At a 2015 event, then-Uber CEO Travis Kalanick pondered aloud the future of his ride-hail company in a world of self-driving cars. There were still a ways off, he acknowledged—five or 10 or 15 years. But to him, Uber’s role was clear. "Are we going to be part of the future?" he said. "Or are we going to resist the future, like that taxi industry before us? For us, we're a tech company, so we've said, let's be part of that.”
Kalanick isn’t at the head of Uber anymore, but the specter of disruption remains. Ten years later, self-driving vehicle companies that mostly didn’t exist in 2015 are readying robotaxis for passenger rides. Moreover, nearly every player in the currently hot robot car space has something in common: They’ve signed a deal with Uber.
Yes, really. Take a look:
It’s the classic “throwing spaghetti on the self-driving cars to see what sticks” strategy. Uber’s interest in self-driving makes a ton of sense. The business estimates it spends $2 per mile to have a pesky human behind the wheel, and Dara Khosrowshahi, Kalanick’s replacement, said in a recent interview that Uber pays drivers a global average of 80 percent of riders’ fares. (Many drivers believe Uber takes much more.)
How much more money could Uber made if robots did the driving? “We think it’s an enormous, enormous long-term opportunity,” Khosrowshahi said.
This year alone, Uber has announced tie-ups with China’s Baidu, Pony.ai, and Momenta; Volkswagen; the Michigan-based developer May Mobility; and this month, the Bay Area self-driving vehicle company Nuro and Arizona EV manufacturer Lucid, who together say they’ll launch 20,000 robotaxis over the next six years, starting in a US city next year.
As the world, and the taxi business, hints at big changes on the roads, Uber seems poised to maintain its status as the Kleenex of ride-hail, a name brand synonymous with an entire category. It's inconsequential who builds the tech—when you call a robocar, Uber wants you to use its app.
“To them, it doesn’t really matter who ultimately succeeds,” says Sam Abuelsamid, who writes about the self-driving-vehicle industry and is the vice president of marketing at Telemetry, a Michigan research firm. “If you’ve got a car that works and can drive safely, you’re welcome to come onto Uber and provide rides.”
Still, it’s too early to say whether the Kleenex gambit will work.
Plenty has changed since 2015. Kalanick is no longer at Uber, deposed by a hostile board in 2017. The company marked a grim milestone in 2018 when one of its own testing self-driving vehicles struck and killed a woman. The incident, for which federal investigators later found the ride-hail giant partially responsible, led to a suspension and then reorganization of Uber’s self-driving development effort.
In 2020, Uber sold off its autonomous vehicle unit to a competitor. In some ways, though, this asset-light existence—where Uber serves as the middleman for drivers and riders, without owning its own (robo)car—seems to have worked for the company. Under the guidance of CEO Dara Khosrowshahi, the company finally recorded its first profit last year.
One potential issue for Uber is that its particular role in the autonomous vehicle industry won’t be super useful for a while. Uber is powerful because it’s already on the phones of some 160 million active monthly users all over the world. The company is good at matching people driving cars with those millions of people who want rides. But there likely won’t be millions of robotaxis for a while.
Waymo, the US leader in robotaxis, has about 1,500 vehicles operating in five cities. Baidu says its next city, Dubai, will have 100 robotaxis by the end of this year. “This is a marketplace that for quite some time will be supply constrained, not demand constrained,” says Len Sherman, a professor at Columbia Business School who has written about Uber. Self-driving car developers want access to Uber’s network—but because there simply aren’t that many self-driving cars, the company is less useful in the near-term.
This leads to another potential issue: Uber may have less power to get a big chunk of each fare in the robotaxi world. The company has spent billions figuring just how much they need to pay individual drivers to take on fares. Robotaxi tech developers who have spent their own billions building self-diving software will likely look to take a bigger portion of each fare. After all, companies including Tesla and Waymo run their own ride-hail apps. Do they really need Uber? “I guarantee they’ll drive a harder bargain,” says Sherman. (A spokesperson for Uber didn’t provide financial details of its existing partnerships.)
Chinese Uber competitor Didi—which acquired Uber’s China business in 2016—seems to be following the old Uber self-driving playbook. It has its own autonomous vehicle technology subsidiary, which is building autonomous vehicle software. It said last year that it would work with EV firm GAC Aion to mass produce robotaxis starting this year.
It may be that Uber hasn’t totally closed the door on owning some of its own robotaxi tech. Earlier this summer, the New York Times reported that Kalanick was back, and in talks to acquire the US arm of the Chinese AV company Pony.ai—with a financial assist from Uber. A spokesperson for Pony.ai declined to comment on the report. Uber told the Times that it plans to work with many AV players globally. The Kleenex strategy, in other words.
One company is conspicuously missing from the tall stack of Uber’s autonomy partnership press releases, of course. In a February interview, Uber CEO Khosrowshahi seemed to indicate that’s not for lack of trying. Tesla appears to want to own its whole self-driving car operation: the technology, the cars, the maintenance, and the app that powers it—but Uber could still be a great robotaxi partner, Khosrowshahi said. “Ultimately, we’re hoping that my charm and the economic argument gets Tesla to work with us as well,” he said.
