GM, Honda put Cruise behind the wheel
Kyle Vogt’s company has sped to the front of the autonomous-vehicle pack
SAN FRANCISCO — Kyle Vogt was 14 when he saw the light, sitting behind the wheel of his father’s car as it crossed the arrow-straight highways of the American Southwest. “I had my learner’s permit, and as I was driving from our home in Kansas to Las Vegas for a remote-controlled battle-bot competition and I thought, ‘Surely this can be done by robots, just read the lane markers and keep the wheel steady,’” Vogt says with a laugh. “So that was the beginning.” Nearly two decades later, Vogt, 32, has made that epiphany his mission in life.
As founder and CEO of Cruise, a self-driving automotive tech company that General Motors bought for $1 billion in 2016, Vogt has sped to the front of the autonomous car pack thanks most recently to a big partnership deal with Honda.
There are dozens of car companies and tech startups tackling the selfdriving car space, given that ride-hailing fleets with no drivers promise untold riches assuming tech costs and car ownership both decline.
But Cruise arguably has only one main competitor: Alphabet-owned Waymo, which has been at this for a decade, and for a year now has been busily testing hundreds of self-driving Chrysler Pacifica minivans in Phoenix.
Waymo is aiming to make its betatesting program public this year.
Cruise is aiming to start picking up passengers here in its Chevy Boltbased self-driving cars sometime in 2019.
Vogt is vague on whether the program will be limited to a select group of riders at first, in an echo of Waymo’s current beta program.
But he wishes to make one thing clear: Cruise’s testing across this hilly, traffic-snarled city will ultimately yield more capable robot cars.
That’s because since 2016, Cruise has deployed a small group of cars (at first 50, now 20) on what he describes as the comparatively easy-to-navi-
gate streets of Phoenix.
“Just look at construction zones, for one,” Vogt says during an interview with USA TODAY inside a spartan conference room at Cruise, whose headquarters just south of downtown San Francisco seems as heavily guarded as Fort Knox.
“Those (zones) pop up 30 to 40 times more often for our cars in San Francisco than they did when we were testing our own cars in Phoenix,” he says. “That means as an engineer working in Phoenix, you can just write that off as a problem. But we’re dealing with that, with emergency services vehicles, with bicyclists. It’s a lot more challenging.”
Vogt also likes to highlight how the software and hardware in Cruise vehicles is fully integrated thanks to Cruise being wholly owned by GM, a comparison perhaps to Waymo’s vehicle partnerships with Fiat Chrysler and Jaguar Land Rover.
“Look at these sensors, they’ve all got air- and water-cleaning nozzles built into the housing, and all the connections and wiring to our computers are autograde materials,” he says during a brief tour of Cruise’s cavernous engineering facility.
One floor down is another key part of the Cruise self-driving service plan: a room that has echoes of an air-traffic control center, with glowing screens and about a dozen workers on headsets.
This is where video and data comes in from a self-driving car that is momentarily stumped by its surroundings and is requesting remote-control help.
“The cars off-load to humans about
0.1 percent of the time during the (test) rides, but we still want to have this here to build trust with our users,” Vogt says. “Besides, if we’re driving one mile, we want to get the maximum value, the maximum knowledge, from that mile.”
❚ Quality, not quantity, of test miles:
Most mobility experts say that the race toward autonomy is not a winner-takeall proposition. In the end, much like the airline industry has big players like Boeing and Airbus providing planes for travelers to fly on, a few big companies will be making the self-driving cars that shuttle us around in the decades to come.
But in the short run, Vogt is indeed implying that the self-driving fleets that win over those early-adopters (polls show that half of respondents remain nervous about riding in autonomous cars) will be companies that have demonstrated an ability to handle extreme urban road conditions and, eventually, extreme weather.
Given that self-driving ride-hailing fleets remain largely a mobility vision of the future, the company that dominates this space will need brilliant minds and lots of money.
While Waymo’s funding appears unlimited, Cruise will have to keep an eye on its cash pile, says Bryan Reimer, a research scientist and self-driving expert with MIT AgeLab.
“As part of Alphabet (the parent company of Google), Waymo has a limitless budget and no concerns about immediate financial returns. Cruise has GM’s resources, the recent ($2.25 billion) investment from (tech investors) SoftBank, and now the Honda deal ($2.75 billion over 12 years). We’ll see if it’s enough,” he says.
Vogt acknowledges that challenge and adds that the cost of a self-driving car ride has to come down quickly to guarantee enough consumer adoption to make the program financially viable.
❚ Startup and automaker combo:
Collaboration will be critical to that goal, says Karl Brauer, publisher of Cox Automotive. He says companies that go it alone in this space risk lagging behind.
“What’s better than one automaker aligning with a tech company like Cruise? How about two automakers, GM and Honda,” says Brauer. “What’s particularly impressive is that Cruise has avoided becoming a subdivision of GM. That could have torpedoed their success. But Vogt has steered clear of that.”
Vogt credits GM CEO Mary Barra with letting Cruise stay in San Francisco, where the company has been hiring feverishly – the company has 1,000 employees and expects to add hundreds more in 2019.
Cruise’s decision to use San Francisco to test its 180 Bolt-based electric cars – which in the future will be joined by a new Cruise-branded vehicle made jointly by GM, Honda and Cruise – hinged on that fact that urban centers, with their high Uber and Lyft usage, can support higher prices for those initial self-driving car rides.
But the calculus that drives Vogt and his investors involves a scenario where self-driving vehicles are circulating 24/7 in major U.S. cities and are charging a fraction of what a current Uber or Lyft ride commands.
“There are 3 trillion miles driven each year in the U.S. now,” Vogt says, “and Uber and Lyft only access about a half a percent of those miles, because having a human chauffeur you around is still very expensive, so most people prefer to have a car.
“But when autonomous cars drive down that cost-per-mile number, and people decide to save on car and insurance and gas payments, we see us capturing 50 percent of those miles,” he says. “Even at a dollar per mile, when you’re in the trillions of miles, that’s big money.”