Autonomous driving startup Waabi just deployed its first robotic semi-trucks for on-road testing in the U.S., a little over a year after it came out of stealth, and founder Raquel Urtasun thinks it may be able to commercialize the technology faster than bigger, better-funded “previous-generation” rivals.
Toronto-based Waabi, which powered up in June 2021 with an $83.5 million funding round, styles itself as a second-generation AV company, in contrast to Alphabet’s Waymo, General Motors-backed Cruise, Amazon’s Zoox and other big players in the self-driving tech space. Rather than starting by putting lots of cars and trucks on the road in the real world to slowly learn the unique features of individual cities and streets to refine its software, the company focuses on aggressive virtual testing with the “Waabi World” simulator CEO Urtasun developed that allows the AI-enabled system to experience nearly limitless types of road conditions.
“Because we can do everything on the simulator we are already ready with a generation that is much more advanced,” she tells Forbes. For example, even the placement of cameras and laser lidar and other sensors that Waabi’s test trucks need to see their surroundings were optimized as a result of extensive simulator testing, rather than by trial and error on the road.
“What we’re doing is in contrast with how, typically, hardware and software work in the industry where first you build a prototype and then software has to be created to handle that kind of hardware,” Urtasun said, declining to critique the approach of any specific companies.
Still, there are questions about the company’s future at the moment. Urtasun declined to say which truckmakers and trucking companies Waabi will be working with, whether its test trucks will be hauling revenue-generating loads and how soon the system will be available for commercial sale or licensing. Its first trucks will operate in a U.S. state she didn’t name and though Urtasun also wouldn’t say how many it’s testing, three trucks are currently registered to Waabi in San Francisco, according to the U.S. Department of Transportation.
Waabi’s optimism–as well as Urtasun’s claim that its 2021 funding round provided a “multiyear runway–is something of a contrast to growing pessimism around when autonomous vehicles will make financial sense. Ford and Volkswagen’s October decision to shut down Argo AI, their joint-venture self-driving vehicle company, was a bombshell move in the industry, particularly after the automakers had provided more than $3 billion of funding. TuSimple, a leading robotic truck developer, aimed to commercialize its system by as early as next year but recent management and board turbulence, as well as potential federal reviews, have clouded its outlook. Waymo continues to expand the scope of its robotaxi service in Phoenix and aims to begin operating in San Francisco and Los Angeles, but it’s a slow process compared to its initial more optimistic outlook.
Urtasun, a University of Toronto professor and former chief scientist for Uber’s autonomous vehicle operation, says Waabi can move faster because it’s learned from the experience of its competitors. Rather than mastering one city or environment at a time, she believes the best way to scale up autonomous technology is to teach the system to handle an extensive range of conditions in simulation and then fine-tune that learning with physical trucks on the road in individual locations.
“The negative sentiment (for autonomy) actually affects the previous generation of companies. It’s becoming more and more clear that the technology is not there and it may take a very long time, if ever, to be solved with that approach,” she says. “They are very capital intensive and when the economy, the markets look like they look today, that is extremely hard when you have a burn rate of half a billion, $1 billion or $2 billion.”
And unlike Waymo, Cruise and the recently shuttered Argo AI, with thousands of employees, Waabi is moving toward commercialization with a staff of about 100 people.
“There’s still a lot of excitement about this technology and what it can provide for the world,” Urtasun said. “But I think there is a realization that we need a different approach.”