass auto market.

The Aug.12 crash of the Nio ES8, resulting in the death of the 31-year-old driver, has also had repercussions throughout the autonomous vehicle industry, with many fearing the prospect of tougher regulation and the loss of public confidence. Xpeng Motors and Li Auto last month quickly dropped the terms “autonomous” and “advanced” in describing their ADAS systems, respectively. 

The fatal crash: The accident occurred on a highway in Putian city in eastern Fujian province. The driver, Lin Wenqin, had placed his 2020 ES8 into Nio’s Navigate on Pilot mode, which basically takes control of the car during highway driving. The sports utility vehicle struck a highway maintenance vehicle stopped in the same lane, according to a statement (in Chinese) posted by local police on Chinese microblogging platform Weibo on Aug.18. The cause of the crash remains under investigation by Putian city police.

Shortcomings of ADAS: Pending results of the police investigation, whether the incident was triggered by a software glitch or human error remains an open question. It appears, though, that either Lin or the in-car system failed to recognize the stationary highway car in front of the ES8 and to move to another lane in response. 

• Similar to Tesla’s Autopilot system, Nio’s ADAS technology can keep a car advancing in its lane, maintain a safe distance behind traffic ahead, and can even change lanes automatically in some cases. However, currently these systems have difficulty detecting parked vehicles and braking for them.
• Nio’s ADAS system uses cameras, powered by computer vision algorithms, and radar sensors to detect and avoid obstacles, but there is room for errors when a vehicle encounters new situations which its AI algorithms had not detected during training.
• The radars of partially autonomous systems are not very good at distinguishing types of stationary obstacles, Raj Rajkumar, a professor at Carnegie Mellon University, told Wired in an interview about Tesla’s similar ADAS technology; the radars, therefore, are designed to ignore such obstacles in order to avoid false braking events.
• Similar incidents have occurred with Tesla drivers. The first Tesla fatality in the US happened in 2016 when a Model S with Autopilot active crashed into a white semi-trailer crossing the highway. A dozen more Tesla vehicles have since been reported for ramming into static obstacles including fire trucks and police cars; US regulators last month finally launched a broad investigation into the company’s technology, reported the New York Times.
• So far, autopiloting technology has been a regulatory blind spot in China and no higher authorities are known to have launched a broader probe into the Nio case in addition to local police.

Nio’s image in tatters: The deadly incident comes at a crucial time for Nio. Having struggled to gain a foothold in the luxury EV segment, the seven-year-old automaker is pushing to roll out its first mass-market car, eyeing a segment of the market where competition is fierce and margins are thin. Now its hard-won reputation as a high-quality premium brand is under threat.

• Nio has built up and benefited from an enthusiastic customer base similar to that of Tesla’s. However, the once incredibly loyal user community is becoming fragmented, as indicated by the response to a group letter (in Chinese) from 500 Nio owners, published online on Aug.18, in defense of the company. 
• More than 10,000 users joined in an online debate with the hashtag “objection to the joint statement” (our translation) in the chat room of Nio’s mobile app, disputing the group letter’s contention that there was “no misleading information” in Nio’s advertising of its ADAS technology. In the chat room, some Nio owners criticized the company’s service staff for overstating the capability of Nio Pilot before their purchases, while some blamed the company for providing little information about the ADAS functions and its limitations, according to a South China Morning Post report.
• In the latest development to hit Tesla’s challenger, Lin’s family contacted the Putian city police, alleging that Nio employees tampered with data from the crashed vehicle; Nio denies the charge.

Far-reaching consequences: Nio’s user manual warns that the ADAS system cannot detect stationary objects, including “roadblocks,” nor can it brake for them. Drivers are required to take control of their cars immediately when these situations arise. This means the liability for such accidents will probably lie with drivers themselves.

• Meanwhile, the auto industry is expecting strengthened regulation in automotive software to ensure safe operation and to tackle security issues for intelligent and connected cars. The central government earlier this month proposed new data security rules for autos, a move that Nio’s local competitor, Li Auto, last week said could result in more efforts to develop an assisted driving function in compliance.
• The publicity nightmare has also cast a shadow upon Nio’s business, highlighting the challenge for the company to maintain strong connections with a rapidly expanding user community, Chinese media reported, citing Zhou Zhanggui, a brand management consultant.
• Having gone through a liquidity crisis and aiming for an all-round expansion, Nio is at a critical juncture and must take steps to restore its image. The luxury carmaker is accelerating the pace to launch its first mass-market model under a new brand, reportedly scheduled for early next year, with plans to almost double its store count to 366 in the domestic market by the end of this year. “We want to provide better products and service at prices lower than Tesla’s,” said Nio’s CEO William Li last month.

Also in the news:

Xpeng plans foray into the premium market: As Nio moves to the mainstream market, Xpeng Motors is doing the opposite. The Alibaba-backed EV maker, which has maintained a price range between RMB 150,000 ($23,225) and RMB 300,000, is looking to expand in the domestic market by entering the premium-market segment with a high-end model scheduled for release in 2023.

• The new model will be sold for at least RMB 400,000, equipped with the company’s technology that could set it apart from its competitors, CEO He Xiapeng said during an earnings call on Aug. 26.
• He added the company is on track to roll out the Xpilot 3.5, the company’s ADAS technology, early next year and a 4.0 version in 2023, enabling the vehicle to automatically steer on city streets, not just on highways.

Internet giants doubling down on self-driving tech: Although the arrival of a truly self-driving car remains delayed indefinitely, Chinese tech giants are still betting heavily on self-driving startups with the intention to own a large share of the driverless driving future. Their investments come at a time when the Chinese government is establishing a looser framework with an expanded scope for testing self-driving vehicles, the South China Morning Post reported.

• Qcraft, a robobus startup formed by a group of former Waymo engineers, recently raised $100 million in a funding round from investors including YF Capital, a private equity firm founded by Jack Ma, and Longzhu Capital, the investment arm of life-service app Meituan.
• Xiaomi is acquiring Deepmotion, a Beijing-based startup working on high-definition maps for autonomous cars, as the Chinese smartphone maker ramps up its efforts to develop driverless car technology and mass produce its first EV in the next three years.

Smartphone giant Xiaomi on Wednesday announced that it is acquiring Deepmotion, a Beijing-based startup that develops digital mapping technology for autonomous vehicles. 

Why it matters: The acquisition is Xiaomi’s latest move in its bid to build its own intelligent connected cars. An expansion into China’s auto sector could greatly expand Xiaomi’s mobile ecosystem and create new revenue streams for the company.

Details: Xiaomi has reached an agreement to acquire Deepmotion Tech Ltd in a cash-and-stock deal valued at $77.37 million, according to the smartphone maker’s quarterly results, released Wednesday. The company did not reveal when it expects the deal to close.

• In an earnings call on Wednesday, Xiaomi’s president Wang Xiang said the purchase is aimed at accelerating the consumer electronics giant’s plan to develop autonomous driving technologies, which Wang called the most crucial part of intelligent and electric vehicles.
• Wang added that the company has been aggressively recruiting automotive engineers, and has established its self-driving team with a batch of 500 experts after kicking off its electric vehicle project in March.

Context: Xiaomi has struck several deals to invest in autonomous driving startups in recent months, as the Chinese tech giant ramps up its efforts to develop driverless car technology and mass produce  its first EV in the next three years.

• The company earlier this month raised its stakes in Geometrical Pal, a startup that develops software solutions that allow radar sensors in AVs to sense the environment. Xiaomi also invested in self-driving software developer Zongmu Technology in June, Bloomberg reported.
• Deepmotion was formed in mid-2017 by four computer scientists from Microsoft Research Asia, the research arm of the US tech company in the Asia Pacific region. In March 2018, the startup raised “dozens of millions of US dollars” from venture capital firms Redpoint China Ventures and Source Code Capital.

Neolix, a startup developing autonomous vehicles (AV) for delivery services, said on Wednesday it has raised “hundreds of millions of RMB” from Softbank Ventures Asia, an early-stage investment arm of the Japanese tech giant, among others. Hundreds of millions of RMB means the total funding amount is between about $15 million and $154 million.

Why it matters: The funding marks the latest bet by Softbank on Chinese startups as the Japanese tech giant seeks to dominate the future of artificial intelligence.

Details: Softbank Ventures Asia and CICC Capital, the private equity unit of Chinese investment banking firm CICC, led a Series B investment round in Neolix that closed earlier this year, a spokesperson of the Beijing-based startup said on Wednesday.

• Existing investors and venture capital firms Yunqi Partners and Glory Ventures also joined the round. 
• Neolix did not disclose the valuation at which the funds were raised.
• Neolix planned to use the funds mainly to expand its delivery service, while accelerating its steps into overseas markets, according to a Wednesday announcement (in Chinese).
• The three-year-old company added that it has deployed a fleet of nearly 1,000 robots in around 30 domestic cities, delivering over 1 million orders to around 300,000 users for retailers and restaurants such as KFC and Pizza Hut in China since mid-2020.

Context: Softbank has been a leading funder of ambitious Chinese companies for decades, but amid a wide-ranging crackdown on tech observers have asked if it is repositioning.

• Softbank CEO Masayoshi Son last week said only 11% of its Vision Fund investment has been directed into China since April. The company later told Chinese media that Son’s comments were misinterpreted, insisting that its commitment to China remains unchanged and that it continues to invest in the Chinese market.
• Softbank’s Vision Fund in February 2019 led a $940 million Series B of Nuro, a California-based self-driving grocery delivery service, TechCrunch reported.
• Neolix raised RMB 200 million in a Series A+ financing led by Chinese electric vehicle maker Li Auto in March 2020.

Qcraft, an autonomous vehicle startup backed by several prominent investors, announced on Monday that it had secured a $100 million investment. Investors include YF Capital, a private equity firm founded by Jack Ma, Longzhu Capital, the investment arm of life service app Meituan, and others.

Why it matters: Founded by a group of former engineers at Waymo, a self-driving car company owned by Google’s parent company Alphabet, the funding round serves as a stamp of approval for the two-year-old company. Chinese tech giants are betting big on Qcraft as they move more seriously into the robocar space.

Details: Qcraft has raised a new $100 million Series A+ led by YF Capital and venture capital firm Genesis Capital. Longzhu Capital, a venture capital fund of Chinese on-demand service giant Meituan, participated in the round. 

• Currently piloting around 70 robobuses open to the public in five cities around China, the two-year-old company plans to expand its test fleet to include more than 100 vehicles by the end of this year, according to its Monday statement (in Chinese).
• Qcraft is also exploring new applications in self-driving technology, with plans to launch a ride-hailing service in the eastern city of Suzhou later this year, LatePost reported on Monday, citing the company’s CEO Yu Qian.
• Qcraft declined to disclose its valuation when contacted by TechNode on Monday.

Context: Self-driving car companies are investing more time and energy in autonomous trucks and buses. The trend is driven by the fact that self-driving trucks and buses tend to have more predictable routes and are easier to manage than self-driving taxis.

• Baidu earlier this month rolled out a new version of its autonomous minibus Apolong. It started deploying the robobus in partnership with the government of the southern city of Guangzhou. The search giant said it has been operating an undisclosed number of first-generation Apolong buses in 22 domestic cities since mid-2018.
• Chinese robotaxi unicorn WeRide in December secured $200 million in a funding round from China’s biggest bus maker Yutong. Since January, the company has begun testing its fleet of robobuses in Chinese cities including Guangzhou and Nanjing.
• Qcraft in early 2020 closed a $24 million seed round from investors including IDG Capital and Vision Plus Capital, an investment firm led by Wu Yongming, an Alibaba co-founder, according to figures from business information platform Crunchbase. ByteDance also reportedly planned to invest $25 million in the company in March, according to Bloomberg.

Li Auto closed down 0.85% on its first trading day in Hong Kong Thursday. The Chinese electric vehicle startup opened at an issuing price of HK$118 ($15) per share. 

Why it matters: Li Auto is the latest Chinese tech firm listing in the US to seek a dual-primary listing in Hong Kong. Tech companies increasingly see Hong Kong as an attractive market as they seek to hedge risks when both Chinese and US regulators accelerate regulatory scrutiny.

Details: Li Auto’s Hong Kong debut met with a lukewarm market response. The company’s shares closed at HK$117 ($15.03), 0.85% lower than its issuing price, falling by as much as 2% soon after starting trading. 

• Speaking to reporters on Thursday in Hong Kong, Li Auto’s president Shen Ya’nan said the company has been considering a listing in the mainland, without revealing details.
• The company said it will use the proceeds from the Hong Kong listing to develop new car models and autonomous driving technology, and to expand charging infrastructure and sales networks.

Context: Backed by Chinese life services giant Meituan, Li Auto first went public on Nasdaq last July. The company is the second Chinese EV maker to seek a Hong Kong listing. Its rival Xpeng Motors raised $1.8 billion in Hong Kong in June.

• Li Auto so far has only one model for sale. The company delivered 8,589 cars in July, surpassing both its competitors Xpeng and Nio in vehicle deliveries for the first time in July.
• Both China and the US have issued new regulations that make it more difficult for Chinese companies to raise money in the US markets. In July, Chinese regulators proposed new rules requiring some Chinese companies to seek official approval before listing in overseas markets. The US has threatened Chinese companies with delisting over a dispute about accounting procedures.

Read more: Drive I/O | The untold story of Li Auto

Inceptio, a China-based robotruck startup, said it has closed a $270 million Series B on Tuesday. JD Logistics, Meituan, and PAG led the investment round. 

Why it matters: It’s unusual for two Chinese tech majors to join the same funding round, signaling that they see Inceptio as a key player.

Details: The funding round is led by JD Logistics, online retailer JD’s delivery arm, life service platform Meituan, and private equity firm PAG. Other investors include express courier Deppon and IDG Capital, Inceptio said in a Tuesday announcement. The investors didn’t provide a valuation for the company. 

• Inceptio has raised more than $490 million since its founding in 2018.
• The Shanghai-based robotruck startup will use the new funds to expand in-house research on self-driving systems and accelerate investments in electric trucks, the announcement said. 

Context: Both JD and Meituan have invested in autonomous driving, but progress on driverless technology has been slower than many expected. JD appears to be behind schedule on its own self-driving truck project.

• JD unveiled a highly autonomous truck model in Beijing in mid-2018 and anticipated its testing fleet would begin commercial operation in 2020, The Paper reported (in Chinese). But the company has yet to announce any updates. A JD spokeswoman did not respond to TechNode’s request about meeting the target.
• Meituan has also been investing in self-driving cars for a few years. The company invested in Haomo.ai, a startup developing automated driving systems for passenger vehicles and electric vehicle startup Li Auto. 
• Inceptio announced in March that the company will mass produce two self-driving truck models by the end of this year. The models will be co-developed with Chinese automakers Dongfeng and Sinotruk, respectively.
• Last November, Inceptio raised $220 million in Series A. Investors include battery giant CATL, logistics firm GLP, and Nio Capital, an investment firm partly formed by electric vehicle maker Nio. 

Tesla said on Saturday it will recall more than 285,000 vehicles in China to address safety concerns in its autopilot system, marking the automaker’s largest recall in the country. Tesla told local news the decision is not linked to previous safety incidents.

Why it matters: The recall raises questions over the carmaker’s future in China. The company’s prestigious image has soured quickly as Chinese Tesla owners this year began blaming the company for car malfunctions, including sudden accelerations and brake failures. 

• The recall affects over 90% of Tesla vehicles made and sold in China, according to figures released by the China Passenger Car Association. 

READ MORE: Safety questions and shady sales tactics are chilling the China-Tesla love affair

Details: Tesla will recall 285,520 cars, including Model 3 and Model Y vehicles built between 2019 and 2021. Affected customers can receive fixes remotely through system upgrades, without bringing the cars back to the dealers. 

• China’s market watchdog, the State Administration for Market Regulation, said it found safety risks in Tesla’s autopilot cruise-control systems. Drivers can easily activate the system by accident, causing the vehicle to accelerate suddenly, the regulator said in a statement (in Chinese). In some extreme cases, this problem can lead to collisions, the regulator said.
• The watchdog said the recall is Tesla’s response to a safety investigation initiated by the regulator. 
• However, Tesla said in its statement (in Chinese) that the recall is a result of the company “acting responsibly to the customers” and that it reported the recall voluntarily to the regulator (our translation). 
• A Tesla spokesperson insisted that the recall was proactive and unrelated to previous accidents in an interview with the National Business Daily (in Chinese). 
• The affected models include more than 211,000 Model 3 vehicles made in China between December 2019 and June 2021, nearly 36,000 imported Model 3s manufactured during 2019, and 38,600 Model Ys made in China since the start of this year. 
• Tesla did not respond to TechNode’s emailed request for comment.

Context: Since early last year, Tesla has faced mounting pressure in China over safety concerns and customer service complaints. The company also faces national security concerns in China. 

• After a car owner launched a high-profile protest at a car show in April, the US carmaker issued an unusual public apology, pledging to respect Chinese customers and China’s laws, and cooperate with the government on its investigations. 
• In late May, the company also established a data center in China, after the Chinese military reportedly banned staff from using Tesla vehicles due to concerns over the cars’ ability to collect confidential data. The center will store and process information generated by locally-made Teslas.
• Chinese government agencies have begun requesting staff to refrain from purchasing Teslas, due to data security concerns, a person with direct knowledge of the matter who asked not to be named told TechNode in early June. 
• In February, Tesla announced a recall in China affecting 36,126 imported Model S and Model X vehicles over touchscreen failures. Three months later, the company issued another recall involving 5,974 imported Model 3 due to safety risks posed by defective bolts, CNBC reported.

Baidu on Thursday unveiled a new robotaxi model, called Apollo Moon, with a manufacturing cost significantly lower than competitors. The Chinese search engine giant hopes to expand its business and commercialize an autonomous ride-hailing service.

Why it matters: The robocar is not being sold, but manufacturing costs are now comparable to the price of a high-end consumer car.

• High cost is one of the main barriers for robotaxi to see wider use. French market intelligence company Yole Développement estimated in 2018 that a robocar cost at least $200,000 on average. 

Details: Baidu’s Apollo Moon will cost the company RMB 480,000 (around $75,000) to manufacture. It costs the company less to manufacture than its rivals, but it’s hard to compare with since these are internal costs making. 

• Ride-hailing giant Didi’s autonomous vehicle costs the company about RMB 1 million (around $155,000), about two times Apollo’s, according to a Chinese media report last June. Baidu said at a Thursday press event in Beijing that the robocar is at a third of the cost of competitors’.
• Co-developed with Chinese automaker BAIC Group, the electric test vehicle runs on Baidu’s driverless software and has a suite of cameras and sensors, including two lidar sensors that provide the car surrounding visuals.
• The company also announced plans to add more than 1,000 of these vehicles to a ride-hailing test fleet while aiming to commercialize a nationwide robotaxi pilot service over the next three years.
• According to Baidu’s announcement, the company currently has a testing fleet with more than 500 vehicles and logged over 12 million kilometers (around 7.5 million miles), since its founding in 2013. The travel distance is about a third of Waymo’s, Google’s self-driving unit.

Context: In mid-2019, Baidu began testing a public ride-hailing service in a downtown area of Changsha, the capital city of central Hunan province, after road testing in suburban areas and closed test sites for six years. 

• The company has since expanded the robotaxi service to more Chinese cities, including Beijing and Chongqing, but only in limited areas. It began charging passengers with 10 selected testing vehicles on the outskirts of Beijing last month, becoming the first company allowed to do so by the Chinese government.
• BAIC Group also partnered with Chinese telecommunication giant Huawei to deliver a consumer-facing car model called Alpha S by the end of this year. Huawei will provide software for a self-driving mode. Drivers still need to stay attentive in the self-driving mode.

