China’s Waymo challenger Pony.ai hits the accelerator to speed up to a robotaxi fleet of 200
Pony.ai said it will gradually add its new driverless cars to China and the US, with 100 vehicles in each, as it tries to catch up on autonomous driving leader Waymo of the US.
Pony.ai, China’s most valuable autonomous driving start-up, is picking up speed, with plans to boost its robotaxi fleet to more than 200 from the current 20 by early next year.
The move will give the start-up more driving mileage and therefore more high-stakes data needed to put self-driving cars closer to commercialisation.
“Our next target is to build a fleet, as what is most important is to turn autonomous cars into mass production and achieve scalability,” said James Peng, co-founder and chief executive of Pony.ai, at a briefing on the sidelines of the World Artificial Intelligence Conference in Shanghai on Tuesday.
“The fleet will allow us to hone overall stability and reliability of the self-driving system, readying for getting closer to commercialisation,” he added, without providing a date.
The revolution under way in the role cars play in daily life is attracting a diverse field of big tech firms, start-ups and traditional car makers, with the US and China battling for dominance.
Major US players include General Motors and Alphabet, Google’s parent company. China’s long list of players includes search engine Baidu, which launched a pilot taxi programme in Chongqing in May, gaming and social media giant Tencent, Alibaba Group, the owner of the South China Morning Post, and Pony.ai and fellow start-up Jingchi, which were both founded in Silicon Valley but relocated their headquarters to mainland China last year. Waymo, Alphabet’s autonomous car subsidiary, is considered to be ahead in the race.
Pony.ai said it will gradually add the new driverless cars to China and the US, with 100 vehicles in each.
The Guangzhou-based company brought 10 refitted self-driving cars to the AI conference, offering test rides. The cars drove themselves at a speed cap of 30 kilometres per hour (about 19mph) in a closed park and in intermittent drizzle. Passengers queued up, and received certifications after their test ride at the finish point. A person in the driver’s seat served as a monitor, with no hands on the wheel. A screen between the two front seats showed a map of the ride in real time.
Still, Pony.ai’s ramp up in fleet size leaves it far behind Waymo, which is buying 62,000 Chrysler Pacifica hybrid minivans to expand its fleet.
Waymo is expanding its pilot project under way in Phoenix, Arizona, where on-demand robot rides are provided for free. By the end of the year, passengers will pay.
So far, it is not possible to formally compare Pony.ai with its rivals. It only applied for road tests with the California Department of Motor Vehicles last spring, which meant it was not included in an annual “disengagement” report in January on how much human drivers have to take over from a car’s self-driving system.
“The starting point is not most important,” Peng said, when asked to compare Pony.ai with Waymo. “Start-ups can win by fast iteration … The solution Pony.ai has brought [to the conference] is the third version introduced in less than two years of its establishment.”
Zhang Yimeng, head of Pony.ai’s perception research, who joined last month after working with Google for six years, echoed the remarks.
“It’s true that Waymo had an early start with nearly a decade of work. But it was only in recent years that the technologies became mature enough to transform into a product,” Zhang said.
Pony.ai announced the completion of a US$214 million round of fundraising in July, marking the biggest series A round by a self-driving company in China. Roadstar.ai, another in the driverless car battle, pulled in an A round of US$128 million one month earlier.
Like Pony.ai, Roadstar.ai started out in Silicon Valley but is now pushing forward their development efforts on the Chinese mainland, with the goal of introducing robotic taxi services in the country.
Betting big on the core technology behind cutting-edge applications such as self-driving, China’s State Council last July laid out a three-step road map for artificial intelligence supremacy. It included the goals of building a domestic AI industry worth about US$150 billion and to make the country an “innovation centre for AI” by 2030.
China has tremendous possibilities for driverless vehicles, Peng said, because of its big population and strong demand.
“With less than 10 per cent of usage rate for private cars, Robotaxi has always been our goal, and we will keep working toward it,” he said.