PCMag editors select and review products independently. If you buy through affiliate links, we may earn commissions, which help support our testing.

Uber COO on Robotaxis: The Economics Don't Work...Yet

Self-driving car hardware is 'quite expensive,' Andrew Macdonald says at Web Summit, but Uber still plans to 'feather in' autonomous vehicles alongside human drivers.

 & Rob Pegoraro Contributor

Our team tests, rates, and reviews more than 1,500 products each year to help you make better buying decisions and get more from technology.

Our Expert
LOOK INSIDE PC LABS HOW WE TEST
65 EXPERTS
43 YEARS
41,500+ REVIEWS
(Credit: Smith Collection/Gado/Getty Images)

LISBON—Self-driving vehicles may be accelerating up an on-ramp for Uber, but robotaxis that would yield a profit for the ride-hailing company aren't yet crawling along an assembly line. 

Uber COO Andrew Macdonald offered that measured take on autonomous cars in a Web Summit panel here this week. "This is commercialization that's starting to happen right now," he said after name-checking Uber's addition of Waymo robotaxis in Atlanta and Austin as well as its recent launch of service with WeRide self-driving vans in Abu Dhabi.

"The hardest problem to solve has been the technology component," Macdonald told his onstage interviewer, Bloomberg's global automotive editor Craig Trudell. However, while Uber has "largely solved" the engineering challenge of writing software that can drive "10x as safe as a human," using that technology to make money off self-driving electric vehicles remains an ongoing venture. 

To get there, Macdonald said, Uber needs not just "supportive regulatory frameworks" (a phrase he used more than once) but cheaper cars. "The underlying hardware is still quite expensive. The economics don't work—today."

He suggested that it will require a substantial drop in the price of a robotaxi: "It probably needs to go somewhere from a half to a quarter of where it is today."

Uber is optimistic about that happening in China, with Macdonald citing the work of such Chinese firms as WeRide, Pony.ai, and Baidu. "The Chinese automotive sector has a proven track record of being able to iterate and have great tech but also bring the costs of that tech way down with scale," he said. 

'This Is Going to Be a Delicate Transition'

Uber's plans for autonomous vehicles don't involve any abrupt cutover from human to digital drivers, even though AVs can run almost nonstop. Macdonald said that the average robotaxi in Austin "is busier than 99% of the drivers on our platform," which invites questions about that extreme 1% of drivers.

"From day one, autonomous vehicles generally aren't going to be able to handle every situation, every type of weather, every road, every speed," Macdonald said. Waymo, for example, has taken a tentative route to offering highway rides in its self-driving Jaguar i-Pace EVs

"They're also not going to be fleeted to peak levels of demand because otherwise you'll just have a heavily underutilized asset," Macdonald added. He predicted that this would not result in human drivers being put out of work, instead suggesting that making ride-hail service more widely available would boost overall demand.

"The reality is, this is going to be a delicate transition going from fully human labor to AI in the physical world in the form of autonomous vehicles," he said. "Our network will help us sort of feather in those autonomous vehicles." What would happen to Uber drivers at the end of that transition did not come up in this discussion.

Uber spent years trying to develop its own self-driving vehicles, but abandoned that effort after a test vehicle hit and killed a pedestrian in Arizona in 2018. The company is now sticking to a partnership strategy that keeps folding in more companies. Over the last year, Uber has announced autonomous-vehicle initiatives with such firms as Volkswagen, EV maker Lucid, and (as of last month), chip giant Nvidia

"We've announced 20 partnerships today," Macdonald told Trudell. "Eighteen months ago, that number would have been five or six."

He called the Nvidia partnership "a really exciting one for us" because it would both yield "software that can be deployed across a variety of OEM partners" and allow Uber to contribute "a ton of ride-sharing specific data that's very rich for the problems that we're trying to solve" to Nvidia's AI models.

Trudell closed out this 20-minute session with a hint that expanding robotaxi service to Lisbon—Uber already plans to start self-driving testing in London and Munich next year—would require policy changes in Portugal. "We need a supportive regulatory environment," he said. "I have no reason to believe that won't develop, and I promise if that develops, we will bring autonomous vehicles to Lisbon next year."

Disclosure: I'm moderating three panels at Web Summit, with the conference's organizers covering my hotel and airfare.

About Our Expert

Rob Pegoraro

Rob Pegoraro

Contributor

Rob Pegoraro writes about interesting problems and possibilities in computers, gadgets, apps, services, telecom, and other things that beep or blink. He’s covered such developments as the evolution of the cell phone from 1G to 5G, the fall and rise of Apple, Google’s growth from obscure Yahoo rival to verb status, and the transformation of social media from CompuServe forums to Facebook’s billions of users. Pegoraro has met most of the founders of the internet and once received a single-word email reply from Steve Jobs.

Read full bio