As Chinese automakers pour money into autonomous vehicles (AVs), they’re relying on another emerging technology to be the eyes of self-driving cars: lidar. Chinese carmakers are promising that models with lidar will hit the road in the next six months, likely marking the first time the tech sees widespread commercial deployment.

What is lidar? Well, it’s a lot like radar, but it uses lasers. It can pick out details and see small things better—a small dog crossing the road, a pothole. It can see things other systems, such as cameras and radar, might miss. 

But established lidar systems are bulky contraptions that are proving hard to integrate into consumer cars. They’re expensive, too, driving up the price of cars that use them for self-driving functions. For now, it’s mostly seen on prototype robo-cars.

Despite the challenges, most Chinese AV contenders are counting on lidar.

Five Chinese lidar startups say that they’re close to making it work. It’s a tough act: the device has to be small enough to fit in a sedan, reliable enough to trust on the road, and cheap enough to fit into the price of a consumer car. While they won’t be the first to deliver road-ready systems, Chinese companies could be the first to do it at a practical price. 

In this week’s issue, we’ll meet China’s leading lidar players and see how they’re trying to make the emerging technology work.

Robo ski-racks

Most lidar systems on the road today are mechanical spinning lidar on AV prototype vehicles. You’ve probably seen one—they’re the ones that look like half a jetski, or three portly Alexas strapped to a ski rack. If you saw it in China, it was probably made by Hesai, the Baidu-backed startup that’s the dean of the field.

Hesai has dominated the experimental generation in China, making the systems used on most Chinese and some international prototypes. At least 10 out of the top 15 robotaxi startups worldwide are reportedly (in Chinese) among its clients, including Baidu, Didi, and Pony.ai. 

But to address size and durability, lidar makers are now turning to “solid state” sensors that eliminate most moving parts. These can fit the system into a small box, around the size of a lunch box, which fits easily into the grill or tucks under the roof of a car. But miniaturization creates new problems with range, price, and reliability.

In early 2019, Hesai unveiled its latest solid-state device, called Pandar GT and boasting a detection distance of 300 meters, but it is still validating the product and negotiating with auto clients, according to a prospectus filed by the company in January. 

So far, Hesai hasn’t found a customer to put its solid state technology into a production vehicle. Baidu, a leader in China AV tech, has skipped lidar for its self-driving package, known as Autonomous Navigation Pilot, despite years of collaboration with Hesai in mechanical lidars for its test fleets. Speaking to Chinese media during this year’s Auto Shanghai expo, Baidu’s vice president Wang Yunpeng said the company is developing a “reliable and affordable” lidar sensor for production cars with partners, without giving further details.

The key challenges

Five Chinese companies have made real progress on consumer-ready lidar, using a variety of approaches that strike different balances between range, price, and reliability, and reaching deals with major automakers to put their sensors into cars. But they each have difficult technical problems to solve. 

Huawei and Robosense, a Chinese lidar upstart backed by Alibaba, are betting on a technology called micro-electro-mechanical systems (MEMS), which uses a tiny mirror (1 mm to 7 mm in diameter) to steer light. With only this piece of glass moving, the whole unit can be smaller than one that has to rotate as a whole. Robosense is currently making lidar s¯ensors for US electric vehicle startup Lucid Motors.

Both MEMS players are struggling with range: the latest offerings from the two companies only work at distances up to 150 meters.

Experts believe self-driving systems will need to spot objects at least 200 meters away to have enough time to react. 

The MEMS solution has proven to be superior in terms of size, speed, and cost over other types of lidar sensors, according to an article published by three University of Florida engineers last year. However, a short detection distance due to the small mirror is a key flaw and, to deal with it, systems will likely need a larger detector, complicating assembly, the paper said.

With its latest offering boasting an impressive distance of 250 meters, Sunnyvale and Suzhou-based Innovusion seem to have solved the range issue. Their solution uses lasers at a wavelength of 1,550 nanometers, rather than more common 905-nm lasers. Considered a “sweet spot” by lidar developers, 1,550-nm light allows longer-range measurement and poses less danger to human eyesight. When using 905-nm lasers, power is usually restricted to avoid blinding people.

But Innovusion has faced challenges with production, for a physical reason: traditional silicon chips can’t detect 1,550-nm light, and therefore developers have to make custom sensors with an exotic material called indium gallium arsenide (InGaAs), which is more costly and more complex to manufacture. Setting up a production line for this less common technology is no easy feat, and the product may not be cheap.

Speaking at an online conference in March, Innovusion technology chief Li Yimin said getting lidars to work well on production cars had turned out to be more difficult than he expected. Nonetheless, he said his staff have been working “day and night” to meet the early 2022 timeline target set by partner Nio. The Chinese EV maker has promised to deliver its first sedan model enabled with its lidar sensors, the ET7, early next year.

“We have to pull ahead the production schedule of many advanced technologies including lidar … This has posed a lot of pressure on our teams and the partners. We are fully focused on achieving this goal and pushing ahead despite all those challenges,” Nio’s chief executive William Li said during an April earnings call.

Xpeng Motors, with partner Livox, claims it will be the first Chinese automaker to deploy lidar on production cars this October. But it is facing other problems. Livox’s sensors boast a unique method of scanning objects in a spiral or flower pattern, rather than in traditional horizontal linear scanning patterns. This helps its sensors create a higher-definition map of the world and could enable more reliable autonomous driving capabilities, the DJI-backed lidar maker has claimed.

However, the unusual scanning style requires the sensor’s motor driver to operate at a high rotation speed of over 6,000 revolutions per minute, more than five times that of sensors made by major French lidar marker Valeo. These speeds pose a big technical challenge for the five-year-old startup to meet reliability requirements for autos, since high rotational speeds usually come along with high abrasion and reduced lifetime for motors.

Livox recently said that it has resolved the issue with manufacturing improvements, based in part on DJI’s expertise in mechanical engineering from making drones, according to a Chinese media report published last week. However, Xpeng CEO He Xiaopeng last month during an earnings call acknowledged that the company is still testing lidars from multiple suppliers and is “very open” to other choices for new models scheduled for launch over the next two years.

“With an all-round sensing performance on our cars and our production capabilities, we’re very confident that we can be complementary to some of the disadvantages of lidar technology,” He added.

Some Chinese automakers and lidar startups are also seeking overseas partners. In addition to the Robosense-Lucid hookup, Chinese legacy automaker Great Wall Motors, a manufacturing partner of BMW, has teamed up with Germany’s Ibeo as its source for lidar sensors on production cars.

The price is right

After technical barriers, lidar-enable cars will have to leap another hurdle: cost. The sensors don’t come cheap.

China’s low-cost manufacturing advantage appears to apply to lidar, with the offerings of local suppliers usually costing 80% less than international competitors, or below $1,000, French market intelligence firm Yole Développement wrote in a report published last August.

However, lidar cars don’t look cheap. The latest premium electric sedan announced by Huawei and BAIC in April, equipped with three lidar sensors, has a starting price of RMB 388,900 ($60,785), more than 50% higher than that of Tesla’s locally-built Model 3. 

R&D and onboard computing could be driving the cost. The Chinese telecom giant in April announced that it will double its annual auto R&D budget for self-driving cars to $1 billion this year, without giving a breakdown of its investments. Apart from three lidar sensors, the hardware stack of the BAIC-Huawei sedan also includes five more cameras, and five more radars than a Tesla Model 3’s. Although cameras usually take significant computing power in the vehicle, the task of combining data from multiple sensors also requires much computing power and a more complex vehicle architecture. 

Mixed opinions

Not everyone agrees that AVs will need lidar. Tesla has been heavily relying on a cheaper, camera-based approach. Nissan and Baidu, are also skipping lidar, relying on cameras, radar, and ultrasonic sensors for AVs. 

Most other major players, including Google’s Waymo and General Motor’s Cruise, consider lidar an essential part of developing safe autonomous cars. “Lidar sensors contribute to the redundancy and overlapping capabilities needed to build a car that operates without a driver, even in the most challenging environments,” wrote Cruise CTO Kyle Vogt in a post in 2017.

Chinese EV makers are betting on the lidar-based approach in competing against Tesla, and have gained chances to validate the technology. “At the current stage our top priority is not to secure as many contracts as possible, but to fine-tune our products and hit volume production,” (our translation) a Livox spokesperson told TechNode last month.

But lidar prices are falling. As the sensors get cheaper, the case for them looks more and more tempting. “Lidar guarantees high reliability for self-driving cars when vehicle autonomy is still in its early stage. Such redundancy is worth taking in the name of safety,” (our translation) Paul Gong, a China auto analyst at UBS, told TechNode last month.

Chinese ride-hailing platform Didi filed for an initial public offering on Thursday. The company plans to trade on either the New York Stock Exchange or Nasdaq. 

Why it matters: Valued at $62 billion, Didi is among the world’s five highest-valued unicorns. The company’s listing could be one of the biggest IPO this year. 

• The Beijing-based startup is also the world’s biggest mobility platform, with around 156 million monthly active users (MAUs) in the three months ended March 31, 2021, according to its prospectus. To compare, Uber had 98 million MAUs during the same period.

Details: Didi’s IPO filing highlights its quick recovery from the impact of the Covid-19 pandemic. The company reported a net income of RMB 196 million ($30 million) in the three months ended March 31, up from a net loss of nearly RMB 4 billion a year earlier.

• Didi’s revenue for the first quarter of this year was RMB 42.2 billion, doubling its figures from the same period in 2020. The company reported a RMB 10.7 billion loss on RMB 141.7 billion in revenue for 2020. 
• Didi plans to use 30% of the proceeds from its IPO to fund its international expansion. The company intends to spend another 30% on technology development, including electric vehicles and autonomous driving, according to the filing.
• Didi is ramping up its push into electric vehicles (EVs) to cut operational costs. Last November, the company released its first EV, the D1, with Chinese carmaker BYD. It also partnered with Guangzhou Automobile Group Co Ltd (GAC Group) to develop self-driving cars last month.
• Didi didn’t disclose how much money it seeks to raise, but listed a placeholder amount of $100 million on the filing.

Context: As part of its rapid expansion plan for the next three years, Didi is expanding into overseas markets and aggressively entering new verticals. 

• In November, the Chinese mobility giant launched its ride-hailing service in Auckland, New Zealand, three months after making its debut in Russia. It is also reportedly eyeing western Europe, where it seeks to enter the UK, France, and Germany.
• Domestically, Didi has been diversifying its business by expanding into grocery delivery and mini-van services for logistics. In January, the company reportedly closed a $1.5 billion round of funding for its mini-van service. It plans to spin off its grocery delivery service in a separate listing as early as next year.

The local government of Beijing on Tuesday granted the country’s first-ever permits for commercial deployment of delivery robots to JD.com, Meituan, and Neolix, allowing the companies to charge clients for driverless delivery services.

Why it matters: Robot vehicles are going into commercial use on Chinese city streets for the first time. It’s not the first time such vehicles will operate on city streets: China has previously granted permits to test passenger and commercial vehicles on city roads, and self-driving vehicles have gone into use in limited circumstances.

• Robot delivery services have previously been allowed to operate in geo-fenced areas such as university and industrial campuses.
• During lockdowns last February, local authorities in Beijing and Wuhan temporarily authorized Meituan and JD.com to use delivery robots delivering life and medical supplies to residents and hospitals on a limited set of public streets.

Details: Beijing-based tech giants JD.com and Meituan, as well as robotics startup Neolix, have been authorized to operate robot delivery services commercially within designated parts of the city’s Daxing district, state-owned media Beijing Daily reported Wednesday (in Chinese).

• The companies’ vehicles can operate on public streets in a designated area of 225 square kilometers, as well as the city’s mammoth new Daxing International Airport, and a 143-km stretch of highways, the report said. Robots face a speed limit of 15 kilometers per hour.
• Around 100 driverless cars deployed by JD.com and Meituan have begun delivering groceries to customers in the areas. Meanwhile, Neolix, backed by Chinese EV maker Li Auto, says that it expects to provide food delivery using a fleet of more than 150 vehicles by the end of June.

Context: The government is pushing automated passenger and freight transport services. Vehicle intelligence is one of the major goals of China’s current five-year plan, running to 2025.

• The Beijing municipal government earlier this year gave Baidu a green light to charge for rides on a fleet of 10 self-driving cars in an industrial park in the west of the city.
• Shenzhen in March started consultation on allowing companies to charge fees for their self-driving transport services in the city, expecting applications, such as last-mile delivery and robotaxis, to grow significantly over the next three to five years, Caixin reported (in Chinese).
• JD.com has been testing a fleet of 100 vehicles for grocery delivery in over 20 domestic cities, while Alibaba in March said its delivery robots have gone into operation into 15 university campuses in 11 cities.

With contributions from Emma Lee.

Huawei’s auto push won’t include making its own cars, the company said Monday. The statement comes on the heels of a series of high profile moves into auto technology by the telecoms giant, and reports that it plans to manufacture its own vehicles. 

Why it matters: Huawei’s statement comes amid unease from existing carmakers that Huawei will enter the industry by manufacturing its own cars.

Details: Huawei has not invested in any automakers and is not interested in acquiring majority stakes in car companies in the future, the company said in a statement on Monday.

• The Chinese smartphone maker reaffirmed that it will stick with a “long-term strategy” of manufacturing key components for intelligent and connected vehicles.
• “Persistent rumors that Huawei is investing in its own car production capabilities, or that we own shares in car manufacturers, are unfounded and do not stand up to scrutiny,” Huawei said.
• Shares of BAIC Blue Valley and Changan Automobile, two of the company’s major auto partners, plunged 10% on Monday following Huawei’s announcement. Both companies’ shares slumped a further 4.8% and 4.4%, respectively, on Tuesday.

Context: China’s tech and auto industries have long swirled with rumors of Huawei buying stakes in domestic car companies.

• The smartphone maker seeks to tap into the autonomous and electric vehicle market as its core businesses faces pressure amid US sanctions.
• According to a Reuters report in April, Huawei was looking to acquire a controlling stake in the EV unit of lesser-known domestic carmaker Chongqing Sokon, in a move that would enable the tech giant to make Huawei-branded cars. Sokon’s latest model, the Seres SF5, has been on sale in Huawei stores since last month.

Huawei has appointed the head of its smartphone business to take charge of its young vehicle technology unit, part of a wider management reshuffle as the telecommunications giant tries to break into the fast-growing autonomous and electric vehicle sector.

Why it matters: The appointment is expected to initiate a round of restructuring which will place Huawei’s nascent intelligent automotive solution (IAS) business unit and the team that develops and sell in-car services for automakers under its core consumer business group.  

• The IAS unit was set up in May 2019 to develop self-driving system as well as key components for autos and was previously under the Information and Communications Technology Infrastructure managing board.
• Huawei’s consumer business group, is seeking adoption for an Android alternative called HarmonyOS targeting various connected devices including autos other than smartphones.

Details: Richard Yu, chief executive of Huawei’s consumer business group, was appointed concurrently CEO of the auto solutions unit. Current head Wang Jun will remain as the president of the unit, a source with direct knowledge of the matter told TechNode on Wednesday. Chinese media first reported the shift, citing an internal memo dated Tuesday.

• Yu was also relieved from his role as CEO of Huawei’s cloud and artificial intelligence business unit, an appointment made three months ago and reported by the South China Morning Post. Zhang Ping’an, current president of Huawei’s cloud unit will be promoted as the CEO of the unit and led by Eric Xu, Huawei’s rotating chairman.
• A month after putting EVs on sale in dozens of its flagship stores, Huawei is ramping up a push into electric and connected vehicles. Yu recently set an ambitious annual target of selling 300,000 EVs next year, Chinese media reported Wednesday citing sources.
• The company has secured around 6,500 orders for the Seres SF5, a plug-in hybrid launched by its partner Sokon last month, according to the report. It has planned to sell EVs in at least 200 shops by the end of July and increase that number to more than 1,000 by year-end.

Context: Huawei has been seeking new growth drivers as its smartphone sales plunged globally last year. The smartphone business is running out of key components from US suppliers while being cut off from Google’s Android by the US sanctions.

• The telecoms company last month co-launched Alpha S, a premium electric sedan targeting Tesla’s Model 3, with state-owned automaker BAIC and pledged to start delivering self-driving capabilities for highways and urban streets at the end of this year.
• Ford manufacturing partner Changan is also reportedly (in Chinese) on track to unveil a new premium EV brand, co-developed with Huawei and Chinese top battery supplier CATL, later this month. Huawei last month announced to spend $1 billion in research and development for autos this year.

Traditionally a time for automakers to flex their muscles, the Auto Shanghai expo this year held a surprise: It was China’s big tech firms that took the spotlight, outshining some of the country’s leading EV makers. 

Huawei made a big splash, unveiling its complete self-driving car technologies as it gears up to compete as a central player in China’s autonomous vehicle (AV) industry. Baidu, China’s biggest internet search firm, was not to be outdone, proclaiming itself the undisputed AV industry leader. The company said it expected to equip 1 million new cars in five years with its software.

Some of the biggest startup unicorns such as chipmaker Horizon Robotics were also busy, forging alliances with a list of automakers during the event as they work to establish themselves in the booming industry.

Highlight 1: Chinese tech giants bet on smart EVs

Overshadowing traditional carmakers displaying flashy concept models and production-ready cars, Chinese tech giants generated big buzz at Auto Shanghai this year. 

Tech giants unveiled advanced connected and autonomous driving solutions along with ambitious growth strategies, generating headlines and lending cachet to lesser-known auto partners. In particular, deep-pocketed Huawei and Baidu showed how they are ramping up aggressive pushes into the industry.

Huawei was one of the biggest draws at the show. Crowds swarmed the Arcfox-branded Alpha S electric sedans on display at its booth, equipped with the telecom giant’s hardware and software and made by automaker BAIC. 

After three years of co-development, the two companies said that they are on track to deliver the Alpha S by year-end. According to Huawei and BAIC, the vehicle features “best-in-class” self-driving capabilities for highways and busy streets to customers in China’s four biggest cities. Its other customers that hail from outside of the four cities will get the function via over-the-air software updates within the next two years as Huawei continues to work on its AV mapping.

To reach this target, Huawei has been plowing resources into its new auto business. Its Automotive Solutions unit will beef up headcount 25% to 5,000 employees this year, Wang Jun, president of Huawei’s intelligent Automotive Solution business unit, told Chinese media during the show.

Hands-free driving on busy city streets is widely considered a key milestone for mass AV adoption, one that Tesla has offered in its full self-driving (FSD) package since March. Eager to offset its flagging smartphone sales Huawei has been chasing this capability as it ranks auto among its top-priority businesses, though it is years behind industry leaders. At the company’s global analyst conference a week before Auto Shanghai, deputy chairman Eric Xu announced that Huawei will nearly double its annual auto R&D budget to $1 billion this year.

Lingering questions among industry analysts TechNode spoke with include understanding what progress Huawei has made on the self-driving front so far—a question it has not yet addressed—and how much safer its self-driving cars will be compared with traditional autos. The tech heavyweight faces a significant uphill climb. Many automakers remain skeptical that the “wounded tiger” will manage to make cars itself, these analysts said.

Huawei’s moves into the auto industry present a significant threat to Baidu. Wang Yunpeng, a vice president at the search firm, recently went on the counter-attack in a talk with Chinese media during the auto show, insinuating that even by throwing money at the challenge, competitors stood little chance of quickly catching up. 

Baidu, Wang said, is in the same camp as Google’s AV unit Waymo—it’s on the verge of commercializing its technologies. To compare, “companies like Huawei and Didi are probably still at the stage of testing their vehicles on fixed routes,” Wang said (our translation).

Baidu’s robocars have logged 10 million kilometers (6.21 million miles) on public roads, around a third of Waymo’s. During the event, Baidu launched what it boasted was China’s most advanced driver-assist system. Called Autonomous Navigation Pilot (ANP), the technology enables autonomous driving capabilities for vehicles made by Baidu’s automaker partners. The system will be first available to owners of these vehicles in 20 cities by year-end and then over 100 cities by 2023, the company said. Baidu said its self-driving tech will power at least one new model per month beginning in July and equip more than 1 million cars with its software over the next five years.

