After a high-speed crash in Arizona, the ride-hailing giant grounds its autonomous fleet.
In the era of self-driving cars, a scary but otherwise uneventful car crash can be huge news. This was the case in Tempe, Arizona, on Friday, when an Uber self-driving car was hit so hard that it rolled onto its side. There were no serious injuries reported.
Uber has grounded its fleet of self-driving cars in Arizona as a result, a spokeswoman for the company told me. “We are continuing to look into this incident, and can confirm we had no backseat passengers in the vehicle,” an Uber spokesperson said in a statement provided to The Atlantic. Uber also suspended testing of its self-driving vehicles in Pittsburgh and San Francisco “for the day, and possibly longer,” The New York Times reported. In addition to its global ride-hailing service, Uber has been testing its self-driving car technology on public roads in Arizona, Pennsylvania, and California for several months.
The vehicle involved in the Arizona crash was in autonomous mode at the time of the collision—meaning the car was driving itself with a human riding behind the wheel—but police in Tempe say Uber wasn’t to blame for what happened. A human-driven vehicle failed to yield at a traffic signal, and collided with the Uber SUV, police said in local news reports.
The incident is a reminder of the need for this technology in the first place: Humans are abysmally bad drivers. But it’s also a reminder of how much Uber has riding on the success of self-driving cars.
And how much is that? Everything, basically.
If self-driving cars are adopted on a mass scale and Uber isn’t leading the way, its current business—which revolves around humans driving cars—is made obsolete. But if Uber finds a way to dominate in the development of self-driving cars, it can remove those costly human drivers from its business model—a scenario that could mean a windfall for Uber. Succeeding on this front “is basically existential for us,” Uber’s CEO, Travis Kalanick, told Bloomberg Businessweek in August.
The ride-sharing company is uniquely positioned in the self-driving car space this way. Google’s self-driving car project, now rebranded as Waymo, could fail and its parent company would still have a massively profitable search engine to fall back on.
“Every other company isn’t betting the company’s future on self-driving cars,” said Arun Sundararajan, a professor at New York University’s Stern School of Business. “Google will be fine. Uber is the one company in the world who has really made this all-or-nothing bet.”
The crash in Tempe is unlikely to derail Uber’s self-driving car aspirations—especially if the incident played out the way law enforcement described, and Uber was not at fault. But Uber is facing several other major problems. For one, there’s the federal lawsuit.
Waymo is suing Uber for intellectual property theft, claiming that an Uber engineer who used to work for Google stole 10 gigabytes of “highly confidential data” from Google’s servers, then used it to copy Google’s designs for a self-driving car. The lawsuit is a “particularly damaging development for Uber,” given how much is at stake, Sundararajan told me.
There are also reports of internal strife on Uber’s self-driving car team, which has escalated into a “mini civil war,” according to the tech-news site Recode. Leaders from Uber’s self-driving car unit gathered in San Francisco for a “critical summit” on the matter last week, hoping to solve leadership problems and figure out a way to stem departures from the embattled company. “With every engineer that defects,” Recode wrote, “Uber is feeding the fire of its competitors, which are growing by the day, both big and small.”
“Uber has tied their fortunes to the imminent arrival of fully autonomous cars, which is highly risky,” Sundararajan told me. “It’s an outcome with a lot of variability, right? It could be three years, it could be 10 years.”
In other words, anything that gets in the way of Uber’s work on self-driving cars, he says, is “particularly troubling” for the company’s survival.
This post originally appeared on The Atlantic.