PITTSBURGH -- Bryan Salesky has been entrusted with $1 billion by Ford Motor Co. and given a mission to bring the pioneering American automaker into the age of driving robots. All that money to figure out tomorrow from a company feeling more than a little insecure about its present comes with an unavoidable psychic weight. It becomes top of mind when Salesky looks down at the floor.
“It’s not like I lined the place with marble,” Salesky says as he walks through Argo AI’s Ford-funded offices, spread across two floors of a new building in a hip section of Pittsburgh. “I don’t want everybody to think that we’re so well-funded that we can pour money on things ridiculously. So we’ve got polished concrete floors instead of beautiful tile like a law firm. This is not a law firm.”
The investment in Argo AI, a self-driving startup, is Ford’s way of trying to catch up to Google affiliate Waymo and General Motors in the driverless derby.
Ford took a majority stake in Argo last year and expects to deploy autonomous vehicles running its system in a money-making business by 2021. But this move to make up for lost time in next-generation technology comes as Ford is beset by plunging profits, investors stampeding out of a stock that has fallen 23 percent this year, and a credit rating that’s now one step above junk. Wall Street is openly questioning the leadership of CEO Jim Hackett, after only 16 months on the job, and predicting the dividend will be cut.
After Hackett announced an $11 billion restructuring in July, a Morgan Stanley analyst asked the CEO if he’ll even be around to see it through. “Hell yes,” he responded.
Salesky and his founding partner, Pete Rander, certainly have the pedigree to play saviors, with backgrounds in Google’s and Uber’s self-driving units. And there might even be an opening as the autonomous-vehicle industry comes off the sugar high of early, inflated expectations and grapples with the serious technological and manufacturing challenges that remain before we turn the wheel over to the robots.
The whirlwind formation and funding of Salesky’s 22-month-old company is grounded in an epiphany he had while engineering Google’s earliest self-driving prototype: Silicon Valley cannot realize the dream of driverless cars on its own. It needs the help of a “legacy automaker,” as old-line carmakers are disparagingly known in tech world. Salesky is now determined to ensure an American automaker becomes one of the power players.
Salesky is a blue-collar engineer in a field dominated by Ph.D.s from Stanford and Carnegie Mellon. The 38-year-old native of working-class Detroit has a fresh face and mop of red hair that, combined with his tech chops, make him seem much younger.
He took an unconventional path to becoming one of the masterminds of autonomous cars. “I grew up in a small community of very, very hard-working people and grew an appreciation for what a hard day’s work could earn,” says Salesky, whose father worked at a steel mill. “Motor City never left me.”
Eventually, he ended up in the same crucible that shaped so many leaders now plotting the future of autonomous vehicles. He competed in a 2007 Pentagon-funded contest that proved the viability of primitive self-driving technology. From there, also like many of his peers, he ultimately moved into driverless development at Google.
Since the Ford deal, Salesky and Rander quietly built up a team of 330 people and deployed squadrons of driverless Ford Fusions in three U.S. cities. It’s an impressive achievement in a short time. But his main rivals, Waymo and GM, have been at it longer and are farther along. Since Waymo began life in 2009 inside of Google, it has logged more than 9 million miles of on-road testing in 25 U.S. cities. Argo’s vehicles have driven a fraction of that, although the company doesn’t reveal specifics.
“Ford, even if they have amazing technologists and even if they are making a lot of progress, they don’t have nearly as many vehicles on the road in testing in nearly as many cities,” said Mike Ramsey, auto analyst for researcher Gartner Inc. “At a minimum, Waymo is two years ahead of Ford.”
Waymo plans to roll out the first-ever driverless taxi business later this year in Phoenix. GM promises a similar service next year, which will probably be in San Francisco. Ford’s robot rides are still three years away.
This gap can also be mapped financially. GM’s self-driving unit, Cruise, landed a $2.25 billion investment in June from Japan’s SoftBank, boosting the automaker’s stock and credibility. Waymo, meanwhile, has a deep-pocket parent company in Alphabet Inc. Argo is not set up like Cruise to take on Softbank-sized capital. Instead, Salesky says he’s seeking investment by finding additional automaker customers besides Ford.
Without Google-sized money of its own, Ford decided over the summer to open the door to outside backers --and the legitimacy that goes with big investments -- by creating a freestanding autonomous vehicles unit. Over the next five years, Ford said it is investing $4 billion in the group to develop self-driving cars and new mobility businesses, and to harness big data to provide new services in cars.
But Argo remains the crown jewel of Ford’s autonomous investments, which is why Hackett brought the automaker’s board to Pittsburgh this month to visit Salesky’s new headquarters and ride in Argo’s self-driving cars. Hackett, in an interview, speaks enthusiastically about the way Argo vehicles navigate around parked cars on narrow city streets and how the startup’s open, airy offices will attract hard-to-get tech talent.
“He’s designed a building where I would like to work, compared to where I work now,” says Hackett, a former CEO at office-furniture maker Steelcase Inc. “He’s emblematic of my dream of the way organizations ought to be led.”
Self-driving cars are expected to upend the transportation industry and become a business worth $7 trillion by mid-century, according to a report last year by Intel Corp. and Strategy Analytics. Automakers who get it right -- whether by developing their own technology or forming alliances with Silicon Valley -- will be on track to more than double the industry’s meager profit margins and build a more prosperous future. But those who fail may not have a future at all.