With blurred lines between vehicles and technology, how much tech is in a Baidu- or Huawei-enabled smart car? Using as an example WM Motor’s W6, the latest crossover from the Baidu-backed EV maker, the tech giant is responsible for most of the digital technology in the car, from the voice assistant to the map navigation in the operating system. WM Motor also sources Baidu’s self-driving software and hardware suite including 12 cameras, 12 ultrasonic sensors, a radar system, and a computing platform, while it independently develops the car’s mechanics, such as the powertrain system.

Chinese carmaker Chery is also clamoring to join Baidu’s friend circle, while BAIC is one of Huawei’s oldest allies in the automotive industry. However, some of the bigger names in auto want full control in developing the next-generation of vehicle architecture. For that reason, China’s biggest automakers, SAIC and Dongfeng Motor, displayed their latest offerings with software developed in-house or by Chinese AV unicorns they have backed.

During the expo, SAIC began to take orders for its first sedan, the L7, under its new premium EV brand IM. Short for “Intelligence in Motion,” SAIC co-launched the brand with Alibaba in November to compete against Tesla. The Volkswagen partner recently raised its holdings in Chinese AV upstart Momenta, aiming to offer urban self-driving capabilities early next year. Meanwhile, Dongfeng announced (in Chinese) that it aims to sell a total of 1 million EVs and master fully driverless technologies within the next five years.

Experts TechNode spoke with were optimistic about Chinese automakers’ moves into smart, electrified cars, thanks in part to local tech giants. Domestic players could account for 70% of auto sales from the current 40% within the next 10 years, Liu Guanghao, an investment director at Shanghai-based venture capital firm BeFor Capital told TechNode. “These driver assistance features are industry-leading, and the car interiors, such as the digital dashboards, appeared forward-thinking. This could help traditional automakers reposition their brands to be more premium,” (our translation) Liu said.

Highlight 2: EV Big Three momentum slows

Amid the hubbub from big tech and traditional auto companies, Chinese EV contenders were comparatively quiet, with no mention of new models at Auto Shanghai.

Well-funded Nio, Xpeng, and Li Auto are considered emerging EV leaders and the most promising of China’s Tesla challengers. Now, as competition heats up, they are collaborating with smaller tech unicorns—such as Li Auto’s partnership with Chinese chipmaker Horizon and Xpeng’s partnership with DJI’s Lidar unit, Livox—in an effort to maintain their leadership positions in the sector. 

But their outlook may be clouding over after internet giants overshadowed them during the expo.

On the first day of the show Nio kicked off a massive expansion of its charging infrastructure, announcing that it would open 100 battery swap stations and 500 supercharging stations in an area spanning eight northern provinces during the next three years. Meanwhile, Nio president Qin Lihong acknowledged to Chinese media on April 19 that big tech’s push into EVs was a challenge for the company considering Huawei’s established retail network, and reaffirmed its goal to expand its sales network by 60% to 366 stores nationwide by year-end.

There has been growing concern over EV upstarts lagging larger players in new product and technology development going forward. Nio CEO William Li last month expressed confidence that it would release the ET7, its next-generation electric sedan, on time, slated for delivery early next year. It would happen, he confirmed, despite steep challenges in advanced technology adoption. The company said it is doubling its R&D budget to RMB 5 billion ($774 million) this year. “Auto intelligence is where this game may be decided,” Li told Chinese media during the auto show.

Li Auto is seen as falling behind its peers in the AV race, having not yet delivered highway self-driving functionalities to its customers. Feeling the heat at the auto show, CEO Li Xiang said April 20 on Chinese social media platform Weibo that its self-developed AV system will be able to compete head-to-head against those by Huawei and Tesla next year. The EV startup in September announced plans to adopt Nvidia’s advanced supercomputer Orin for its second model, scheduled to launch in 2022.

The six-year-old automaker also turned to Chinese AI unicorn Horizon Robotics for help, and the two companies during the show deepened their partnership to an “in-depth cooperation in building upgradable smart and electric vehicles” (our translation). Despite its best efforts, Li Auto may be too late to catch up and gain a competitive advantage, as tech heavyweights venture into EVs, an analyst told TechNode at the show. 

Li Auto in February assured investors that it will triple its R&D spending to RMB 3 billion ($464 million) this year. Since December it has raised around $2 billion from a new share offering and bond sales to ramp up in-house R&D capabilities.

Xpeng Motors is ahead of its peers in driverless technologies, but also failed to wow the crowd during the show, despite unveiling its second sedan, the P5, which it displayed at a press event in Guangzhou a week earlier. Touted as China’s first production model equipped with two Lidar sensors, an expensive and essential component for 3D perception, the P5 is expected in the first half of 2022 to self-navigate driving scenarios such as being cut off on busy streets.

However, Xpeng did not release the P5’s pricing information as planned, spurring concern from industry insiders that the company’s best days are behind it. Several insiders and analysts that TechNode spoke with said that the P5 launch fell short of expectations while the cost of the vehicle’s hardware suite has remained high, pressuring Xpeng in pricing the new product, people close to the company told TechNode during the show.

Xpeng fired back on April 22, saying on its Weibo account that it had secured more than 10,000 orders of the P5 in 53 hours after opening orders (with refundable RMB 99 deposits). “The market feedback was beyond our expectation,” (our translation) a company spokeswoman said to TechNode on Wednesday. 

Big tech disruption

Chinese tech giants at the Auto Shanghai 2021 disrupted the already-breathtaking pace of China’s new energy and autonomous driving world by doing what they were there to do: build consumer brand awareness and deliver advanced car technology solutions. The disruption is boosting the perception of Chinese-built vehicles—no longer synonymous with cheap, low quality cars—up the industry value chain.

This disruption is pressuring Chinese EV upstarts’ lead in the industry. These EV firms will have to convince investors that, after notching early wins, they can maintain their momentum in an increasingly crowded playing field. 

“Big tech’s entry into the market would inevitably erode the influence young EV makers have in the industry. This has created an alternative regarding the competitive landscape in the next five to 10 years,” (our translation) Paul Gong, China auto analyst at UBS, told TechNode on April 21.

Chinese search firm Baidu is launching a fully driverless, paid ride-hailing service in the outskirts of Beijing beginning on May 2, the company confirmed on Thursday.

Details: Baidu will begin charging passengers for rides with its fleet of 10 driverless vehicles in western Beijing’s Shougang Industrial Park beginning on Sunday.

• All rides will be offered without a human driver behind the wheel, though a driver will sit in the passenger seat to ensure safety, a spokeswoman told TechNode on Thursday.
• The launch follows Baidu’s fully driverless tests within geo-fenced areas in several domestic cities including Beijing and the northern city of Cangzhou during the second half of 2020.
• Shougang Park is more than 20 kilometers (12.4 miles) away from downtown Beijing, and is formerly the site of a plant belonging to state-owned steel company Shougang Group. The Beijing 2022 Olympic Games Organizing Committee is located at the industrial park.
• A Shougang Park representative told Beijing Daily (in Chinese) last week that a fleet of more than 100 self-driving vehicles will be in service during the 2022 Winter Olympics in February. The fleet will include robotaxis, robobuses, and delivery robots offered by Baidu, JD.com, and other companies.  

Context: Baidu in March became the first Chinese company granted permission to offer robotaxi rides to paying customers in Cangzhou, a city near Beijing in northern Hebei province.

The city of Cangzhou in northern China granted search giant Baidu a permit to begin commercial robotaxi services on some of its streets, the company said on Monday.

Why it matters: Baidu is the first Chinese company with permission to offer robotaxi rides to paying customers, which requires additional permits, and is a strategic milestone for its costly, years-long quest for self-driving cars.

Details: Baidu will be allowed to operate its autonomous ride-hailing vehicles with safety drivers on public roads spanning 229 kilometers (142 miles) in areas including the city’s downtown, the company said Monday in an announcement. The company can also begin testing out trip fares with its volunteers using discounts and coupons, according to a deployment permit issued by the government on Friday.

• A total of 35 robotaxis will kick off the deployment, the company said, though a spokeswoman did not specify a timeframe when asked.
• Cangzhou also granted permits for another 10 Baidu AVs to drive without a human driver behind the wheel, state-owned media Hebei Daily reported Wednesday (in Chinese).

Context: Cangzhou, the third-biggest city in northern Hebei province, was late to the AV race, lagging Beijing and Shanghai by over a year. However, it is catching up quickly by leveraging its partnership with Baidu, which is accelerating its autonomous transportation initiative.

• Baidu began piloting its robotaxi service to public in Cangzhou in August, but was prohibited from charging fees from riders until recently.
• Cangzhou granted its first AV testing permit to Baidu for a fleet of 30 vehicles in late 2019. The city has the country’s third-largest self-driving testing road network, following Beijing with 700 kilometers and Shanghai’s 560 kilometers.
• Guangzhou, China’s southern gateway, and Changsha, the capital city of the central Hunan province, have both opened around 160 kilometers of public roads for AV tests.
• Baidu has not yet received a green light from Changsha authorities to charge fees for robotaxi services. Neither has WeRide, a Guangzhou-based AV upstart.

Chinese robotruck startup Inceptio Technology plans to mass-produce trucks with intermediate autonomous driving functionalities as early as the end of this year, the latest stage in the commercialization of self-driving technology.

Why it matters: Commercial vehicles, including trucks and buses, are viewed as a more achievable application for self-driving technology than private passenger vehicles, and the market has been attracting significant investment.

Details: Inceptio‘s two self-driving truck models, co-developed with Chinese automakers Dongfeng and Sinotruk, are in their final stages of development. Mass production is set to begin at the end of this year, the company said at a press event Wednesday in Shanghai.

• The new models will be equipped with semi-autonomous driving functions, or Level 3 capabilities, such as lane changing and following the vehicle ahead in traffic jams, according to the company’s technology chief Yang Ruigang, a former Baidu scientist.
• It will also install a hardware suite capable of being upgraded to Level 4 self-driving capability, Yang added. The rigs will also feature multiple sensors including two Lidars which enable the detection and recognition of objects a kilometer (0.6 miles) away, a capability that competitor TuSimple said it was testing in 2018.
• Inceptio plans to install its self-driving technology onto more than 80,000 tractor-trailers in 2024. Each truck is expected to travel 250,000 kilometers per year on Chinese highways by that time, allowing the company to compile 20 billion kilometers of driving data per year and helping its push into full vehicle autonomy.
• A spokeswoman declined to discuss details about the company’s near-term delivery target.
• Level 4 autonomy refers to a fully autonomous system which can handle emergency situations, while Level 3 still requires that a driver intervene in emergency cases, according to a rating from the Society of Automotive Engineers (SAE).

Context: Inceptio is among several local robotruck startups backed by big auto and logistics names. In 2018 it closed a funding round for an undisclosed amount from Chinese battery maker CATL, and Nio Capital, an investment firm formed by the EV maker, and others.

• The Shanghai-based company last year raised a total of $220 million in two rounds from investors including Singapore-based logistics firm GLP and Tencent-backed fleet management company G7, Reuters reported.
• TuSimple, a US- and China-based self-driving truck startup, filed for a public listing in February, the Wall Street Journal reported. Morgan Stanley was tapped to lead the IPO, after helping the company raise a $250 million pre-IPO round last year.
• Toyota-backed robotaxi startup Pony.ai is stepping up its autonomous truck business to attract new investors. The company was granted a license in December by the Guangzhou municipal government in southern China to test trucks on public roads.

Li Auto reported losses of RMB 792 million ($121 million) in its first annual result as a public company, significantly reducing losses from a year earlier, but has drawn criticism for underinvesting in future innovation. Its shares declined 9.8% on Thursday.

Benefiting from rising electric-vehicle demand in China, Li Auto earned nearly RMB 9.5 billion in 2020. Its first model, the Li One, was China’s best-selling electric SUV during the year, according to figures from China Passenger Car Association. However, its delivery guidance of 11,500 vehicles in the first quarter of this year was almost 30% lower than the preceding quarter, which it attributed to the Spring Festival holiday and an uptick of Covid-19 cases in parts of the country.

Cost controls gone too far

The company narrowed its loss per share of $0.28, or net loss attributable to shareholders of $121.4 million, a 76% decrease from the previous year. This was partly aided by net income of $16.5 million in the fourth quarter from “short-term investment income” according to CFO Li Tie during the call with analysts. The EV maker also benefited from streamlining its sales operations, spending RMB 1.1 billion on selling, general, and administrative costs for the full year, 40% of what NIO spent on the same expense in the first three quarters of the year.

However, Li Auto’s investment into research and development was substantially less than its peers, raising concern among investors. Company executives had promised investors during an online briefing held a few weeks ago that it will accelerate the launch of new models to ease concern about its transition from EREV to all-electrics, according to a report released by investment bank China International Capital Corporation (CICC) last week.

In a conference call with analysts on Thursday, CEO Li Xiang said it has been on track to expand its range of products as part of a strategic move to prioritize business growth over cost control. The company promised to launch at least one new model every year starting 2022, including its first all-electric model scheduled for 2023.  

Ambitious outlook

The goal is to occupy a larger share of the market from mainstream to premium for an annual sales target of “several hundreds of thousands of vehicles” by the end of 2024, Li said (our translation). It also expects to build out a retail network of at least 1,000 stores by that time. The company had 52 stores in 41 Chinese cities as of December; NIO and Xpeng Motors had promised a respective 200 and 150 shops by year end.

The Beijing-based EV maker currently has only one model for sale and mainly focuses on extended-range electric vehicles (EREVs), a technology which features a small internal combustion engine dedicated to recharging the vehicle battery, designed to resolve range anxiety. However, recent policy changes in China is pressuring the company to accelerate its transition to all-electric.

Policy influence

Following Beijing, the Shanghai municipal government early this month unveiled a new policy for new energy vehicles, which excludes new purchases of plug-in hybrid vehicles, including EREVs, from free vehicle registration starting in 2023. Company president Kevin Shen on Thursday reassured investors, saying he expects EREV sales will continue to be strong until then. The company confirmed that it will release its second EREV model, a full-sized SUV with advanced driver assistance capabilities, in 2022.

Li Auto vehicles combine popular features and an affordable price tag, making it a more attractive choice than most internal combustion and electric vehicles in China over the past year. However, the company lags significantly rivals where self-driving technology is concerned— NIO and Xpeng Motor have emerged as major rivals to Tesla. The Li One crossover does not offer intermediate self-driving capabilities, such as navigation from on-ramp to off-ramp on Chinese highways, similar to Tesla’s Navigate on Autopilot and those NIO and Xpeng have both introduced in their vehicles.

CFO Li said the company will increase its R&D investment to at least $464 million this year and it will exceed $1 billion by end-2024, with half of the budget to be used in vehicle autonomy. CTO Wang Kai said that the size of its self-driving team will double to around 600 engineers by the end of this year as it opens its new R&D center in Shanghai with the end goal of 2,000 total employees.

Bigger rivals, including Tesla and a number of Chinese tech giants, pose a real and urgent threat. Wang said 2021 will be “the year of preparation” for the release of Li Auto’s new vehicle architecture next year, powered by Nvidia’s most advanced auto processor, Orin. “Similar features offered by our rivals, along with some brand new features, will also provided to customers for sure,” Wang said.

Correction: An earlier version of this article incorrectly stated that Li Auto plans to double the size of its R&D team to 600 engineers this year, not that of the self-driving team. 

SAIC Motor, the biggest automaker in China, will use processors for its self-driving cars from a domestic chip startup, throwing its weight behind a young upstart as Beijing accelerates plans to replace foreign-made chips with homegrown.

Why it matters: For one of the world’s biggest automakers to gamble a major strategic push on a young and relatively untested chipmaker signals the importance that Beijing places on rapid acceleration of self-reliance in advanced chips.

• Horizon Robotics lacks a global reputation compared with larger rivals such as Nvidia and Mobileye. In August 2019, it launched the Journey 2, its first auto chip to meet global auto stress-test standards, and began shipments in March.

Details: SAIC, China’s largest automaker and Volkswagen’s manufacturing partner, will use processors and software from Horizon Robotics, a rising Chinese chipmaking startup, for its upcoming car models that include advanced driver-assisted capabilities, according to a joint announcement released Monday (in Chinese).

• The state-owned auto manufacturer will also collaborate with the chipmaker to build and mass produce its next-generation, highly autonomous driving technology that is finally “capable of competing with Tesla’s full self-driving (FSD) capability,” the statement said (our translation).
• The company is valued at around RMB 30 billion ($4.64 million), a 50% increase compared with early 2019, and is eyeing a public listing on China’s Nasdaq-style STAR Market later this year, persons close to the company told TechNode.
• A company spokeswoman declined to comment on the valuation and the public offering when contacted by TechNode on Monday.

Context: SAIC is among a list of state-backed automotive majors now shifting towards Horizon Robotics as a domestic source for semiconductors. The chipmaker is considered to be China’s only alternative to global chip-making giants for auto processors.

• US sanctions of Chinese technology giants including Huawei, as well as a pandemic-fueled global chip shortage, is renewing Beijing’s urgency to cultivate a domestic chip sector by 2025.
• Carmakers are receiving state support. The Shanghai municipal government announced (in Chinese) Saturday that it is partnering with Horizon Robotics to establish its global research and development center in the city in an effort to accelerate the development and adoption of China-made central processing units for intelligent vehicles.
• In an interview with Chinese state broadcaster China Central Television (CCTV) last month, Horizon Robotics vice president Zhang Yufeng said that it is currently the only Chinese chipmaker with computing platforms for mass-produced vehicles.
• The company said recently that it shipped 160,000 of its Journey 2 artificial intelligence chips, which it boasts is more power efficient than Nvidia’s offerings, to Chinese automakers including SAIC, Changan, and Chery as of December, nine months after shipments began. It has outsourced production of its processors to Taiwan Semiconductor Manufacturing Corporation (TSMC) since 2017.
• Journey 5, its next chip model for advanced self-driving functions, is scheduled to launch before June and ship in 2022. Its computing power is expected to reach 96 trillion operations-per-second (TOPs), higher than the 72 TOPS of Tesla’s FSD computer.
• The company has an annual shipment goal of more than 1 million units this year.
• The five-year-old startup announced earlier this month that it had closed its $900 million Series C from investors including China’s State Development & Investment Corporation (SDIC), China’s biggest EV maker BYD, as well as Great Wall Motor.
• Competition with Tesla’s advanced self-driving capabilities is a catalyst for many Chinese AV manufacturers, and SAIC is no exception.

Correction: An earlier version of this story erroneously stated that the Journey 2 was Horizon Robotics’ first auto chip instead of the company’s first auto chip to reach global stress test standards.

Chinese self-driving vehicle startup Uisee has raised around $150 million in a new round of funding led by a state-backed venture capital fund, as the Covid-19 pandemic fuels demand for driverless vehicles.

Details: A national investment fund formed by a group of government agencies and state-owned enterprises led the RMB 1 billion funding round, according to an announcement released Monday. The fund has registered capital of RMB 147.2 billion. The Uisee valuation was not disclosed.

