A war of words broke out between two of China’s best-known technology entrepreneurs on social media on the eve of the 2023 Shanghai Auto Show. Claims of autonomous driving are all “nonsense” and “hokum”, declared BYD’s Founder Wang Chuanfu after his company delivered its best-ever quarterly profit on the back of overtaking Tesla as the world’s largest seller of vehicles that run either partially or fully on electric power. “It’s ultimately a higher standard of assisted driving,” Wang said, describing the electric vehicle (EV) industry’s obsession with outdoing each other in self-driving as “the emperor’s new clothes”. Three weeks later, the man in charge of spearheading smart-driving technology at China’s largest telecommunications supplier hit back. “Either he does not understand this technology, or he said it deliberately” to knock the industry, retorted Huawei Technologies' rotating Chairman Yu Chengdong, who also heads the company’s smart car unit.
The argument not only played out on social media, but gets to the crux of a debate in the fastest growing EV market, where 60% of new vehicles entering China’s roads are expected to be electric by 2030: what does the perfect EV look like? Increasingly, an EV’s appeal is defined by how smart it is, not how far it can go or how cheap it is, as technological advances helped overcome drivers’ so-called range anxiety and lower production costs. “Young people in China no longer treat EVs as just vehicles, they want them to function like smartphones,” said Cao Hua, Partner at the Shanghai private-equity firm Unity Asset Management, which counts artificial intelligence (AI) and vehicle robots among its investments. “Making the cars autonomous and intelligent can draw more Chinese buyers.”
Intelligence is measured by the vehicle’s digital bells and whistles, manifested in such built-in features as voice-activated controls, facial recognition, over-the-air software upgrades, phone-linked features, and self parking. At the top is autonomous driving – graded in five levels of sophistication – due to the technological complexities of ensuring that the vehicle can move about on its own at high speed without endangering the occupant.
All 10 of the models ranked “the smartest” among China’s EVs could drive themselves with minimal human intervention in so-called conditional (Level 2 to Level 3) autonomous driving, according to a December 2022 survey by JD Power and Shanghai’s Tongji University. Nine of the country’s 10 bestselling EVs last year had this feature. The watershed moment came in 2022 when Chinese carmakers made headway to turn self driving into a reality, said UBS’ Analyst Paul Gong. “Chinese companies are developing self-driving technology fast, while striving to produce and sell cars fitted with these systems on a large scale,” he said. “They have shown their capability in controlling costs, and it is an encouraging sign that they have taken the leap forward in commercializing the technology.” The most complex component in self-driving is the light detection and range sensor (LiDAR), which uses pulsed lasers to map out a 3D view of the vehicle’s surroundings. This technology, widely adopted by Chinese EV assemblers, has turned Hesai Group into a billion-dollar company on the Nasdaq market two months after its initial public offering (IPO).
China’s highway laws bar Level 5 full automation – requiring zero human intervention all the time – except on designated routes in controlled environments. Most of the self-driving capabilities found in China’s EVs are either classified as L2 or L3, where sensors are used to give a vehicle “environment detection” capability, enabling it to decide whether to pass a slow-moving car. But the technology still requires human override and the driver must be alert and ready to take control. “It will take some time before fully autonomous driving passenger cars can be commercialized,” said Wang Xiaogang, Co-founder and Chief Scientist with the AI firm SenseTime, a key supplier of autonomous driving technology in China.
“Eventually, driverless cars are supposed to be safer than cars with humans behind the wheels.” SenseTime provides a visual perception system to achieve autonomous driving. It also develops car analysis and behavior detection technologies to ensure driving safety, such as facial recognition technology, which can tell whether the driver is drowsy. It has already supplied autonomous driving systems to about 30 car making clients, including Nio and Geely’s premium EV brand Zeekr. Wang, also head of the company’s smart car business division, said SenseTime had signed agreements with the assemblers to install a total of 31 million autonomous driving systems in their cars.
China’s congested roads, large population and long urban commutes have created voluminous data for self-driving systems, getting them into the mainstream sooner than in many other markets, analysts said. An example of that advantage was the navigation-guided pilot (X-NGP) installed in Xpeng's G6 mid-size sports-utility vehicle (SUV), the South China Morning Post reports.