BEIJING – I recently hopped into a Chinese robotaxi and was surprised to see just how promising this technology is – though it might not reach the U.S. anytime soon.
On my ride, I was seated in an elegant chair, with plenty of space usually reserved for a front passenger. With no driver at the wheel, the vehicle glided smoothly through Beijing’s busy streets, masterfully maneuvering around double-parked cars and unexpected turns. It wasn’t long before I forgot there was no human in control.
The overall experience felt pretty similar to the driverless taxis from Waymo, part of Google’s Alphabet umbrella. While American inventions like Waymo and Tesla’s robotaxis are constantly in the news, Chinese companies are developing their own markets and claiming their share in Europe, the Middle East, and Southeast Asia.
Experts at UBS predict that the global market for self-driving technologies could potentially be worth hundreds of billions in the near future, which encourages these companies to push forward.
In China, major firms like Baidu, Pony AI, and WeRide have deployed hundreds of robotaxis offering services without human backup drivers.
Leo Wang, CFO of Pony AI, mentioned how their vehicle production costs are significantly less than those of Waymo, which could be a strong advantage as they ramp up their operation.
HSBC analysts estimate that by the end of next year, China’s fleet of robotaxis will soar to tens of thousands, a steep rise from just a few thousand on the roads today. In contrast, Waymo is testing its vehicles only in Japan and has plans for London, while Tesla’s services are confined to Austin, Texas.
A spokesperson from Waymo asserted that they are at the forefront of autonomous technology, boasting about their operational scale and performance safety.
During my rides with WeRide and Pony AI, I was impressed by the vehicles’ decision-making skills. You should have seen it navigate around a double-parked car during one ride – it slowed a bit for safety, only accelerating once the obstructing human moved aside.
The leader in this space is Baidu, the Chinese internet giant with its well-known search engine. Baidu already has over 1,000 autonomous vehicles, predominantly operating within China, and claims to have completed more than 14 million rides as of August. For comparison’s sake, Waymo surpassed the 10 million ridership number recently.
In July, Baidu made headlines with a deal with Uber to deploy even more vehicles in Asia and the Middle East. Next on their agenda is an agreement with Lyft, rolling out cars in Germany and the UK, and talking about future expansion into Southeast Asia.
It’s looking pretty unlikely that American roads will see Chinese robotaxis. The U.S. trade barriers through hefty tariffs make importing these electric vehicles a tall order, and both the Biden and Trump administrations have openly voiced worries about the data collection practices of Chinese manufacturers.
Industry players feel this global race is still just warming up, with operations beyond the U.S. and China being mainly small-scale trials. Waymo’s financial strength and extensive experience across different U.S. markets provide them a solid advantage should they decide to speed up global operations.
Meanwhile, companies in China leverage local EV factories to create affordable vehicles, paving the way for luxurious features, like massage chairs fitted in the latest Apollo Go models, or cozy seats with armrests, which I notably enjoyed in my WeRide experience.
As WeRide’s marketing head Maeve Zhang mentioned, “It’s not just about the tech; the passenger’s experience matters too. We’re working on features that let you chat with the vehicle, saying things like, ‘Play me some Taylor Swift.’”
Another competitor worth noting is Momenta, which despite not manufacturing its own vehicles, supplies its driver-assistance tech to many cars and gathers anonymous data for AI development. For Uber, they aim to implement their technology into robotaxi services in Europe, which kick off in Munich next year.
The increasing deployment of robotaxis highlights a reliance on simulator training that prepares them for unique situations which may happen infrequently. “Good synthetic data can be invaluable in this context,” commented Wang from Pony AI.
However, the business model for robotaxis still faces skepticism – as shown when General Motors nixed their Cruise robotaxi plans after pouring in a whopping $10 billion.
Pony AI and WeRide are also navigating troubles with revenue; although increasing, it’s still only hitting the low million-dollar figures per quarter, and both firms just debuted on the Hong Kong exchange with mixed reactions, giving way to conversations about their sustainability prospects.
Safety is a paramount concern. While the Chinese robotaxi companies seem to have avoided severe accidents, they’ve dealt with incidents such as a Pony AI car bursting into flames last May, fortunately without injuries.
In the words of Musk, during a Tesla earnings call regarding robotaxis, “The world doesn’t quite grasp how quickly this will phase in.” Yet, he echoed concerns from his peers about cautious deployment, noting the repercussions that even one accident could provoke in global headlines.
For further inquiries, you can contact Peter Landers at Peter.Landers@wsj.com.