• China’s Ministry of Finance is the fund’s top shareholder holding a 15.3% stake, according to information on Chinese business research platform Tianyancha. Other major stakeholders also include China Development Bank Capital, the investment subsidiary of the state-owned policy bank, as well as China Insurance Investment Fund, a venture capital firm set up by a group of top Chinese insurers.
• Uisee said the proceeds will be used to enhance the development of its fully self-driving vehicle technology for use cases from passenger transport to logistics. The company’s public demonstrations have primarily showcased its logistics applications, enabling cars with self-driving technology to transport goods in and between warehouses.
• The Beijing-based AV startup said it is partnering with several big automakers to commercialize its technology, including equipping a fleet of more than 100 electric vehicles as trailers to transport auto parts in an SAIC-GM-Wuling plant in the southern Guangxi province. SAIC-GM-Wuling is a General Motors joint venture with its Chinese manufacturing partners.
• Uisee said that the fleet has logged more than 300,000 kilometers (186,411 miles) unmanned as of January in the 15 months since partnering with the carmaker. In an interview with Chinese media in April, founder and CEO Wu Gansha said that it is planning to enable daily operation of 1,000 AVs and achieve revenue of RMB 200 million by the end of 2021.
• Wu, formerly the director of Intel’s research institute in China, formed Uisee in early 2016. In February, it closed a Series B for an undisclosed amount led by Robert Bosch Venture Capital GmbH, the German auto supplier’s investment arm. Other early investors included Chinese venture capital firm ZhenFund, as well as Sinovation Ventures co-founded by former Google China head Kaifu Lee.

Context: Uisee is benefiting from pandemic-driven demand for unmanned transport, as one of a handful of Chinese AV startups getting a funding boost.

• Guangzhou-based WeRide earlier this month said it had closed a $310 million Series B from investors including China’s biggest electric bus maker Yutong, as the robotaxi startup begins expanding into self-driving minibuses.
• Plus.ai, a self-driving truck company backed by Sequoia Capital China, in November raised $100 million from Guotai Junan International, a Hong Kong-based investment and securities firm, and Hedosophia, a London-based VC firm.

Xpeng-backed self-driving startup WeRide has raised $200 million in a round of fresh funding from China’s biggest electric bus maker Yutong, it announced Dec. 23. The two companies also announced the development of a fully automated minibus with no controls.

Why it matters: The deal comes as Chinese automakers are pushing into the country’s autonomous vehicle industry, and as local authorities are allowing AV testing on public roads in bid to catch up in the global battle for tech dominance.

• The central government has reaffirmed a goal to “adopt highly autonomous cars on a massive scale” by 2025 in its finalized blueprint (in Chinese) for intelligent and electric vehicles released late October.

Details: Yutong put $200 million into WeRide in a solo Series B1 investment, according to a joint announcement released Wednesday. Based in the central Henan province, Yutong is China’s biggest medium and large-sized electric bus maker with more than a third market share in its domestic market, followed by BYD and Dongfeng Motors, among others.

• WeRide said that it and Yutong have jointly developed an autonomous minibus, which has no steering wheel, accelerator, or brake, without revealing details.
• The Guangzhou-based company did not disclose its valuation, nor details of other investors involved in its Series B.

READ MORE: The Chinese startup bringing robotaxis to the masses

Context: WeRide has been testing fully driverless vehicles on open roads in Guangzhou since July. It is the second company in the world to test fully driverless vehicles on open roads, following the US’s Waymo.

• The company in December 2018 closed an undisclosed amount in a Series A from the Renault-Nissan-Mitsubishi Alliance, along with Chinese artificial intelligence unicorn Sensetime, Xpeng CEO He Xiaopeng, and others.
• WeRide claimed to have offered more than 147,000 rides to 60,000 passengers in a robotaxi pilot program in a 55-square-mile area on the outskirts of the city as of November. 
• WeRide peer Pony.ai revealed last month an undisclosed amount of funding from the state-owned FAW Group, which it claimed was the first bet on AV by a Chinese state-owned automaker.

Chinese electric vehicle maker Li Auto on Tuesday said it will partner with Nvidia Corp to provide its next-generation SUV with a chipset and software platform that can be used for self-driving functions.

Why it matters: The partnership is the latest in a series of Li Auto’s efforts to develop its own autonomous driving capabilities to catch up in a race led by Tesla.

• The collaboration also means that Li Auto, currently a partner of Intel’s automotive sensor company Mobileye, is switching to Nvidia for a custom-designed chip and to retain control over the development schedule.

Details: Li Auto is teaming up with Nvidia and its Chinese partner Desay SV Automotive to develop a self-driving platform based on the Orin chipset and software stack for its next large-sized premium SUV which will launch in 2022, the companies announced Tuesday.

• Li Auto will be the first automaker to use Nvidia’s newest processor to facilitate upgradeable autonomous driving functions for its EVs, ranging from assisted driving functions and eventually, vehicle autonomy, according to an announcement.
• Nvidia in 2019 unveiled Orin, its next-generation system-on-a-chip (SOC) for automobiles, capable of performing 200 trillion operations per second (TOPS) using just 45 watts. The SOC is scheduled for production in late 2022.
• Its previous generation chip, Xavier, delivers 30 TOPS and consumes 30 watts of power, was first included in Xpeng’s latest P7 sedan which it began delivering in June. The two companies formed a partnership in late 2018.
• Li Auto currently offers assisted driving functions on its first Li One model based on the Mobileye Eye Q4 vision processor, which is also deployed on Nio’s crossovers.
• Li Auto’s new technology chief Wang Kai said to Chinese media in Beijing on Tuesday that Mobileye’s data center offerings, including algorithms for vehicle perception, was “sophisticated but not open enough” (our translation), leaving limited room for self-improvement.
• The Meituan-backed EV maker recently kicked off Level 4 autonomous driving development a year ahead of schedule. It is also ramping up plans to offer a hands-off Level 3 automated navigation driving function, similar to Tesla’s Navigation on Autopilot, as early as 2021.

Context: After big cash injections from US stock markets, young Chinese EV makers are speeding up efforts to close the gap with Tesla.

• Nio raised $1.7 billion earlier this month with a follow-on share offering. The Tencent-backed EV maker plans to use part of the proceeds to enhance self-driving technologies, following the hiring of a Chinese computer vision expert to lead its AV team of 200 employees.
• Li Auto currently has 60 self-driving scientists and engineers, and is planning to triple the size to 200 by early next year, according to Chinese media reports. It appointed Wang Kai, a former global chief architect at American Tier 1 supplier Visteon, last week as CTO to lead AV development.
• Li Auto’s collaboration with Mobileye will continue—the Israeli self-driving firm makes chipsets for the automaker’s first production model, the Li One.

China’s biggest search engine Baidu has transported more than 100,000 passengers in autonomous vehicles as part of a robotaxi pilot program, and the number will “soon surge to” more than 1 million, CEO Robin Li said.

Why it matters: Accelerating passenger numbers for Baidu’s self-driving pilot underscore China’s growing efforts to win the global autonomous vehicle (AV) race. It also signals that the Chinese government is lifting restrictions on driverless vehicle tests across the country.

• China’s top industry regulator on Tuesday revealed plans to release revised national guidelines for AVs to provide more public roads for passenger transport and logistics testing in self-driving cars, reported Shanghai Securities News (in Chinese).

Details: More than 100,000 riders have tried out Baidu’s autonomous ride-hailing service in cities such as Beijing and nearby Cangzhou, as well as Changsha in central Hunan province, and southwestern municipality Chongqing, Li said on Tuesday during the annual Baidu World 2020 technology conference.

• Li estimated that the robotaxi passenger number will soon surpass 1 million. In the future, with AVs and supporting roadside infrastructure deployed at scale in 2025, he added, more efficient public transport would contribute up to 4.8% to the GDP.
• Authorities granted the company permission on Tuesday to test two fully driverless vehicles in Changsha, meaning that a human safety driver is no longer required behind the wheel.
• Still, Chinese regulators require a human driver stay alert in the passenger seat, as well as a safety operator for remote takeover, according to a local media report (in Chinese).
• Changsha currently allows driverless testing on a public road 15 kilometers (around nine miles) in length and a closed testing facility, but has allowed Baidu’s AV fleet to test passenger transport in 130-square kilometer area in downtown since April.
• The internet giant expects to expand its testing scope significantly as more regional governments ease restrictions. Last month, Beijing local authorities issued licenses for the company to offer rides to the public on 700 kilometers of roads in the city outskirts.
• Baidu is operating a fleet of 500 AVs in a total of 27 cities worldwide. It is ranked in the top spot within China with 6 million kilometers test driven as of August, followed by Toyota-backed AV startup Pony AI, which has logged 2.5 million kilometers.

Context: Baidu is not the only company in China testing AVs without a trained driver behind the wheel. Weride, a self-driving startup backed by Nissan, Renault, and Mitsubishi, has been testing 10 AVs without a human driver present in the southern city of Guangzhou for two months. Weride earlier this month announced its robotaxi pilot has completed upwards of 90,000 rides.

• Home to Weride and Pony AI, Guangzhou currently leads the market from a legal perspective. In July, it allowed Weride to test fully driverless vehicles on public roads across an area of nearly 100 kilometers, making it the first company to do so in China.

Ye Jieping, the head of Didi Chuxing’s artificial intelligence research team, is stepping down after five years in the role as the Chinese ride-hailing platform sharpens focus on sustainable growth and profitability.

Why it matters: Ye is the latest in a series of departures from Didi this year. The country’s biggest ride-share app is streamlining its businesses to focus on revenue growth and efficiency, and scaling back on non-core projects.

• Didi was hit hard during the Covid-19 outbreak, and it is seeking out new revenue streams such as logistics and grocery deliveries. Didi president Jean Liu in May said its ride volume had recovered to 60% to 70% of pre-Covid levels in an interview with CNBC.

Details: Didi chief technology officer Zhang Bo is taking over to lead AI Labs, a team of around 200 scientists and engineers, from departing director Ye, Chinese media reported Monday citing people familiar to the matter.

• A fellow of the Institute of Electrical and Electronics Engineers (IEEE) and an associate professor at the University of Michigan, Ye joined Didi in 2015 when the company was in urgent need of top-tier AI talent in its head-to-head competition with Uber China.
• The AI academic led the deployment of machine-learning algorithms to manage dispatching vehicles on Didi’s ride-hailing platform to improve demand forecasts.
• Ye was named a deputy head of Didi Research Institute in April 2016, months before Didi merged with Uber China. Its oversight of AI Labs, which focuses on developing AI use cases for urban mobility services, came two years later.
• In an announcement sent to TechNode on Tuesday, Didi thanked Ye for his service to the company, saying that his work helped DiDi “break many new trails in AI application to transportation.”
• China’s ambition for wide AI applications has been hindered by security issues and a lack of high-quality data, among other challenges, according to a SCMP report, with scientists returning to academia from roles with local tech giants.

Contexts: Rumors linked Ye’s departure with the recent shift in positioning AI Labs as engineering-driven rather than research-led. Previously, the company’s new growth goals for the next three years triggered a series of management departures.

• Tiger Qie, a former Google scientist and the CTO of Didi’s ride-hailing business, recently left the company after being assigned to lead research collaboration earlier this year, Chinese media LatePost last month reported citing people familiar with the matter.
• Other executives who had left the company this year include Fu Junhua, senior vice president overseeing Didi’s public transport services; Tony Qiu, COO of its global business group; and Jia Zhaoyin, chief architect of the self-driving business.
• As it prepares for a potential listing in Hong Kong, the Chinese ride-hailing giant has reportedly refreshed its core values in recent months, calling for greater responsiveness to change.

Tusimple, a Chinese self-driving startup backed by delivery giant UPS, is reportedly seeking a US listing as early as the beginning of 2021.

Why it matters: If Tusimple does successfully go public in the US, it would be the first self-driving company in the world to do so on a major financial market. The initial public offering (IPO) could also blaze a trail for peers in need of capital.

• The IPO would test investor appetite for risk as the development of robocar technology has been slower than expected.

Details: Based on both Beijing and San Diego, Tusimple is planning to file IPO paperwork for a US IPO in the first quarter of 2021 at a valuation between $3.5 billion and $7 billion, Chinese media reported citing persons familiar with the matter.

• Meanwhile, the autonomous vehicle company is about to close a last-minute pre-IPO financing next month from a number of strategic investors, with commitments for a $300 million Series E at a target valuation of $2.9 billion, the report said.
• Volkswagen’s truck unit Traton is expected to pour $140 million into the startup, with Chinese media giant Sina Corp following with a $100 million investment. Sina Corp has been a longtime backer since its RMB 50 million ($7.2 million) Series A in 2016.
• The pre-IPO funding would take the unicorn’s total funding to more than $600 million. Still, that figure is nowhere near enough to sustain the company until it can lift its profitability, which CEO Chen Mo recently estimated to cost $1 billion, according to Chinese media.
• The company expects to achieve profit with an unmanned fleet of 5,000 trucks with revenue of $300 million annually, according to Chen. It currently pilots autonomous trucking services with 70 roborigs on the highways of several cities in Arizona and Texas.
• Tusimple declined to comment when contacted by TechNode on Monday.

Context: Chinese automakers and AV startups have also been experimenting with autonomous trucks in a number of domestic cities in bid to cut labor and fuel costs, but have made slow progress because of testing restrictions.

• China’s biggest carmaker SAIC in May proposed in May at China’s annual political event allowing the testing of Level 3 autonomous vehicles, which enable hands-off driving under limited conditions, for passenger transport and freight delivery services on Chinese highways.
• Tusimple’s self-driving truck fleet in China has been limited to testing in a geo-fenced suburban area in Shanghai where the city’s Yangshan Port is located. Tusimple CTO Hou Xiaodi in May said its robotrucks have traveled a total of around 45,000 kilometers (28,000 miles), without offering further details.

Chinese electric vehicle maker Nio has quietly hired a Chinese computer vision expert to lead its self-driving unit following the June departure of Jamie Carlson, its tech lead since early 2016.

Why it matters: The management change comes as Nio works on its self-driving technology development to catch up with peers after securing $1 billion in funding from the Chinese government.

• Nio has trailed behind rivals Tesla and Xpeng Motors in making autonomous vehicles after a series of layoffs last year when the company was under a massive cash crunch.
• It dismissed 141 employees in its third round of cutbacks in December, the majority of which came from its AV team, and partnered with Intel’s Mobileye to share the cost of developing robocars. The company currently has around 40 self-driving engineers based in the US, along with 160 in China.
• The latest hire will help Nio build in-house self-driving capabilities and partly offset its R & D headcount in the US, which it has been gradually reducing, people close to the company told TechNode.

Details: Ren Shaoqing, a computer vision expert and co-founder of Chinese self-driving startup Momenta, recently joined Nio as the assistant vice president of autonomous driving, according to three persons familiar with the matter.

• Jamie Carlson, Nio’s AV tech lead since 2016 and a former Tesla and Apple engineer, left the company in June, according to a Chinese media report. Nio declined to comment when contacted by TechNode on Friday.
• Ren will report directly to CEO William Li, taking charge of Nio’s perception solution development, which provides visuals of vehicle surroundings using cameras and sensors.
• Among the most highly cited Chinese researchers in self-driving technology, Ren in 2016 co-founded Momenta, an AV startup that develops camera-based software solutions for self-driving cars.
• Nio has backed Momenta since 2017, when Nio Capital, a venture capital fund established by the EV maker, led its $46 million Series B. German auto giant Daimler was also involved, along with other investors.
• Momenta also builds high-resolution maps that facilitate more accurate road navigation and enhanced safety for AVs. It was granted a permit to draw up high-definition navigation maps from Chinese regulators in 2018.
• Momenta did not respond to a request for comment.

Context: Nio’s progress in self-driving car technology has slowed over the past year. On the other hand, Xpeng Motor has advanced rapidly, and has a growing reputation in automated driving capabilities.

• During a call with analysts on Tuesday, Nio CEO William Li revealed that the proportion of owners who ordered full Nio Pilot self-driving package is around 25%, far lower compared with the 68% of Tesla buyers which opt in.
• Li added that Nio is on track to release its Navigate on Pilot (NoP) solution, which allows the vehicle to change lanes on its own, within this year, while acknowledging its self-parking feature was “not as competitive as Tesla’s.”
• Meanwhile, Xpeng boasts the highest auto-parking success rate among all vehicles available on the market, enabled with a dozen sensors and HD map solutions, and plans to provide its vehicles with an autonomous lane change feature on highways later this year.

Chinese self-driving startup Pony.ai will begin testing self-driving cars in Shanghai as part of a government push for global leadership in the development of autonomous vehicle technology.

Pony.ai will work with city regulators to deploy a self-driving fleet for test drives on public roads in northwestern Jiading district, the company announced Saturday along with the Shanghai municipal government during the annual World Artificial Intelligence Conference (WAIC).

The company did not disclose the number of cars in the fleet or project timeline.

The AV upstart, with headquarters in Silicon Valley and the southern Chinese city of Guangzhou, was valued upwards of $3 billion after securing earlier this year $462 million in a Series B led by Japanese auto giant Toyota. The Pony.ai fleet of more than 100 vehicles has traveled a total of more than 2.5 million kilometers (around 1.6 million miles) in China and the US combined, around a tenth of what Google’s self-driving unit Waymo has logged.

The move will thrust the AV unicorn squarely in the Chinese self-driving race. Mobility giant Didi as well as AutoX, a rival company backed by Alibaba, are piloting autonomous ride-hailing services in Shanghai. The three companies are currently the rising stars in China’s AV competition, and are ranked within the top 10 for self-driven miles in California’s annual self-driving report.

Pony.ai’s Shanghai debut will come just two weeks after Didi began offering rides to members of its early rider program within a geo-fenced area of around 100 square kilometers (39 square miles) in Jiading district.

Still, Chinese AV startups may be a long ways from mass-producing fully automated cars because of costs and technical and regulatory hurdles. Each of Didi’s custom-built Volvos are equipped with nearly 20 sensors including three Lidars, seven cameras, and a bunch of radars, and cost more than RMB 1 million ($143,000) per unit. Didi expects to operate more than 1 million self-driving cars on its platform by 2030, Meng Xing, COO of Didi’s self-driving subsidiary said last month in a webcast.

Weride, an AV startup backed by the Renault-Nissan-Mitsubishi Alliance, kicked off its robotaxi program with a fleet of 20 Nissan vehicles in its home city of Guangzhou late last year. Guangzhou in southern China on Friday gave the green light to Weride to test 10 self-driving cars without safety drivers on public roads. A Weride spokeswoman confirmed to TechNode on Friday that it was the second company worldwide to test fully driverless vehicles on open roads, after Waymo.

Chinese AV startups have accelerated moves to transport passengers via self-driving cars as the government is eager to make inroads in the technology’s development. The Beijing municipal government released China’s first rules for AV road testing in December 2017, while Shanghai issued in September the country’s first permits for AV passenger service pilot programs to SAIC, Didi, and BMW.

US electric carmaker Tesla is expanding its Chinese engineering team to accelerate the launch of self-driving features in the country as it pursues “full vehicle autonomy” by the end of this year, CEO Elon Musk said on Thursday.

“I really want to emphasize that it’s not just copywriting sort of stuff from America to work in China. We will be doing original design and engineering in China,” Musk said in a recorded video speech played on Thursday during Shanghai’s annual World Artificial Intelligence Conference (WAIC).

The electric vehicle giant maintained an earlier statement that its vehicles will be capable of “basic functionality for Level 5 autonomy completed this year,” according to Musk.  

Level 5 (L5) autonomy refers to a fully autonomous driving system which can handle all driving tasks without the need for human guidance, according to definitions set by the Society of Autonomotive Engineers (SAE).

Musk also said that Tesla has already produced the hardware needed for full self-driving capabilities, including an in-house designed AI chip known as Autopilot Hardware 3. The company can achieve L5 autonomy “simply by making software improvements,” he said.

Tesla has been ramping up its hiring in China, creating positions in departments from data engineering to server architecture as part of a broader strategy to localize software and user data in the world’s biggest auto market, according to a report from Chinese media. It had 3,200 employees in China as of late last year, Reuters reported citing its chairwoman Robyn Denholm.

