Autonomous Driving in China Robotaxi Accident at the Worst Possible Moment

Related Vendors

FAULHABER Drive Systems

Wuxi InfiMotion Technology Co., Ltd.

At the beginning of December, a robotaxi in Zhuzhou, China, hit two pedestrians, one of whom was so severely injured that they had to be admitted to the intensive care unit. A tragic accident, especially for the pedestrians, and a setback for autonomous driving.

According to reports on China's social media, a robotaxi from the Chinese mobility service provider Hello (formerly Hellobike) hit two pedestrians. One of them was run over by the vehicle and had to be freed by passersby who lifted the robotaxi together. The accident victim is already in the intensive care unit of a local hospital, CarNewsChina reported.

Hello, whose largest shareholder is the Alibaba Group, had only received its license for test drives in August, which were restricted to certain roads in Hunan province. Like all other robot taxi companies in China, however, it had already operated accident-free for many kilometers by Saturday. However, the images of the accident victim lying on the side of the road waiting for an ambulance after being freed will preoccupy the entire industry and possibly result in stricter safety requirements, especially in China.

Many Test Drives

Just one day before the tragic accident, it had seemed as if the robot taxi industry was doing better than ever before. Even small milestones were celebrated in the trade press, such as a robotaxi from Didi Autonomous that was allowed to drive around the clock. In a district of Guangzhou, the “Chinese Uber” has been allowed to transport passengers “24/7” on a trial basis since the beginning of this month without a driver in the car, the company announced.

Initially, this case involved a shuttle service during a sporting event, the "15th National Games." According to Chinese media reports, a longer test then began in the Huangpu district of the southern Chinese metropolis. On their route, the robotaxis currently pass by "schools, shopping centers, office buildings, and residential houses," it is reported.

Just one of many reports in recent weeks and months suggesting that the deployment of robotaxis in China is gaining new, strong momentum. For the first time, it appeared that a financially sustainable commercialization of this business model might be possible in the long term.

Didi was not the only provider of driverless taxi services that could hope for a breakthrough. Apollo Go from Baidu, the largest operator, already has more than 1,000 autonomous vehicles on the road, most of them in China. By 2027, the company aims to have 20,000 of them operating worldwide. According to the company, Apollo Go reached the milestone of 14 million total test drives last year.

WeRide, a provider from Guangzhou, has already gained 1,700 days of experience with a commercial robotaxi service in designated zones of the city as of the beginning of this month, reports Tech in Asia. Pony.ai, a Chinese startup for autonomous driving solutions, operates a fleet of 100 robotaxis in Nansha district of the same city, generating 1,000 paid rides per day. Nationwide, the startup has deployed more or less autonomous robotaxis in more than 300 cities.

High Pressure Weighs on the Industry

Every single ride counts in this competition. On the one hand, providers collect important data, which, similar to individual chat inquiries in chatbots, are used to further improve the systems with the help of artificial intelligence. Robotaxis thus become gradually safer. What everyone feared at the same time, however, was an accident like the one that has now occurred in Zhuzhou.

Although initial reactions on Chinese social media include comments that “something like this could also have happened with a human driver at the wheel,” the wave of headlines around the world will certainly give momentum to opponents of the new technology. The accident happened at a time when all providers were already under pressure to prove the commercial viability of their robot taxi businesses in the coming years. Now they will undoubtedly have to talk about the safety of their systems even more often than before. For the young industry, this incident therefore represents a serious crisis. So far, only Apollo Go in Wuhan has reached the break-even point; everywhere else, investors' and companies' money is still being burned at a massive rate.

Scaling is the Order of the Day

Scaling, i.e., the deployment of more and more new robotaxis, would be the order of the day at the moment. The more of these vehicles are built and deployed, the more their purchase price will fall. This also applies to the roadside hardware that is currently being installed in some places in China for connected driving variants. And the more kilometers can be “collected,” the greater the chance for providers to obtain further approvals at home and abroad. All of this is now being called into question, at least temporarily, by the accident.

Comparable serious accidents suggest that the Chinese legislature will likely respond by tightening regulations but will not permanently impact the entire industry. If previous response patterns repeat, the accident will lead to a phase of reflection and deceleration, which will then transition into a new growth phase.

It is now even more important than before the accident to stay in the game for the long term in the race for the robotaxi business. Those who don't have to quit, like Volkswagen or Ford did back in 2020, can look forward to long-term profits.

Positive Forecasts

The industry has recently seen a significant upward trend in terms of revenue. Goldman Sachs recently published a forecast predicting that total sales in the Chinese robot taxi market could increase tenfold from around $46 million today to around $46 million in 2035. UBS is even more optimistic, predicting a market of around $170 billion for robotaxis in China by the end of the 2030s. And demand is high in China alone. There are 139 cities with more than one million inhabitants.

The Chinese government has recognized the new industry as a growth driver for the domestic economy and is promoting it with the issuance of licenses and the approval of more and more pilot zones, even though it had already started to gradually tighten road traffic safety regulations before the accident.

So far, only a few players in the Chinese market have had to deal with temporary setbacks. Didi Autonomous, which dominates the market for Uber-like ride-sharing services in China, had some issues with regulators between 2021 and 2022 after it tried to go public in New York against the wishes of central planners. Not only was Didi subsequently denied the acquisition of new customers for its taxi app for a while, but the company's robot taxi plans also became conspicuously quiet for a time. However, since the beginning of 2023, after the regulatory authorities and the Communist Party were appeased, Didi has been making a comeback in the autonomous driving and robot taxi segments.

Technologically, the group has built up “full-stack” capabilities for level 4 autonomous driving. Investors continue to provide it with money, as a recent new round of financing has just proven. In addition, Didi is in the fortunate position of earning a lot of money with its ride-hailing app “Didi Chuxing,” which has a share of about 70 percent of the huge market for such ride-hailing services in China.

Enhance Safety

Other providers, such as startups like Pony.ai, have to build their customer base for robotaxi rides much more laboriously than Didi, which can simply offer the robotaxis in its app that almost everyone in China has installed on their smartphone from elementary school age onward. There is therefore no concern that Didi will run out of money for further developments.

It remains to be seen how significant and long-term the impact of the weekend's accident will be on the Chinese robotaxi market. Wishing the two victims a speedy recovery. Manufacturers must be called upon to develop an even stronger awareness of the risks associated with driverless cars. (se)

Machine Authentication

Machine Operator Authentication - Why knowing "who" matters