The announcement comes as competition for market share with Chinese EV companies has intensified amid slowing growth. Chinese Tesla challenger Nio partnered with Intel’s automotive sensor company Mobileye to jointly mass-produce highly automated vehicles, which are scheduled for release in 2022. Alibaba and Xiaomi-backed Xpeng Motors, meanwhile, released their first sedan, the P7, with an advanced driving-assist platform which the company said was optimized to handle Chinese traffic conditions. CEO He Xiaopeng in April said the company will introduce a highway self-driving function to car owners with over-the-air updates next year.

Traditional automakers are also catching up. Changan Automobile launched earlier this year what it said was China’s first volume-production vehicle model with Level 3 autonomy. The state-owned automaker sourced self-driving chips for vehicle perception from Horizon Robotics, a Chinese chipset startup backed by Intel, Hillhouse Capital, and Sequoia Capital China.

Tesla pulled ahead of local automakers with the delivery of a record 14,954 China-made vehicles last month, a fifth of the country’s total EV market share. Meanwhile, Nio’s June deliveries almost tripled year on year to 3,740 units, while Meituan-backed Lixiang followed with sales of around 2,000 vehicles during the month.

Young Chinese EV makers sold a total of 9,470 units in June, accounting for 14% of the EV segment, compared with a mere 7% market share the same period a year earlier, according to figures from the China Passenger Car Association (CPCA).

Self-driving startup Tusimple is looking to raise $250 million in a funding round which will support plans to remove safety drivers from its robotruck fleet as early as 2021, a person close to the company told TechNode.

Why it matters: The funds would be critical for the company’s expanding efforts to commercialize its technology. However, its valuation is now so high that most venture capital firms have been deterred, said two people with the knowledge of the matter.

• San Diego-based Tusimple has been viewed as the player with the most potential in the self-driving truck sector, achieving unicorn status with a valuation of $1 billion after closing its Series D1 early last year.

Details: Tusimple is seeking to add $250 million to its war chest, appointing investment bank Morgan Stanley which recently sent proposals to potential investors on why the company is poised to succeed, TechCrunch reported Friday citing people familiar with the matter.

• A person close to the company confirmed the fundraising efforts to TechNode on Tuesday, adding that the AV startup intends to close the pre-IPO funding round within the next six months for a possible US stock market listing as early as 2021.
• Tusimple did not respond to request for comment.
• The company is looking to start offering driverless freight delivery services in the US next year along with partners such as auto supplier ZF, CTO Hou Xiaodi recently told Chinese media.
• Hou added that it currently operates a fleet of 50 trucks with logistics centers in Dallas, Houston, and San Antonio, Texas.
• It has been running a pilot cargo service since last March, partnering with UPS to test self-driving trucks on a stretch of highways between Phoenix and Tucson in Arizona.
• The US logistics giant took a minority stake in the company months later as part of its $215 million Series D closed last September.
• The AV startup in March announced it was doubling the number of trips in its service for UPS to 20 round-trips in Arizona and Texas per week.
• Speaking with Chinese media in late 2018, CEO Chen Mo said it plans to launch a combined fleet of 1,500 roborigs in US and China in 2021, and poised for IPO after 2021 when safety drivers are completely removed.

Weride, a Chinese self-driving startup backed by the Renault-Nissan-Mitsubishi Alliance, is making its autonomous vehicles available for ride-hailing on Alibaba’s map platform Amap, also known as Autonavi. Starting Tuesday, riders in Guangzhou can summon one of WeRide’s self-driving electric cars for a ride through the app, the company said.

Why it matters: Autonavi is currently the most popular mapping and navigation service provider in China and the partnership is expected to enable the AV startup to accelerate the pace to scale up the robotaxi business and make the technology more widely available for public riders.

• Autonavi ranked in April as top of the most popular mapping service with 902 million monthly active users (MAUs), nearly double that of Baidu’s mapping app, according to figures from Chinese moble internet research firm Trustdata (in Chinese).

Details: Customers can hail one of Weride’s self-driving cabs via Autonavi or proprietary ride-hailing app “WeRide Go” in a geo-fenced area of 144.7 square kilometers (around 55.8 square miles) across the Huangpu and Guangzhou Development districts, the company announced Tuesday.

• The service is available from 8 a.m. to 10 p.m. every day, with over 200 pick-up/drop-off spots. The size of the fleet has recently doubled to 40 Nissan electric cars running on public roads on the outskirts of Guangzhou, the company said. It has a testing fleet of 100 vehicles in China and the US.
• China currently requires all AVs to have a safety driver behind the wheel while on the road.
• Guangzhou-based Weride launched China’s first robotaxi pilot service available to the public in the city late last year. The company claimed completion of 8,396 orders for 4,683 passengers with zero accidents in December, the first month of operation, without revealing the latest figures.
• Googles self-driving unit Waymo completed 4,678 trips for a total of 6,299 passengers in its first month of a robotaxi pilot project in September last year in California. Waymo’s robotaxi service is limited to employees and their guests, Techcrunch reported.
• Weride said the number of orders in April tripled from a month earlier, after a temporary suspension during the Covid-19 outbreak.
• Weride has been working on a Series B funding round since September last year. Its early backers include RNM Alliance, Chinese venture capital firms Qiming Venture Partners, Kai-fu Lee’s Sinovation Ventures, and Harry Shum, former chief of artificial intelligence research at Microsoft.

Context: Chinese self-driving startups and mobility giants have been pushing hard to meet the technical and regulatory challenges needed in a journey towards a driverless future.

• Speaking at an online meeting hosted by South China Morning Post on Tuesday, Weride CEO Tony Han said China lags behind US in self-driving technology advancement by about one to two years, but is “more advanced in operations.”
• Didi Chuxing is about to launch a commercial robotaxi pilot service in Shanghai “very soon,” Meng Xing, COO of Didi’s self-driving unit said on Tuesday during an online conference, expecting mass production of AVs to begin in 2025. 1 million AVs could be available on the platform by 2030, Meng added.
• Alibaba-backed AV startup Autox claims it won Shanghai approval to trial passenger transport services ahead of Didi, adding that Shanghai has been conservative in handing out licenses, Financial Times reported last month, citing CEO Xiao Jianxiong.
• It in April also announced a partnership with Autonavi, with plans to begin autonomous ride-hailing services in Shanghai, without revealing further details.

It was 2017, and the future of driving was right around the corner: Fleets of autonomous cars would cruise city streets while self-driving buses swerved around pedestrians. Three years ago, tech companies around the world, including Nvidia and Audi, felt confident enough in AVs to predict this driverless future would be a reality by 2020.

Venture capital funds snapped up self-driving startups, plowing cash into dozens of these companies in China and around the world. Pitchbook figures show the global deal count in the AV sector nearly tripled to 127 in 2017.

The industry had to grow up eventually, and it’s happening now. Small players are leaving the market as it matures around a few success stories; in China, central planners are pushing back targets to match reality. Easier, less flashy applications like delivery-robot autonomous trucks are getting more attention from investors.

Plenty of engineers are still working on the dream of L5 fully automated cars. But for now, we’d better get used to the existing L2 parking-assist features and L3 office park shuttles.

Q1 AV fundraising

Competition in China’s self-driving market is heating up, driven by a few companies that dominate fundraising.

Only ten Chinese AV startups won investment in the first quarter of 2020, yet these ten startups conquered a third of all investments in 2019, according to an analysis of public records and data from TechNode and Beijing-based consultancy EO Intelligence.

Some of the lesser-known companies winning new war chests claim to control 90% market share in their own domains, a possible sign of maturity among these firms.

As investors realize that the commercialization of AV technology is still a long way off, they are betting larger amounts on more mature companies instead of making smaller investments in a wider range of early-stage startups.

In fact, much of this year’s activity was driven by just two companies: self-driving startup Pony.ai and lidar maker Hesai.

In January, Hesai closed its $173 million Series C, led by German Tier-1 supplier Bosch, among others. A month later, Pony.ai announced it had raised $462 million at a valuation of $3 billion, in what was at the time the biggest-ever funding round in China’s self-driving industry.

Most other companies did not disclose the value of their funding rounds, instead saying they raised “dozens of millions of RMB.” The two exceptions include an autonomous mining startup that claimed to have closed a RMB 100 million ($140,000) round and a delivery robot maker that doubled that number.

China is aligned with global industry trends. Around the world, mobility deal volume and total investment fell while a few late-stage AV companies raised larger sums.

Plans scaled back

A 2017 government plan anticipated that more than half of all cars sold in 2020 would be equipped with autonomous driving functions—but the installment rate of major assistive driving functions on cars was less than 20% in 2019.

Although investors and innovators are rushing to get L3 vehicles on the road, those cars aren’t ready for real traffic conditions. So far, deploying full autonomy means lowering the speed (to roughly under 40 km/hr) and restricting them to a very limited area, which usually means a local community, a school campus, or a park.

“We’re following special-case AVs very closely,” Qi Lei, the investment principal at Alliance Venture, Renault-Nissan-Mitsubishi’s global investment organization, told Chinese media. She brought up parks and old people’s homes as being easier sites for robots to navigate safely.

But the problems of getting L3 passenger vehicles on the road were highlighted by Baidu’s 2017 self-driving minibus model, Apolong. With the high price tag of RMB 1.5 million per unit, Apolong’s market performance fell short of expectations last year. Baidu immediately denied the reports with the release of a second-generation model, without revealing sales and cost details. It is unlikely that Apolong is affordable enough to be rolled out widely.

In the latest blueprint released by the National Development and Reform Commission earlier this year, the top economic planner declined to give a specific goal for AV development, instead by saying the country would need to reach “mass production” of intelligent vehicles with conditional automated driving functions by 2025.

Still, another action plan released late last year by China’s industry ministry shed some light on Beijing’s hopes for the AV sector. The report predicted that sales of intelligent and connected cars are expected to constitute 30% of new car sales over the next five years.

“The national guidelines will drive growth in China’s AV industry … facilitating cost reduction and efficiency improvement as the supply chain will move in the same direction,” according to analysts. However, as China currently lacks legislation governing self-driving cars, analysts expect mass adoption of L3 automation of passenger vehicles will probably happen no sooner than 2021.

The government remains confident enough to start writing rules for these future cars. Beijing promises to finish drafting technical standards for commercial vehicles—including those for driver monitoring systems and automated lane changing—by the end of 2020. Research on regulations on driverless passenger transport and unmanned delivery are also among the priorities, indicating that legislation for unmanned vehicles has been put on the table.

Autonomous deliveries

Given the difficulties of building affordable passenger AVs, growing emphasis is now being put on autonomously delivering goods. The COVID-19 pandemic also has driven the need for safe, contactless deliveries.

AV companies are racing to fulfil this niche. Three out of the 10 Chinese AV startups raising funds in Q1 are making robots for grocery delivery, according to TechNode’s analysis of funding data. Meanwhile, six AV companies that secured financing over the past two months claimed that their sensor-based algorithms could facilitate trucking rigs with the capability to drive themselves on Chinese highways.

This trend partially explains why investors have piled into Chinese robot delivery startups during the first three months of this year:

• Uisee, formed by Wu Gansha, a former director at Intel’s laboratory in China, secured an undisclosed amount of fresh funding in February in its Series B from German auto supplier Bosch, among other investors.
• In May, Uisee said its fleet of 75 self-driving cars has driven more than 15,000 km in a pilot project in working with SAIC-GM Wuling, General Motors’ light-vehicle joint venture with Chinese automakers SAIC and Wuling Motors.
• In March, Neolix, a partner of Baidu’s self-driving platform Apollo, announced having closed an RMB 200 million Series A+ led by Chinese electric vehicle maker Li Auto (aka Lixiang) and followed by existing backers including Yunqi Partners.
• In the same month, Beijing-based White Rhino raised an undisclosed amount of funding from Chinese investment firm Estar Capital. The company was formed by a group of former Baidu engineers and transported medical supplies to a makeshift hospital in Wuhan during the outbreak.

Venture funds are also pursuing self-driving trucks for freight deliveries on Chinese highways, as the Chinese government forced the installation of autonomous emergency braking (AEB) systems on commercial vehicles last year.

• In March, Maxieye, a company that claims a market share of nearly 90% in the AEB sector, announced it had closed its Series A from Chinese Tier-1 supplier SORL Auto Parts, without revealing financial details.

• Maxieye said its perception algorithms combining data from a variety of sensors could enable warning and braking based on navigation with merely 1% perception errors from 1 to 50 meters, compared with the 50% error rate achieved by some competitors.
• Soon afterwards, Inceptio revealed a $100 million new war chest from Singapore’s logistics giant GLP, among other investors. The Shanghai-based robotruck startup is pushing forward the mass production of L3 autonomous trucks with partners including Dongfeng Motor, China’s second-biggest automaker, by the end of 2021.

“Innovators and VCs have been through a learning process over the past several years since 2016. We are having a better sense of the fact that there is a very high ceiling to achieve vehicle autonomy—and that the lifecycle either of the technology per se or of the business operation is a lengthy and complex one.”

—Inceptio CEO Julian Ma, speaking to TechNode

Looking ahead, self-driving companies are still among the primary targets for VCs, but the rise of unicorns means more difficulty for early-stage startups to raise capital. “It’s a race with incentive capital over a very long term,” said Ma. As automakers and startups struggle to find nearer-term solutions to monetize their technologies, they’re hoping that regulators will remove the barriers in their path.

China’s biggest automaker SAIC Motor has proposed supportive regulations for highly autonomous vehicles among other rules on the sidelines of the country’s annual political event in Beijing on Wednesday.

Why it matters: China’s largest annual political gathering of legislative delegates and political advisers, known as the “two sessions” or “lianghui,” kicked off on Thursday. The suggestions from the country’s top automaker could shed a light on the government road map for AV adoption in the coming year.

Details: SAIC, manufacturing partner to Volkswagen and GM’s, urged the central government to pass legislation that will allow the road testing of Level 3 and above self-driving cars on Chinese highways first in certain areas and then nationwide, Chen Hong, president of SAIC wrote in a proposal (in Chinese).

• Chen is a member of the Communist Party of China and delegate to the National People’s Congress.
• He also called for piloting passenger and freight delivery services using self-driving cars without safety drivers behind the wheels in some “developed areas” including two Shanghai suburbs.
• SAIC has so far invested in five Chinese AV startups, including AutoX which is about to launch its robotaxi pilot project in the northwestern Jiading district of Shanghai around the end of this month, along with ride-hailing giant Didi.

Robotrucks: The Shanghai-based automaker is also eyeing the adoption of AV in the traditional logistics industry.

• It has claimed that its first autonomous rig has begun transporting containers at the city’s Yangshan Port.
• The company on Wednesday expected its robotruck fleet to transport 20,000 containers this year with more than 1,000 autonomous trucks to be commercially deployed over the next three to five years.

Context: Top executives of Chinese auto majors have brought proposals ranging from intelligent cars to vehicle electrification in bid to buck the downward trend in the Chinese markets and drive a new auto technology boom.

• Wang Fengying, president of Great Wall Motors proposed to re-provide subsidies to boost development of small-sized electric vehicles. Great Wall Motors in 2018 partnered with BMW to jointly manufacture the electric models of the famous Mini car brand as early as next year.
• Most small-sized EVs, normally with a short driving range, had been deprived from the government incentive plans when Beijing in 2018 raised the minimum requirement for EVs to a range of above 150 kilometers (94 miles). The threshold was further raised to 300 km in the latest subsidies scheme.

Chinese self-driving startup Hongjing Drive on Wednesday announced it has raised “tens of millions of RMB” in its Series pre-A. This is the second venture deal in China’s AV industry in two weeks amid an enhanced national push to drive an automotive technology revolution.

Why it matters: The recent deals reveal a modest recovery of investors’ confidence after government initiatives were introduced.

• China’s state economic planner earlier this year delayed its ambitious plans for five more years of “mass production” of intelligent vehicles with “conditional” self-driving capabilities from 2020 to 2025.
• The central government has set a goal to draft technical standards for intelligent vehicles by year-end to lay the foundation for adoption ramp-up in the next five years.

Details: Hongjing Drive, a Chinese supplier of AV computing platforms, has closed an undisclosed amount of fresh funding led by Silicon Valley venture capital firm BlueRun Ventures. California-based TransLink Capital and existing investor China’s Linear Capital both followed on.

• Hongjing was founded in 2018 by Liu Feilong, a former engineering lead at GM’s autonomous driving business unit
• It specializes in developing a scalable computing platform that serves as the brain for autonomous vehicles.
• The company has offices in Detroit and Shanghai-based and is looking for a presence first in low-level automated driving businesses while exploring use cases in the L3 semi-autonomous market.
• It is working on L3 autonomous trucks with China’s state-owned automaker FAW and Nio’s manufacturing partner JAC that could deliver at least 5% fuel cost reduction and half of labor savings for the traditional logistics industry.

Context: On April 29, Inceptio, a Chinese self-driving truck startup announced it has raised $100 million from Singapore’s Global Logistic Properties Ltd (GLP) among other investors.

• The company is planing on making “several thousands of” L3 robotrucks with OEMs including China’s Dongfeng Motor by the end of 2021, CEO Julian Ma said speaking with TechNode last month.
• The global mobility market venture deals shrank by 40% from a year ago to $33.5 billion in 2019, Wired reported citing figures from the data and research company Pitchbook.
• The market witnessed a rebound with $11.8 billion spent by VCs on global mobility companies in the first quarter of this year, a 62% increase year on year.

How should autonomous vehicles (AVs) be programmed to interact with pedestrians? What about non-intelligent vehicles such as other cars and bicycles? While AV technologies themselves—intelligent computing and AI, LIDAR sensors and so on—receive a great deal of government and popular attention, AV technology domestication is an overlooked topic that will influence how fast the industry can develop.

Human behavior changes when confronted with new technologies; we slow down at a speed hump, we smile in front of a camera, and we cross the road when the pedestrian light is green. Science and technology studies scholars call this technology domestication; how do people consume, modify, reconfigure, and resist technologies? Technology domestication is user experience writ large.

As a post-doctoral fellow at the Hong Kong University of Science and Technology, I ethnographically explored this topic in 2019. Over six months, I met and interviewed dozens of people across Beijing, Guangzhou, Hong Kong, Shanghai, and Shenzhen.

My interlocutors worked inside AV and traditional automobile companies as data scientists, engineers, marketers, and strategy advisers, as well as along the supply chain making sensors and other components. I also spoke with industry analysts, journalists, and lawyers.

I did not expect technology domestication to be so top of mind among my interlocutors, but it was. People involved in actually making AVs, as well as those responsible for getting them onto China’s roads, were especially perceptive. In fact, while other countries have been focused on developing AVs that “fit in” with existing behaviors, Chinese researchers are approaching the topic differently.

Yi Zeng, a prominent computer scientist and Director of the Research Center for Artificial Intelligence Ethics and Safety in Beijing, encourages Chinese AI companies to better understand how people will ultimately use and interact with the products and platforms they are creating.

Fitting in

You might think it is the AVs job to fit in with people. A team of social scientists working on AV behavior at Nissan thought just that. They worked hard to develop AVs with “socially acceptable behavior,” defined as behavior that takes into account existing social and cultural practices related to mobility and human-automobile interaction.

Perhaps due to established road rules and ingrained behaviors around Silicon Valley, where the team was placed, Nissan focused on teaching AVs to simulate how a typical driver navigates a vehicle in the presence of others.

Take a pedestrian crossing that does not have traffic lights (i.e., a zebra crossing): even with clear rules of engagement, it is common for drivers to momentarily make eye contact—maybe also using subtle facial gestures—to signal the pedestrian can cross safely. In fact, the pedestrian may let the driver pass first for various reasons; such is the complexity of this seemingly simple human-automobile interaction.

The team proposed that Nissan’s AVs should be equipped with a device that functions analogously to such cultural signaling, alerting pedestrians with a caption that lights up and flashes “I have seen you, you may cross safely.” In this case, the solution ensures that AVs are programmed to behave based on current social and cultural norms.

Standing up to a bully

My interlocutors in China thought differently. After seeing how people treat AVs during testing, they became convinced fitting-in was not enough. Yan Li (a pseudonym), a deep learning (shendu xuexi) engineer at a large Chinese automobile conglomerate developing their own line of AVs, put it pithily when we met in Guangzhou: “Humans bully AVs.”

Also using zebra crossings as an example, Yan Li explained that time and time again during testing, pedestrians crossed the road and payed scant attention to the AV. “They were so confident the AV would stop they completely tuned out.”

This worries Yan Li because outside testing areas in the real world, where traffic is greater and there are more pedestrians, the AV will get stuck. A flashing light alerting the pedestrian they can cross safely is useless. Yan Li explained, “Chinese just won’t let the AV pass. They’ll keep crossing and even loiter, because they know an AV will not harm a human. We need to break the deadlock. But how? That’s our conundrum.”

Yan Li treated the issue matter-of-factly at first; just another algorithm oddity that needed fixing. Over time, however, she came to see the deep social and cultural realities at the heart of the issue. Part of the challenge, she explained, is the sheer variety of users and behaviors on Chinese roads compared to western environments. “What do you do when a chicken crosses the road?” Yan Li asked earnestly. “This is not as uncommon as you think. Chickens move differently to a cat or a dog. We need to think about all these things.”

It’s a fair point; how many chickens roam the roads of Silicon Valley?

Sammy Wang (also a pseudonym), a senior executive at a large Chinese e-commerce company, had similar experiences and concerns. Sammy is part of a team developing autonomous delivery solutions in which an AV travels to the entrance of a residential compound and then dispatches smaller autonomous units to complete delivery to the customer’s door (these smaller units are even capable of riding up and down an elevator). Sammy explained:

During our testing, we had lots of problems. The biggest one was that people would not let the small unit pass them and often ignored it; they just stood there talking or whatever. Even if they noticed it, they didn’t let it pass. It’s a big headache for us.

Engineers were just beginning to recognize this challenge when I spoke to them.

Both Yan Li and Sammy were trying to figure out a way forward: how can AVs operate in such an environment without hurting others? Some people I met believe Yan Li’s and Sammy’s concerns are not relevant, even in China’s unique environment. These interlocutors explained that AVs will anyway be deployed in highly controlled environments as part of China’s smart-city development agenda.

But when AVs eventually share space with pedestrians and non-intelligent vehicles, as is likely, an AV that asserts itself rather than yields may be necessary. But how assertive should it be and what will an assertive AV actually look like?

Will it edge forward and nudge people with a soft yet harmless bumper bar? Will it announce “I am proceeding slowly, please disperse” and move forward? Right now, we just don’t know. It all depends on how people (mis)behave in the future? Yan Li summarizes the challenge nicely: “How can I program an AV to be assertive, yet not endanger others?”

While we’re not sure of the answer to AV technology domestication, at least now we’re asking the right questions.

Baidu has officially expanded its autonomous early rider program to citizens in the central Chinese city of Changsha as rivals accelerate their plans to carry passengers for dominance in the self-driving arena.

Why it matters: Baidu is doing more road tests in a bid to win more favor from Chinese local governments. Beijing has called for local governments to spend more on upgrading their transportation infrastructures.

• Baidu has secured several contracts for building data centers for vehicle communication networks from local governments this year.
• An official from the National Development and Reform Commission (NDRC) on Monday said “smart transportation” and “smart energy infrastructure” will be among the top priorities of the “new infrastructure” initiative to boost the economy.

Read more: China’s ‘new infrastructure’ projects, explained

Details: Citizens ranging from 18 to 65 can hail a self-driving Hongqi, a luxury sedan model from state-owned automaker FAW, with just “one-click” on Baidu’s navigation and search apps, according to our investigation.

• Free of charge for passengers, the rides are currently restricted in a geo-fenced area with limited numbers of pick-up and drop-off locations. Residential blocks, shopping centers and office parks are covered in a downtown area of around 130 square kilometers.
• China’s biggest search engine began piloting their robotaxi program with a fleet of 45 cars in late September in partnership with Changsha municipality.
• A safety driver is required behind the wheel and testing during rush hours is not allowed.
• The company has served around 700 riders with its fleet logging 150,000 kilometers (in Chinese) in the program.

Context: Baidu is not the first self-driving company allowing public to hail a robotaxi on Chinese public roads.

• Nvidia-backed WeRide completed nearly 8,400 rides with its fleet of 20 Nissan self-driving cars in a 145 square kilometer area on the outskirt of the southern Guangzhou city in a month period since late November.
• Pony.ai joined the force with Hyundai in the battle by taking 150 rides per day in Irvine, California since late last year. The Toyota-backed AV startup has been operating its robotaxi program to a limited pool of seed users in the southern Guangzhou city since December 2018.
• Ride-hailing giant Didi is catching up, as well as Alibaba-backed AutoX, with plans to roll out their autonomous ride-hailing platforms separately in Shanghai next month, persons familiar with the matter told TechNode.

Pony.ai on Thursday announced it has launched a new last-mile delivery service in California. In partnership with Yamibuy, the company’s autonomous vehicles will deliver daily essentials to customers in Irvine.

Why it matters: This is the first time for the Toyota-backed AV startup do autonomous delivery. As the pandemic keeps citizens from street shopping and public gatherings, tech companies have a chance to experiment with new technology in live commercial operations.

Details: Pony.ai on Thursday began piloting a “contactless” delivery service for customers in Irvine through a partnership with Yamibuy, a California-based e-commerce platform featuring Asian snacks and beauty products.

• Customers collect their groceries from Yamibuy on their doorsteps, delivered through its all-electric and self-driving fleet, Pony.ai said in an announcement. The AVs have a human safety driver behind the wheel.
• A Pony.ai spokesperson said all orders in the Irvine area on the Yamibuy platform will be “automatically assigned” to Pony.ai for delivery.
• The company began operating a robotaxi service with a fleet of 10 Hyundai vehicles in the city in November. It claimed it has so far completed more than 100,000 rides in several Chinese and US cities.

Context: Pony.ai is the latest AV company navigating use cases for the commercial operation of self-driving vehicles in delivery services.

• Google’s self-driving unit Waymo began delivering packages in partnership with UPS with an undisclosed size of fleet of Chrysler Pacifica minivans in the metro Phoenix area early this year.
• Alibaba-backed AutoX has been testing delivery service with two retailers in San Jose since mid-2018, reported TechCrunch.
• Pony.ai and AutoX are two of the four self-driving companies granted permit for carrying passenger with self-driving cars in California, with a human safety driver required.

China’s push to lead the world self-driving race is making another step forward: Didi Chuxing and AutoX are both about to launch their own autonomous ride-hailing pilot projects on the outskirts of Shanghai in late May, two sources familiar with the matter told TechNode on Tuesday. The projects are separate. The companies started their partnerships with the Shanghai government around the same time.

Why it matters: As Chinese local governments continue to support road testing, Didi and AutoX, among other newcomers, are attempting to elbow further into the crowded race led by Pony.ai and Baidu.

• The news comes days after Didi’s self-driving unit was reportedly closing a $300 million investment deal with its main backer SoftBank, only a month after another Didi’s investor Toyota poured $400 million in Pony.ai, the biggest funding round in Chinese autonomous driving arena.

Read more: Didi is close to $300 million deal with Softbank

Details: Ride-hailing giant Didi is planning to launch a robot ride-hailing pilot service in Shanghai as early as May and so is AutoX, two persons with direct knowledge confirmed to TechNode on Tuesday.

• “Didi is actively testing robo-taxi in Shanghai and we hope to launch the pilot service as soon as possible,” a company spokesperson told TechNode.
• AutoX on Friday announced the opening of what it claimed the Asia’s largest robotaxi operations center in Shanghai.
• Covering an area of around 750 square meter in the northwestern Jiading district, the Shanghai operation is expected to collect and process “petabytes” of driving data from road testing in Shanghai each week, AutoX said in an announcement.
• The operation will also be used for performance training simulations in a virtual traffic environment, while offering facilities to test hardware in climate conditions such as high-pressure water, and high temperature.
• The company claims that users “will soon be able to hail a ride” in its AVs, although the app is yet to be revealed. Business partners in making this operation center were also undisclosed.
• AutoX is backed by China’s biggest automakers SAIC and Dongfeng, as well as Alibaba.

Context: Shanghai government in September issued China’s first licenses for passenger-carrying self-driving cars in an area of 65 square kilometers to Volkswagen’s partner SAIC, BMW and Didi, followed by AutoX in December.

• Little progress has been revealed since then, as the official greenlight has not been given for testing, according to people close to the matter.
• Didi and AutoX late last year revealed plans to roll out 30 and 100 robotaxis in Shanghai as early as year-end, or early in 2020, respectively, as part of their efforts to lead the driverless mobility future.

Baidu’s Apollo autonomous driving program has thrust the search giant into the spotlight. Named after NASA’s moon missions, the self-driving program recently enjoyed a series of wins when Baidu came out on top in annual self-driving reports released by authorities in California and Beijing.

Waymo has said the reports do not provide “relevant insights” or distinguish their company’s “performance from others in the self-driving space.” Kyle Vogt, the CTO of General Motors-backed Cruise, shared similar sentiments. “The idea that disengagements give a meaningful signal about whether an AV is ready for commercial deployment is a myth,” he wrote in a blog post.

Still, much is expected of Baidu’s self-driving efforts. The company has launched autonomous ride-hailing services in Changsha, the capital of Hunan province, as well as in Cangzhou, in north China’s Hebei province, with a fleet of 30 cars. Baidu’s autonomous driving tests have covered more than 3 million kilometers on public roads across 23 Chinese cities.

Where are Baidu’s self-driving cars headed in 2020? What is the outlook for Baidu in autonomous ride-hailing? We will start with our recent experience in a Baidu robotaxi in Changsha and move on from there.

A ride in a Baidu robotaxi

Robotaxis are all the rage. Around the world, startup and tech giants alike are fighting the war for self-driving supremacy, and autonomous taxis have become the new battleground.

Companies including Baidu, Pony.ai, and WeRide have launched robotaxi pilots across China. Baidu, the country’s designated self-driving champion, began offering its robotaxi service in Changsha last September.

Three months later, TechNode arrived in downtown Changsha. Standing outside a well-known culture and arts center on a sun-washed December afternoon, we waited for a Baidu self-driving taxi to pull up.

The trip showed us how companies are taking vastly different approaches to developing their self-driving technologies, and just how difficult it is to create global benchmarks detailing how these vehicles should perform.

Baidu runs its autonomous taxis in and around Changsha’s downtown Xiangjiang New Area. The trial operation is more of a geo-fenced test on public roads; passengers can pick one of three fixed five-kilometer routes, all starting from the city’s grand theater.

The tech giant has partnered with Chinese state-owned automaker FAW Group, which provides the vehicle for its autonomous system. As the luxury Hongqi model arrived to pick us up on that balmy December afternoon, we quickly took one photo before we were told that pictures were not allowed.

Shortly after we got into the car and entered Changsha traffic, Baidu’s approach to its self-driving program became evident. It was like going for a ride with a nervous student driver.

Companies that develop self-driving technology need to consider not only the safety of their passengers but also the comfort of the ride. Baidu places more emphasis on safety than we had expected, resulting in a trip that was less smooth than AV rides we’d experienced from companies that squeeze more efforts to the comfort of their passengers.

“Our top priority is zero accidents on the road,” our vehicle’s safety driver said while we waited at a traffic light. He offered a glimpse into how the company’s safety precautions are meant to protect the trial project from any sort of controversy. “All of us are required to take a 10-minute break for each hour of work,” the driver told us.

During our trip, Baidu’s robotaxi traveled at speeds of around 30 kilometers per hour and stopped by itself every now and then to yield to pedestrians. Traffic was heavy, with cars filling the six-lane Meixi Lake Road, downtown Changsha’s main avenue.

When the vehicle stopped at a red light in the middle of an intersection, we got to see firsthand the safety precautions that our driver had described: After a few minutes of waiting, the human driver had to take over. Situations like these are typically evaluated as “too risky” for the autonomous system to navigate. Baidu says it has reported “zero accidents” in the past few years because of its “safety-first” approach.

The company has requested that its fleet of dozens of vehicles in Changsha log a certain amount of mileage each day, our safety driver told TechNode, without revealing any further details. Meanwhile, working hours are very limited since the company has not been allowed to test during rush hour. Therefore, overtime work during weekends has become common.

A conservative driving strategy

Baidu is taking a more conservative approach to its AV road testing, emphasizing safety over comfort, a self-driving car engineer said, commenting on TechNode’s observations of our robotaxi ride.

Slower driving speeds, hesitation when turning or changing lanes, and constant stops when facing dangerous scenarios are among the passive driving strategies that result, the engineer said, who asked not to be named because he was not permitted to speak to the media.

A focus on safety, alongside a goal of fewer human interventions, can be achieved by developing a cautious algorithm, helped by some of the high-performance hardware that acts as the eyes of self-driving vehicles.

For years, safety and comfort have been among the top priorities for robotaxi companies offering driverless experiences. “No doubt safety is the key to getting autonomous cars on the roads,” but a better solution could be a wider range of driving styles with safety guarantees to ensure more comfort for passengers, the engineer said. There should have been some “more decisive driving policies” he said, referring to how the vehicle could have taken proactive measures to avoid dangerous situations, such as changing lanes.

Key metrics on AV testing

Baidu’s prudence could be part of the reason the company came out on top in the recent self-driving report released by California’s Department of Motor Vehicles.

Baidu beat Google’s self-driving unit Waymo by reporting the least number of disengagements among all companies operating such vehicles in the state. A disengagement is defined as any time a human driver is required to take over from an autonomous system during self-driving tests.

But within the industry, questions over the relevance of such metrics are on the rise, with experts saying that the measure has limits when trying to gauge whether a company’s technologies are ready to be deployed commercially.

AV companies themselves have also highlighted the report’s limited usefulness. In an announcement, Baidu said disengagement is more of an internal reflection of the speed of technical iterations, and therefore comparison between companies is “not that meaningful.”

However, if disengagement rates offer few relevant insights into the technology, what are the measurable metrics that could indicate progress? Two experts that TechNode spoke with gave the same answer: the variety and complexity of testing scenarios in which a robocar can operate.

Keeping within a lane in urban traffic, recognizing traffic signals, or turning left at an intersection without a “green arrow” traffic signal are some of the most typical and frequently seen scenarios identified and tested by AV players.

However, the real difficulty is to get autonomous cars to operate under “edge cases,” or unusual circumstances, such as a nearby vehicle changing lanes abruptly, a motorcycle coming out of nowhere, or drunk driving behavior from other road users.

These scenarios could be used to create a benchmark dataset that enables companies to train and evaluate their algorithms and compare accuracy rates to effectively evaluate their technologies, much like ImageNet, a renowned computer vision dataset of more than 14 million photographs widely used to evaluate the performance of AI systems.

“The more driving scenarios your cars can handle, the more you can prove the safety of the technology,” said one of the experts. Nevertheless, problems persist because the industry has not reached a consensus on standards.

The self-driving industry has now evolved from being driven by research and development of AV technologies to being mostly pushed forward by testing efforts. The development of key technologies, such as environment perception and car control, have mostly been completed; the priority now is to gain experience in as many driving cases as possible and learn how to deal with them, the experts added.

Every new experience helps a self-driving car to learn, and that’s where some of the world’s AV leaders are ramping up their efforts. Last year, Cruise almost doubled its testing and validation miles from the year prior, and “every mile Cruise tested in California was driven in the very complex urban environment of San Francisco,” it said in its individual filing.

The company, which is mainly backed by General Motors, operates a fleet of 228 vehicles that drove more than 831,000 miles last year, nearly eight times that of Baidu. As of last December, the Chinese search giant claimed its vehicles had traveled a total of more than 3 million kilometers (1.86 million miles).

But wider tests in China are coming as more local governments join in the race to open their roads to robotaxi companies, allowing them to collect more data and develop better evaluation methods. We’ll have to wait and see who comes out in pole position.

Self-driving startup Qcraft on Friday announced it has closed a round of seed funding running into “dozens of millions of US dollars” led by investment firm IDG Capital.

Why it matters: The deal marks another round of investment fervor around Chinese autonomous vehicle (AV) companies, at a time when municipal governments are nurturing emerging technologies to shore up economic growth.

• The Politburo Standing Committee, China’s top decision-making body, in March urged the acceleration of “new infrastructure” construction, an investment initiative focusing on 5G, data centers, and artificial intelligence, according to a China Central Television report.
• Baidu, China’s Google, last month secured contracts from three local governments including southwestern municipality Chongqing for building cloud-based transport infrastructure such as data centers and fleet management systems to enhance AV testing on public roads.

Details: Qcraft has raised an undisclosed amount of funding in a seed round from a list of investment companies including IDG Capital and Vision Plus Capital, a Hangzhou-based venture capital firm formed by Eddie Wu, an Alibaba co-founder. The startup was founded in Silicon Valley and operates both in Beijing and California.

• The AV startup, formed by four former Waymo engineers in March 2019, is developing AV simulation testing, which refers to a system solution that trains self-driving cars to deal with vast traffic scenarios, especially unusual cases, on a virtual road network.
• Simulated testing is considered a safer, more scalable and cost-effective way to get self-driving cars on the roads. Google’s AV unit Waymo in November said it has logged more than 10 billion virtual miles with its simulation software, around 500 times that of actual miles driven on public roads.
• Still, it requires collecting sufficient real-world data to develop an simulation system and effectively train vehicles. Co-founder Wang Kun told TechNode on Friday that the company is currently testing around 10 cars with Level 4 autonomy in China and the US, and hinted at more collaborations with industry players.
• Wang spent three years at Waymo’s simulation team as a software engineer, before forming the company with CEO Yu Qian, a former tech lead at Google Map and Waymo, as well as Hou Cong and Da Fang, two engineers formerly with Waymo’s perception and motion planning teams, respectively.

Context: Qcraft is one of the several AV startups that has recently won a new war chest.

• Space Tech, a Shanghai-based AV company, in Spetember announced an angel round of funding worth “dozens of millions of RMB” from undisclosed investors, reported Chinese media. Founder Ma Guanglin was a former senior manager of assisted driving system development at Tier-1 supplier Delphi.
• The recent self-driving investments boosted confidence of worldwide investors. Pony.ai announced a $400 million funding round led by Toyota just a week before Waymo’s $2.25 billion deal, its first funding from external investors.

Robotruck startup TuSimple has partnered with German auto supplier ZF to develop and commercialize technology for autonomous trucks.

Why it matters: TuSimple aims to begin testing truly driverless trucks—those without safety personnel on board—by 2021.

• ZF is one of the largest automotive suppliers in the world and has made a significant push into autonomous driving.
• TuSimple has tested its technology on a stretch of highway between Tucson and Phoenix, Ariz. since 2018. The company also won China’s first permit to trial driverless trucks in Shanghai.

Details: TuSimple and ZF will work together to develop onboard computers and sensors such as radar and lidar for autonomous trucks, according to a statement released Thursday. The German supplier will also become TuSimple’s “default supplier” when commercializing robotrucks.

• ZF will provide engineering support to integrate and validate TuSimple’s autonomous driving platform into the trucks, the companies said.
• The partnership is an “important milestone,” according to Chuck Price, TuSimple’s chief product officer. He said that the deal allows the companies to scale the technology toward mass-produced autonomous trucks.
• Meanwhile, ZF said that it expects the two companies to create the first commercialized automotive-grade autonomous system for trucks.
• The deal comes shortly after TuSimple expanded a tie-up with UPS, effectively doubling the number of autonomous delivery runs the company makes for the American logistics company.
• TuSimple currently has 18 contracted customers and makes around 20 autonomous trips per day.

Context: Autonomous trucks are expected to reach commercialization before passenger vehicles, presenting huge potential for growth. TuSimple aims to transform America’s $800 billion trucking industry with autonomous rigs.

• The company was founded in the US in 2015 by Hou Xiaodi and splits its operations between the US and China, with offices in Beijing and Shanghai as well as in Japan.
• TuSimple, which is developing Level 4 autonomous long-haul trucks, secured $120 million in an extended Series D in September.

For the first time in history, a Chinese company has taken the top spot among firms testing autonomous vehicles on California public roads.

In 2019, Baidu drove 108,300 miles and reported six disengagements across its four vehicles, making for the lowest disengagement rate of all the companies listed in California’s annual report: 0.055 per 1,000 self-driven miles.

Baidu had drastically improved its performance over last year’s report, In 2018, the company reported one disengagement every 205 miles. This year, that number fell to one for every 18,050 miles. In doing so, the company managed to knock Waymo out of its top-ranked position. Baidu attributed the drop to rapid expansion in testing fields over the past three years. 

But as the industry matures, disengagements are increasingly being seen as a poor measure of performance, since road and weather conditions, which play a huge role in report results, are not included in the data. Meanwhile, Baidu’s reports contain significantly less information about disengagements than its peers, causing industry insiders to raise questions about the quality of the company’s tests.

Baidu’s human intervention

According to Baidu’s report, the company’s vehicles required human intervention in certain situations: when surrounding objects were not detected or were misclassified, when a decision made by the autonomous system was not appropriate to the scenario, or when there was a problem with the hardware.

However, Baidu does not provide any additional information about the situation under which these disengagements occurred, only broad categories. Meanwhile, several of its rivals’ reports provide more detail about each incident that resulted in a disengagement.

For example, where Chinese counterpart Pony.ai said of one disengagement: “Driver precautionarily intervened for a reckless neighboring vehicle cutting into vehicle’s lane,” Baidu would simply say “perception discrepancy,” making it difficult to gauge just how well the company’s AV system functions.

To be fair, self-driving startup WeRide also lacked detailed descriptions in its reports. These companies are not required to include comprehensive accounts of every disengagement. However, many well-established players do, including Cruise, Didi, and Zoox.

Other aspects of the company’s testing regime are also absent. The company does not mention in its report where the tests took place. Most other companies’ reports indicate where they are testing and whether they have expanded their operations in California.

Baidu is predominantly running its AVs in Sunnyvale in very simple traffic scenarios, two industry insiders told TechNode, who asked not to be named due to their proximity to the matter.

By contrast, General Motors-owned Cruise conducted all of its tests on urban roads in San Francisco, the third-most congested city in the US, according to Tomtom’s 2019 Traffic Index. Cruise reported a disengagement rate of 0.082 per 1000 miles.

A Baidu spokesperson told TechNode that the company tests in “diverse conditions,” including urban roads and scenarios involving pedestrian avoidance, left and right turns, lane changes, and traffic light recognition.

Road conditions can have a profound effect on disengagements, with more complex urban roads leading to more disengagements. Conversely, highway driving is typically seen as easy for AVs.

“If I wanted to look even better, I’d do a ton of easy freeway miles in California and do my real testing anywhere else,” Bryant Walker Smith, a self-driving car expert, told The Verge.

China’s major players

While Baidu took the top spot in the tests, four Chinese AV startups also made it into the Top 10. AutoX and Pony.ai came in fourth and fifth—right behind GM’s Cruise—with one disengagement every 10,684 and 6,475 miles, respectively.

Meanwhile, Didi Chuxing took the eighth position, reporting 1,535 miles per disengagement, a good result for a relative newcomer. Didi, China’s biggest ride-hailing platform, started testing in California in June 2018.

In addition, China- and California-based WeRide recorded 151.7 miles driven per disengagement, performing much worse than its Chinese peers but ranking higher than companies such as Apple, Mercedes Benz, and Toyota.

Most Chinese companies conducting tests in California revealed no further details about their operations when contacted by TechNode. However, their individual reports reveal a blurred glimpse into their performance.

Pony.ai, AutoX, and WeRide all claimed to have covered a big pool of testing scenarios in various traffic and weather conditions—either sunny days or heavy rain. However, none of them detailed when and where exactly a driver has to disengage the system. These companies gave no indication of whether these incidents occurred in downtown traffic during commutes or on empty highways at night.

In terms of test areas, all the four companies have vehicles being tested in the South Bay, while Pony.ai further expanded to Fremont, where it launched a pilot robotaxi program providing transport services from a train station to two government offices.

However, most of the areas have modest population density, around one-quarter of that of San Francisco, where GM Cruise tested its vehicles in the city’s “very complex urban environment.”

Among the four Chinese companies, Pony.ai reported that its vehicles covered the greatest distance. Its fleet of 22 vehicles logged 174,845 miles in California, the third-largest number in the ranking, although nearly a fifth of that of Cruise.

The Toyota-backed AV startup also detailed their disengagements in more detail than its Chinese counterparts. In the 27 disengagements recorded over the 12 months ending in November 2019, Pony.ai attributed eight of them to reckless driving by other vehicles, and 11 to suboptimal routes planned by software for the car to maneuver. The situations its vehicles encountered vary from insufficient yielding to reckless driving on the part of other road users.

A flawed system?

Other AV companies reported disengagements resulting from poor detection of road objects or mapping flaws in different traffic scenarios. Although such details were presented in Didi’s reports, the ride-hailing giant revealed few reasons for disengagements, not categorizing them as planning, mapping or control issues.

Alibaba-backed AutoX referred very generally to the company’s three human intervention cases as localization and planning problems. The low number of disengagements may result from fewer miles driven than other companies. Meanwhile, Nvidia-backed WeRide reduced its miles driven by nearly two-thirds in 2019 from the year before, making little progress compared to last year.

The furor over the reports has led an increasing number of experts in the field to call into question the effectiveness of using disengagements as a metric to gauge how a vehicle is able to drive autonomously.

Disengagement reports provide an opportunity to compare AV performance between companies but discrepancies in reporting make the metric insufficient to measure performance, experts say.

In a series of tweets last month, Waymo asked whether disengagement metrics lead to meaningful insights. The company added that most of its real-world driving experience comes from outside California.

Meanwhile, Cruise Co-founder Kyle Vogt shared similar views, saying in a blog post that the reports are “woefully inadequate” to judge whether an AV is ready to be deployed commercially.

An earlier version of this article quoted a TechNode source as saying that Baidu tests AVs on Bay Area interstate highways. In fact, the company denies that its vehicles have been tested on interstate highways, and a review of interview recordings suggested that we may have misunderstood our sources’ comments.

Chinese mobility service provider Didi Chuxing has reportedly been in talks with Softbank for $300 million in fresh funding for its autonomous driving unit.

Why it matters: The investment is a vote of confidence in a Chinese AV startup during a low point in investment activity compounded by the Covid-19 outbreak.

• China’s deal-making activity for startups reached its lowest point, shrinking by more than half year on year to 168 deals worth $1.79 billion as of February, according to the South China Morning Post citing figures from financial research firm Preqin.
• Expectations about widespread AV adoption has flagged in recent years, as the development of the technology proves to be more difficult and time-consuming as initially thought.

Details: Softbank is expanding its commitment to Didi and is on the brink of reaching a deal to lead a $300 million investment into the ride-hailing startup’s self-driving unit for an undisclosed valuation, The Information first reported Monday citing people with knowledge of the situation. TechNode verified Softbank’s investment in Didi with a person close to the matter on Tuesday.

• The other investors involved in the deal are unknown. SoftBank did not respond to a request for comment and Didi declined to comment.
• Rumors of Didi seeking funds for its AV unit have been circulating since July, just a month before the ride-hailing platform spun off its self-driving car department to transfer some of the considerable cost to external investors.
• Didi has been playing catch-up in the global self-driving race. It began testing robocars in California in June 2018, two years after Baidu and lagging Pony.ai by a year. It reported one disengagement from a human driver every 1,535 miles on California public roads last year, a decent result for a first-timer.
• Industry insiders TechNode spoke with view the company as having significant potential to succeed in light of the huge volumes of human driving and public transport data it is able to feed into its algorithm. Didi late last year unveiled plans to launch a robotaxi pilot service in Shanghai, though there has been no progress since then due to regulatory hurdles.
• Softbank has invested aggressively in AV. It poured $2.25 billion into General Motors-backed Cruise for a 20% stake in May 2018, making it the largest deal at the time for the nascent industry.
• This was followed by a $1 billion deal a year later from the Japanese investment giant and other investors including Toyota with Uber’s autonomous driving team, which also became an independent business soon after the investment was unveiled.

Context: Softbank has had a rough past several months. It has been sharply criticized over its once-hyped investment strategy, following the downfall of two of its biggest rising stars, WeWork and OYO, which face falling revenues and plunging valuations.

• The Japanese tech investor wrote down its Wework investment by $3.4 billion late last year, and reported its first quarterly loss in 14 years during the quarter ended Sept. 30, totaling $6.5 billion. It announced plans on Monday to sell $41 billion in assets to buy back shares and reduce debts.

China’s biggest internet search company Baidu has won a bid to build public road infrastructure for self-driving cars in southwestern Chongqing municipality, a deal worth $7.5 million.

Why it matters: Baidu is expanding from developing autonomous vehicle technology to offering cloud-based transport infrastructure for car connectivity amid rising 5G adoption in China.

• China in 2011 began researching vehicle-to-everything (V2X) technology that links cars, transport facilities, and other road agents through a carrier network. It ramped up efforts with the launch of what it said was the world’s largest V2X city network in the eastern city of Wuxi in late 2017.
• Baidu open-sourced its V2X solutions a year later. It then set up a standalone V2X department late last year in response to Beijing’s call to close the gap with world leaders in the self-driving race.

Details: Yongchuan district in Chongqing has offered a RMB 52.8 million ($7.5 million) contract to Baidu to develop cloud data centers for self-driving car testing on city roads, the government said in an announcement released Tuesday (in Chinese).

• Baidu will provide a package of solutions including cloud data centers for vehicle-infrastructure communication and car management, enabling Level 4 autonomous vehicles to test on a 20 square kilometer (around 7.7 square mile) area of public roads.
• The contract also covers deployment of edge servers and signal control systems on roads for detecting objects and transmitting data. The deal was part of a larger RMB 1 billion framework deal struck between Baidu and the district government for an AV test infrastructure project last March.
• In an announcement released on Friday, Baidu said a fleet of more than 100 self-driving cars could drive on the roads to validate concepts of operations and technologies after construction.
• In April, the Chinese tech giant became one of the first seven companies to win permits for AV tests by Chongqing’s government, alongside state-owned automakers including Changan and Dongfeng.

Context: Baidu has reached partnerships with more than a dozen Chinese governments over the past few years. Some of the biggest deals were those with Beijing and Changsha, to monetize its futuristic AV technologies.

• The company inked a framework agreement with Yinchuan, the capital of western Ningxia province, in late December to enable self-driving rigs tested with V2X solutions following a similar deal with the government of the northern Cangzhou city two months ago.
• “It is important for us to gain operational experience as well as derive commercial value from areas, such as smart transportation to ensure that we are meeting market needs with our technology,” Baidu CEO Robin Li said during the fourth quarter earnings call.

China’s Changan Automobile on Tuesday unveiled its new flagship sedan with what it said was the country’s first mass-manufactured conditionally automated Level 3 system, as Chinese automakers ramp up to compete in the global self-driving race.

Why it matters: The development could be a prelude to mass deployment of highly automated cars on Chinese roads, but regulatory and technological hurdles remain.

• Level 3 autonomy conditionally allows drivers to take their eyes off the road and hands off the wheel at low speeds to focus on things other than driving, according to a rating from the Society of Automotive Engineers (SAE). Human intervention is still required.
• Global automakers including Volvo and Ford have argued that the handover of vehicle control is unsafe if human driver vigilance lapses. Most companies had delayed plans to volume-produce L3 vehicles, Chinese media reported last year citing a Baidu executive.

Details: Chongqing-based Changan on Tuesday announced it has developed China’s first mass-production automobile to offer Level 3 autonomy under conditions including highway driving and traffic congestion. The sedan will go on sale in June.

• Named UNI-T, the four-door sedan features an in-house developed computer that can process live feeds from 12 ultrasonic radars, six cameras, and five millimeter-wave radars to detect other road agents and plan its path.
• Traveling at a top speed of 40 kilometers or 25 miles per hour, a driver can take his hands, feet, and eyes off of driving controls for a “long time” on congested highways and freeways when, for example, using a phone or for other in-vehicle entertainment, the company said in an announcement.
• Changan did not immediately respond to requests for clarity on the length a time a driver could safely relinquish driving control to the vehicle.
• The self-driving system can pilot the vehicle when it exceeds 40 kilometers per hour on highways, but the human driver must be attentive. A human driver is required to intervene immediately when alerted, otherwise the car will move to safe mode, which may mean pulling over, the company said.
• Changan said its self-driving vehicles have logged more than 50 million kilometers on public roads in a number of Chinese cities including the southeastern municipality of Chongqing and Beijing.
• It obtained a permit to test self-driving cars on public roads in California from the Department of Motor Vehicles in late 2017, but had not conducted tests in the state as of last year, according to the government agency’s disengagement filings.
• State-owned Changan is one of the entities which lead the drafting of Chinese rules and regulations for driving robocars on roads, which is expected to roll out within the year, Caixin reported citing company president Zhu Huarong.

Context: German automaker Audi unveiled in late 2017 the world’s first production vehicle with Level 3 autonomy, a new A8 luxury sedan, but the model will primarily be available domestically over the next several years, a result of strict regulations and consumer concern over the safety of autonomous cars.

Driverless delivery startup Neolix has raised nearly RMB 200 million ($28.7 million) in Series A+ funding to mass-produce its self-driving vehicles, the company said on Wednesday.

Why it matters: China eased restrictions on delivery robots following the Covid-19 outbreak, resulting in a surge in demand for autonomous deliveries in some of the worst-hit areas.

• The flu-like virus has offered these firms an unprecedented opportunity to put their technology through its paces as Beijing promoted “contactless deliveries” to curb the spread of the disease.
• Neolix began mass production of its Level 4 delivery robots last year, counting companies including Huawei, JD.com, and Alibaba as customers.

Details: Neolix’s latest round of funding, which it closed in February and announced this week, is led by electric vehicle maker and existing investor Lixiang. The company is now Neolix’s second-largest shareholder after CEO Yu Enyuan.

• Meanwhile, Addor Capital, as well as previous investors Glory Ventures and Yunqi Partners, also took part in the round.
• The company plans to use the proceeds to ramp up production and operations of its delivery robots, with the new round coming less than a year after Neolix’s RMB 100 Series A.
• The company said in a statement that it expects to close another new investment in 2020.
• Neolix has a production facility in the eastern Chinese city of Changzhou and aims to sell 100,000 vehicles a year in the next four years, according to Bloomberg.
• The company’s primary focus is food delivery, mobile retail, and security applications—in cooperation with local governments.
• Neolix has produced 225 delivery robots in the two years since it was founded, the company said.

Context: While Covid-19 has thrown delivery robots into the spotlight, the technology still faces technical and regulatory hurdles.

• The Chinese government has allowed lifestyle services platform Meituan and e-commerce giant JD.com to run their autonomous delivery services, but many companies are still only permitted to operate within geofenced areas.
• The unpredictable nature of traffic and pedestrians, especially on small, congested roads, present significant challenges to wider adoption.
• Nevertheless, Alibaba founder Jack Ma predicts that within a decade more than 1 billion packages could be delivered by driverless vehicles daily in China.

As China ramped up its efforts to counter the spread of Covid-19, delivery robots have garnered newfound attention.

Beijing has since pledged to increase its support to upgrade the nation’s freight delivery systems. The government also asked companies for solutions to contain the virus, including various forms of “contactless shopping deliveries,” as people around the country became afraid to leave their homes.

At this moment of crisis, some businesses saw opportunities for largely unproven technologies. In an effort to protect the public, lifestyle services giant Meituan and e-commerce firm JD.com started using their unmanned delivery technologies in some of the worst-hit areas.

Just 60% of deliverymen have returned to work in Wuhan since authorities cut the city off from the world. The remainder have been unable to re-enter the city since the lockdown began. Worse still, those in Wuhan have been under both physical and mental pressure from the burgeoning workload and concerns over the epidemic.

With drivers locked in and locked down, the companies had no choice but to experiment with the new tech. 

JD’s self-driving robot made its first delivery of medical supplies to Wuhan’s Ninth Hospital on Feb. 6. The facility, designated for treating seriously ill patients, is just 600 meters from a JD distribution center. The close proximity put delivery people at risk of infection, Zhou Jianbin, a district manager of JD Logistics in Wuhan, told The Paper.

The majority of deliveries in Hubei include masks, protective clothing, and other medical supplies. However, the process is not completely automated. JD employees need to place orders in the cars before the deliveries begin. Typically, the robots will alert a user that their delivery is ready for collection and wait 30 minutes for them to collect the goods.

The robots are responsible for half of all daily deliveries, around 10-20 orders each day, according to Zhou. Although only two robots are currently being deployed in the city, JD said it is gradually making a shift to serve the nearly Ninth Hospital with fully driverless delivery.

Due to a significant spike in demand for unmanned deliveries in Wuhan and surrounding cities, the commercial launch of JD’s robot delivery service came well ahead of schedule, said Qi Kong, head of autonomous driving and JD Logistics. The e-commerce giant had initially planned to start mass-producing its driverless vehicles by the end of the year, but now expects to roll out more than 50 robots by the end of April.

A week after JD debuted its robots in Wuhan, Beijing-based Meituan began piloting two driverless delivery robots in the city’s northeastern Shunyi district. Running at just 20 kilometers (12 miles) per hour, the pint-sized vehicles deliver groceries to residents of three neighborhoods within a five-kilometer radius of its pickup station. Each robot delivers up to five orders per trip.

The company did not specify how many orders its autonomous fleet delivers per day. According to Meituan, the robots work as an alternative form of last-mile delivery to help alleviate the shortage of delivery drivers. 

The company is also piloting robots at restaurants in Beijing that bring food from kitchens to deliverymen or customers waiting for takeaway meals, in an effort to limit contact between people. The company claims that these robots are not “replacing humans entirely,” as the service currently still requires human-robot collaboration. 

Technical and regulatory hurdles

While the Covid-19 has offered unmanned delivery providers both government support and an unprecedented opportunity to put their technology through its paces, these companies have had trouble driving adoption of autonomous delivery systems, as regulatory and technological hurdles do still present significant roadblocks to companies such as Meituan and JD.

Regulations governing autonomous driving have long frustrated automakers and tech companies, but the situation is even stickier for unmanned delivery services in China.

To begin with, there is no space on roads dedicated specifically for delivery robots, Zhao Bin, head of public affairs at JD Logistics, told Chinese media in February. Before JD launched its Wuhan Ninth Hospital robot delivery service during the outbreak, the Chinese e-commerce giant had to get hasty approval from government agencies to survey the roads and get maps drawn.

Current Chinese laws are not well-equipped to govern self-driving vehicles, which are not legally allowed to drive on public roads. Various pilot programs are able to operate only because the government issues temporary license plates to approved self-driving companies. Without this permission, the use of these vehicles is illegal and companies must bear all liability for accidents.

The Chinese government has given JD and Meituan permission to run robot deliveries, but many more companies can only run their services in geo-fenced areas such as office parks and school campuses.

Meanwhile, other firms are unable to even get their plans off the ground. According to Chinese media reports, one anonymous self-driving company initially planned to use low-speed driverless vehicles to transport meals from a restaurant in Beijing to a nearby hospital for doctors and patients, but the company eventually had to backtrack on its plans.

Even Baidu, the poster child of China’s self-driving ambitions, only gained lackluster support during the outbreak, deploying just two robots for sterilizing the campuses of two colleges in Wuhan, alongside dozens of others in Shanghai, Shenzhen, and Guangzhou. The company claimed one of its invested startups began delivering meals to medical staff in Beijing Haidian Hospital starting Feb. 14.

The industry also faces technological challenges. These vehicles currently face enormous limits in their abilities to operate under certain road and weather conditions. The unpredictable nature of traffic and pedestrians, especially when these vehicles attempt to navigate congested roads within residential communities, present significant challenges to wider adoption. A lack of road markings and bad weather further compound these difficulties.

As Bob Zhang, CTO and co-founder of ride-hailing company Didi, has previously made clear, self-driving technology has a long way to go before it can navigate a wide range of weather conditions safety.

Propelled by machine-learning algorithms and a package of hardware that includes various sensing technologies, a delivery service robot can be quite expensive, with prices starting at RMB 100,000 ($14,220). Fortunately, the cost has declined significantly over the past several years; in the early years of development, JD said in 2017, the outlay (in Chinese) could be as much as RMB 600,000 per robot.

This price tag contrasts sharply with the pay of delivery workers, which ranges from RMB 5,000 to RMB 8,000 per month, according to public information on Chinese job recruiting platforms.

Prospects

Covid-19 has revealed the potential value that autonomous deliveries can play in emergency situations. As Chinese citizens avoided infection by engaging in voluntary isolation, legions of food and grocery delivery drivers became a lifeline, providing a fresh supply of food to millions around the country.

However, there were limits. Many migrant delivery workers had made their yearly trek across the country to their hometowns, leading to a dearth of drivers in major urban centers. With fewer drivers available, deliveries that usually took 30 minutes might now be completed in around two hours.

Costs also increased. In Shanghai, for example, Alibaba’s Hema supermarket charged an additional RMB 6 for deliveries that had previously been free of charge.

The coronavirus outbreak also led to fears over close contact with delivery drivers, who had the potential to unknowingly spread infection to an untold number of other people. In response, companies launched “contactless delivery,” in which orders were left at the entrance of apartment complexes. The model had already been in use at office buildings before the outbreak, but quickly became ubiquitous as the outbreak continued.

In Hubei, the center of the epidemic, the government placed restrictions on deliveries to limit people’s exposure to the disease. Residents in small towns had to contact their party committees to get fresh food and supplies.

Delivery robots could provide a solution to these problems, and are poised to play an important role in China’s logistics industry. In less than a decade, autonomous vehicles will deliver 80% of all goods, according to the research firm McKinsey. These vehicles could increase efficiency and cut expenses in an industry where last-mile deliveries can constitute up to 12% of costs.

Xia Huaxia, Meituan’s chief scientist, told TechNode last year that machines can also be used to complement the work of delivery people by taking night shifts or working during extreme weather conditions. If a delivery robot’s lifespan is more than three years, he said, the cost of the machine will be lower than the cost of human labor.

Observers expect China’s food-delivery market to explode in the next few years. Meituan, which employed 600,000 drivers as of late last year, predicts that its daily orders will increase by 200% per day. According to Xia, in the second half of 2019, the delivery giant completed 25 million orders every day.

Autonomous truck startup TuSimple has expanded its partnership with UPS, doubling the number of delivery runs its vehicles make for the American logistics company per week.

Why it matters: The extended alliance between the two companies is a vote of confidence for TuSimple, which aims to transform the country’s $800 billion trucking industry with fully autonomous rigs.

• UPS’s venture capital arm in August invested an undisclosed amount in the trucking company.
• TuSimple has been testing its technology on a stretch of highway between Tucson and Phoenix, Ariz. since 2018. The company also won China’s first permit to trial driverless trucks in Shanghai.

Details: TuSimple will increase the number of trips it makes for UPS to 20 runs per week, adding an additional 10 trips on a new route between Phoenix and El Paso, Texas, the company said in a statement on Thursday.

• TuSimple will continue running 10 trips a week on an existing route between Tucson and Phoenix.
• At the same time, the company said it had reduced fuel costs in its operations with UPS by 10%. Autonomous driving technologies are often touted as being more energy-efficient than human drivers.
• UPS is using its partnership with TuSimple to explore ways in which autonomous technology can improve efficiency, safety, and customer service, Scott Price, chief strategy and transformation officer at UPS, said in a statement.
• TuSimple currently has 18 contracted customers and makes around 20 autonomous trips per day.

Context: The logistics industry could see increased efficiency by using autonomous trucks as more people do their shopping online, putting increased strain on freight companies.

• The US was short 60,000 truck drivers at the end of 2018, according to estimates by the American Trucking Association, which expects that figure to triple by 2028.
• TuSimple was founded in the US in 2015 by Hou Xiaodi and splits its operations between the US and China, with offices in Beijing and Shanghai as well as in Japan.

A total of 77 self-driving cars have driven more than 1 million kilometers on public roads in Beijing and search giant Baidu accounts for the lion’s share, regulators of the China’s capital city said in a report released on Monday.

Why it matters: Beijing’s self-driving report is the only one of its kind made public and recognized by the Chinese authorities, although self-driving tests are conducted in a number of cities including Shanghai and Guangzhou.

• Meanwhile, after two years of testing, Beijing provided in the report the number of “disengagements” or human intervention events during self-driving car tests.

Details: Baidu’s autonomous vehicles have traveled more than 893,900 kilometers (555,500 miles) in the city over a two-year period as of December, Beijing’s Innovation Center for Mobility Intelligent (BICMI), the city’s official service agency for AV tests, said Monday in a report (in Chinese).

• Baidu’s 45 cars drove around 140,000 kilometers in 2018. The company expanded the fleet size to 52 self-driving cars with more than 754,000 kilometers logged last year.
• Guangzhou-based Pony.ai came second with around 121,300 kilometers of tests as of last year. The number of miles driven by its five vehicles in 2019 increased more than tenfold from just over 10,000 kilometers a year earlier.
• In total, 77 vehicles from 13 companies drove more than 1.04 million kilometers (around 646,400 miles) on Beijing’s public roads as of last year. Japan’s biggest automaker Toyota jumped to the third place with its four cars travelling 11,129 kilometers in 2019. It conducted no tests the prior year.
• Chinese auto and tech companies including Tencent, Didi Chuxing, and Nio were among the top 10, as well as Audi and Daimler. None of the companies beyond the top three exceeded 10,000 kilometers.
• This year’s report revealed details about disengagements: only 14% were attributed to mechanical system or algorithm shortcomings in events such as broken traffic lights or reckless driving by other cars.
• All other disengagements were for reasons such as replacing data-recording equipment, changes in planned routes, or for human break times, the report said.
• A BICMI representative declined to reveal further disengagement details when contacted by TechNode on Tuesday.

Context: The Beijing government released China’s first municipal-level regulations for AV road tests in December 2017. It has opened a total of 503.7 kilometers of roads in four districts in the outskirts of the city as of 2019, more than triple the size a year earlier.

• City officials in December began to allow passenger transport in self-driving cars, and Beijing-based Baidu was the first to receive permission from the city for its 40 licensed AVs. The actual size of the company’s self-driving car fleet is unknown, but it has driven 3 million kilometers in 23 Chinese cities as of December, according to Baidu’s Q4 earnings results.
• However, Baidu is only piloting passenger transport in the central Chinese city of Changsha, and has not disclosed user base or daily rides details.
• Meanwhile Toyota-backed Pony.ai said its fleet of 100 vehicles has traveled more than 1.5 million kilometers in four cities. Its ride-hailing pilot project surpassed 70,000 orders as of the end of last year.

Chinese search engine giant Baidu reported the lowest rate of human driver intervention among companies testing autonomous vehicles (AVs) on California public roads, according to the latest batch of disengagement reports released by the state’s Department of Motor Vehicles.

Why it matters: This marks the first time in the report’s history that a Chinese company unseated Waymo, Google’s self-driving arm and an accepted industry leader, for the top spot.

• California has required data reporting from companies for testing AVs on its public roads, including the number of miles driven autonomously and the number of times human drivers are required to take control of the vehicle, known as a disengagement.

Details: Baidu reported driving 108,300 miles and six disengagements with four vehicles last year, making for the lowest disengagement rate of all the companies listed in the California’s annual self-driving record: 0.055 per 1,000 self-driven miles.

• Baidu’s number dropped significantly from a year ago when it rated 4.86 per 1,000 self-driven miles, which the company attributed to rapid expansion in testing fields over the past three years.
• Waymo again reported the greatest number of miles driven by its 153 robocars, covering 1.45 million miles in California last year. It had one disengagement every 13,219 miles, versus Baidu’s one every 18,050 miles.
• In unusually strident language, Waymo posted a series of tweets on Wednesday questioning if the disengagement metric leads to meaningful insights, adding that its real-world driving experience takes place mostly outside of California.
• The AV leader said in December that it has more than 1,500 monthly active riders for its robotaxi pilot project Waymo One in Phoenix, Ariz. and surrounding areas.
• Doubts about the credibility of the metric are increasingly being voiced, as it is not mandatory for companies to report testing environments, which can vary from downtown traffic to empty highways.
• Executives from companies including General Motors-backed Cruise have expressed views that the metric has little value when there is no clear definition of what constitutes a disengagement.
• Disengagement data has been accepted as a barometer to compare AV companies and assess the commercial readiness of self-driving cars, and is often cited as evidence of Waymo’s leadership.
• In a statement sent to TechNode on Friday, Baidu said disengagement rate is an internal reflection of the speed of technical iterations, but comparison between companies is “not that meaningful.”

Context: Apart from Baidu, four Chinese companies were among the top 10 on the report in terms of disengagement frequency.

• Alibaba-backed AutoX reported one disengagement every 10,684 miles, ranking fourth, followed by Guangzhou-based Pony.ai with one disengagement every 6,475 miles.
• Didi broke into the top 10 for the first time with one disengagement every 1,535 miles, as did PlusAI, a self-driving truck developer which had one disengagement every 940 miles.
• Guangzhou-based WeRide drove 5,917 miles with one disengagement every 152 miles. COO Zhang Li said it has shifted road testing to China, where its fleet of more than 100 robocars drove more than 1.1 million kilometers (around 683,000 miles) and offered more than 8,300 rides within a ride-hailing pilot project as of last year.

Autonomous vehicle startup Pony.ai on Wednesday announced that it has raised $400 million in a funding round from Toyota Motor Corporation, the first-ever and biggest investment to date in a Chinese AV company by the Japanese auto giant.

Why it matters: The latest investment is expected to help Pony.ai widen the gap between the company and its rivals, as well as boost confidence at a time when some major auto and tech companies have scaled back their AV ambitions.

• Global OEMs including Daimler, BMW, and Volvo have delayed the rollout of their autonomous cars amid concerns over a lack of regulation and prospects for profitability.

Details: Guangzhou-based Pony.ai on Wednesday announced that it has secured $400 million in its Series B led by Japan’s biggest automaker and followed by existing investors. The investment is the single largest investment deal in a Chinese AV company, it confirmed, and brings the total amount the company has raised to $462 million.

• The two companies forged an alliance in testing autonomous vehicles on Chinese public roads in August using Toyota’s Lexus vehicles piloted by Pony.ai’s self-driving system.
• The new funds will be used to deepen its collaboration with Toyota on self-driving technological development, while making a push into mobility services in China, the company said in an announcement.
• The AV startup launched a robotaxi pilot project PonyPilot in Guangzhou, capital of southern Guangdong province in December 2018, testing a fleet of 100 autonomous vehicles with a safety driver behind the wheel offering ride-hailing pilot services in the city’s Nansha district.
• It began offering robotaxi services in a partnership with Korean automaker Hyundai in the city of Irvine in southern California starting in November, after being granted a license for passenger transport by the California Public Utilities Commission.
• With another two testing fleets in operation in Beijing and Fremont, Calif., the four-year-old company said that its self-driving fleet has traveled more than 1.5 million kilometers (930,000 miles) as of last year. Google’s self-driving arm Waymo last month reported a record 32.2 million kilometers of driving on public roads after 10 years of operation.
• Pony.ai has built a war chest totaling $800 million from investors including Sequoia Capital China, IDG Capital, and Kunlun, a Chinese games publisher, and was valued at more than $3 billion as of February.

Context: China’s self-driving sector is weathering a rough period amid a broader downturn in investment activity in Asia.

• Venture capital investment in Asia dropped by half to around $60 billion in 2019 compared with the previous year, according to recent figures from KPMG.
• Toyota made the biggest investment deal of the year in China’s mobility sector with a $600 million cash infusion to ride-hailing giant Didi Chuxing in July. It poured $500 million in Uber for joint development on self-driving technology in 2018.

Update: added specifics on the funding round in Details section.

China is postponing plans for massive autonomous vehicle (AV) deployment from its original target by five years as auto and tech companies continue to struggle with the challenges of truly driverless vehicle adoption.

Why it matters: China backing off its ambitious plans underscores the challenging technological leap that self-driving technology has proven to be.

• Global auto tech giants have struggled to make autonomous cars safe for public use. The valuation of Google’s AV arm Waymo was slashed by 40% to $105 billion late last year, while General Motors-backed Cruise delayed the launch of its robotaxi service beyond 2019.
• Venture capitalists are more cautious about the self-driving industry—recent valuations for some AV startups have stalled, adding to fears that the industry will soon undergo consolidation, according to a TechCrunch report.

Details: China is postponing its original goal to achieve “mass production” of intelligent vehicles with “conditional” self-driving capabilities to 2025 from 2020, according to a development plan recently released by the National Development and Reform Commission (NDRC) and the Ministry of Industry and Information Technology (MIIT), among others.

• In a draft version released (in Chinese) by the NDRC for public review in January 2018, Beijing set a target that more than half of new cars sold in China should have autonomous driving functionalities by end-2020. There were 27.8 million new cars sold in China in 2019.
• Beijing has abandoned that goal in the finalized plan, and did not disclose its new production volume goals.
• A plan for clean energy vehicle development released by MIIT in December pinpoints the percentage at around 30%, or 7.7 million units, a calculation based on last year’s auto sales in the country.
• “Conditional driving automation” is also known as Level 3, meaning that a car could drive itself under certain conditions but still requires a human driver ready to intervene, according to a rating from the Society of Automotive Engineers (SAE).

Context: Recent research by business consultancy AlixPartners shows that consumers still have safety concerns about sharing the road with vehicles operating in autonomous mode, as well as limited willingness to pay for the functionality.

• Four out of every five global participants who considered themselves likely to buy higher-level AVs said they will wait at least five more years to purchase, although only 51% of Chinese consumers said as much.
• The survey also shows that the premium which Chinese consumers are willing to pay for full autonomy is weak: only 8%, or $165 more than vehicles with existing Level 2 driver-assisted automation, which include features such as lane assistance and automatic emergency braking.

China’s on-demand service platform Meituan Dianping has made its first grocery delivery in the outskirts of Beijing with self-designed autonomous delivery vehicles, as the country’s tech companies push further into “contactless” initiatives spurred by the Covid-19 outbreak.

Why it matters: Tech firms in China are ramping up “contactless” delivery initiatives as conditions surrounding the deadly virus has created an opportunity to test experimental technologies for wider adoption.

• Earlier this month, e-commerce platform JD.com completed its first delivery of medical aid using a fleet of autonomous vehicles in the central Chinese city of Wuhan, the epicenter of the outbreak.
• A company representative said it has stepped up a plan to commercialize delivery robots, with more than 30 units in production to be deployed in Wuhan.  

Details: Beijing-based Meituan began piloting its driverless delivery service in the city’s northeastern Shunyi district earlier this month, according to an announcement on its official account on messaging platform WeChat released Tuesday (in Chinese).

• Meituan currently deploys two driverless vehicles to deliver groceries to customers in three neighborhoods within a five-kilometer (around three-mile) radius of its pickup station. It requires human employees to place goods into the vehicles before they pull out and begin deliveries.
• The electric delivery robot has a range of 100 kilometers (62 miles), has a maximum capacity of 100 kilograms or about 220 pounds, and can complete a maximum of five orders per trip, the company said.
• A company spokeswoman declined to specify the average number of orders the automated delivery vehicles make per day.
• Orders delivered by unmanned vehicles are generally those with bulkier items or those which require longer rides. Automated vehicles also perform deliveries to communities which have reported confirmed virus cases, a business lead told Chinese media.
• This is the first time Meituan’s delivery robots are running on Beijing’s public roads, and they are proceeding with caution. Vehicles run at a pace of 20 kilometers or 12 miles per hour, the company said, adding that it now works as an alternative last-mile delivery solution to help alleviate human workforce shortages.
• Driven by machine-learning algorithms and a package of hardware including Lidar, radar, and cameras, the price of driverless technology is high. Each of Meituan’s robots cost RMB 100,000 ($14,220) each, Chinese media reported citing a technical lead.

Context: Meituan began work on driverless delivery in 2016, followed by several pilot projects in geo-fenced areas such as university campuses. It launched its open-source platform for unmanned delivery two years later.

• Widescale adoption of driverless delivery is a long-term initiative, as it still requires human and robot collaboration to contend with different traffic scenarios and all sorts of weather.
• Still, machines may be able to help human delivery workers by sparing them from more taxing duties, such as night shifts or extreme weather conditions, Xia Huaxia, Meituan’s chief scientist said during an interview with TechNode last year.
• Restaurant and beverage chains are also innovating operations to accommodate “contactless” deliveries and pickups, Reuters reported. Fast food chains with their own delivery fleets are leaving orders placed via app in specified areas for customers to pick up, while Starbucks customers wait outside cafes while their drinks are prepared.

Israeli startup Innoviz is teaming up with a large Chinese truck maker on self-driving container transport on ports, as the country pushes industrial upgrades for freight deliveries using driverless technologies.

Why it matters: The partnership is an important step for Innoviz, which is going to great lengths to drive down costs for Lidar sensors in order to widen adoption in autonomous vehicles (AV), particularly in the hyper-competitive Chinese auto market.

• The move by the Israeli company comes after its rival Velodyne scaled back its presence in China, replacing its direct sales team with single agents late last year.
• Chinese media reported the US Lidar pioneer lost the market to local rivals such as Alibaba-backed Robosense, which offers an entry-level 16-laser unit at half the price of a Velodyne equivalent.

Details: Softbank-backed Innoviz is working on a pilot project with Shaanxi Heavy Duty Automobile, known outside of China as Shacman Trucks, to deploy the Innoviz Pro solid-state Lidar sensor in autonomous trucks on one of China’s biggest ports, the company said in an announcement released Wednesday.

• The two companies are testing Shaanxi’s trucks to “see” the environment using Innoviz’s solid-state Lidar sensors, which send out thousands of laser points to detect objects and create maps for their surroundings at a range of up to 150 meters.
• The truck maker has set a goal to deploy up to 600 vehicles in the port area with autonomous container loading and unloading capabilities. A China-based Innoviz senior executive declined to specify which port when contacted by TechNode on Thursday.
• Solid-state Lidar, with fewer moving parts than traditional Lidar, is considered smaller, cheaper to build, and more resilient for mass adoption.
• Shacman Trucks is China’s third-biggest truck maker by sales volume and a partner of Chinese AV startup Tusimple. The company recorded sales of 135,000 tractors and trucks last year, behind state-owned FAW and Dongfeng Motor.

Context: Founded in January 2016 by former members of the elite technological unit of the Israeli Defense Forces, Innoviz has secured total funding of $252 million from investors including Softbank, Tier 1 supplier Aptiv, and China Merchants Capital.

• The new entrant has snagged several key customers including BMW, providing the German automaker with solid-state Lidar technology for the mass production of several models in 2021.
• In an interview with Chinese media in 2018, an executive from Innoviz said it planned to outsource production of its premium model Innoviz One to a Chinese manufacturer in two years with a price tag of $1,000. A 64-laser mechanical Lidar produced by Velodyne costs $75,000.

China’s top scientists call for legislation to drive autonomous car industry