Hackett says Salesky’s work is “as important” to Ford as the invention of the internal combustion engine was more than a century ago. “This is going to be essential to the future vitality of vehicles,” Hackett says. “That’s why I’m making an effort to talk to and see Bryan often.”
Mapping the future
Inside Argo’s office, scores of engineers stand at desks mapping the future. Literally. On large computer screens, they plot countless scenarios of situations a driverless car could face and write software that predicts how other motorists will react.
Down below on the streets, rotations of robot Ford Fusions continuously circulate in this neighborhood teeming with bicyclists, runners and pedestrians popping out along every block lined with hipster coffee shops and lunch joints. Argo’s driverless cars are also running the roads in Detroit and Miami, where they’re testing driverless delivery of pizza with Domino’s and groceries with Postmates.
By year’s end, Salesky says, Argo will have more than 100 test vehicles on the road. By that time, Waymo will have embarked on something far more extensive as it continues buying 62,000 Chrysler Pacifica minivans and 24,000 Jaguar I-Pace electric SUVs that will be converted into its driverless fleet. There are already 400 volunteer families using a pilot version of Waymo’s ride-hailing business in the Phoenix suburbs.
When he’s not fixated on the modest cost of his drab office floors, Salesky shows little outward sign of feeling any pressure. But you can rile him up by asking why Argo and Ford appear to lag behind what Waymo and GM are doing on the open road.
“We have the option today to put the public in our vehicles and give them rides,” he says. He views his rivals’ closely watched deployments as limited.
Ford’s plan, Salesky argues, is far grander and will involve up to 100,000 self-driving cars -- with no steering wheel, accelerator, or brake pedal -- operating in multiple U.S. cities, delivering people and packages around the clock. To demonstrate that Ford is further along than people think, the normally media shy Salesky has begun to emerge from the shadows and share his story.
“We probably haven’t shown enough yet,” he says.
Sharing technology, going public
He might even start sharing Argo’s self-driving technology with another automaker, something that’s possible because Argo remains independent. And it could happen soon. “We are in discussions,” Salesky says. “We’re definitely interested in making something happen.”
That could also include going public. “We’re actually thinking about an IPO and making sure that we’re setting the company up for that,” Salesky says. A public stock sale doesn’t require winning a second automotive customer, he adds.
All this began in a booth in a dark corner of an Italian restaurant in Pittsburgh two years ago when Salesky shared a secret with Rander, a mentor who gave him his first job in robotics at Carnegie Mellon University. Salesky told him he had never felt at home in Silicon Valley, and now he was ready to leave Google and strike out on his own. Rander surprised his protégé by disclosing the same thing.
“The two of us at dinner were caught flat-footed, thinking, ‘Whoa, I didn’t expect this,’” recalls Salesky. “That was the spark.”
What followed was a speed round of meetings with venture capitalists and prospective automotive clients before Ford signed on to hand over the keys to its self-driving future to two guys with no staff, no assets, and a company name borrowed from the first ship in Greek mythology.
“We were looking for a car company that understood how difficult a self-driving car is and that didn’t have unrealistic expectations of fielding a car without a human driver in a year,” Salesky says. “That would have been crazy when you’re starting from scratch.”
And then there was the delicate division of duties, in which Salesky became the boss of his old boss, Rander, 49, a software specialist who oversees Argo’s technical activities as president. Salesky’s time in the troika of leadership on the Google self-driving project made him the more well-rounded businessman, and so Rander urged him to take the top job. “I made it clear to him,” Rander says, “if this is going to be successful, you need to be CEO.”
Ford also was attracted to Salesky’s ability to see the business possibilities beyond the science project -- remarkably rare among roboticists. “He gets things beyond the software and the technology,” says Sherif Marakby, chief executive of the newly formed Ford Autonomous Vehicles LLC and also a veteran of Uber’s self-driving car program. “He gets the business and what it takes to make money.”
This is where Salesky’s road to autonomous enlightenment, without the doctorates and research fellowships, might be his most distinctive asset. He’s just a working guy with an engineering degree from the University of Pittsburgh, where his family moved when he was in high school. “He’s a workhorse and totally, totally unpretentious,” says Red Whittaker, the Carnegie Mellon professor who is a founding figure in autonomous vehicles.
Ask Salesky about his academic prowess and you get a qualified answer. “I was a good student, yeah, as long as I was interested,” he says, laughing. “Really, from high school onward, I just wanted to get the academics out of the way as quickly as possible and enter the working world.”
At the Darpa Urban Challenge, the foundational contest for today’s self-driving brain trust that was funded by the U.S. Defense Department in the mid-2000s, Salesky started a life-changing friendship with Chris Urmson. The two later bonded on eight-hour car rides to Peoria, Ill., where they created massive self-driving mining trucks for Caterpillar Inc.
“He’s very Midwestern in terms of his values,” says Urmson, who went on to become an early leader of Google’s self-driving research. “We were able to get along so well together because we had a limited tolerance for BS.”
Dealing with Krafcik
It turned out they also had a limited tolerance for John Krafcik, the automotive veteran hired by Google in 2015 to become CEO of the self-driving unit that would become Waymo. Urmson left first, in early 2016, and founded a self-driving startup called Aurora. He then spent many hours on the phone urging his pal to strike out on his own.
“It was important to me that Bryan find a place where he’s happy,” Urmson said. “And I knew he had aspirations to build something.”
What Salesky has built, at the moment, is a modified Ford Fusion that can smoothly traverse Pittsburgh’s streets in the thick of the lunch-hour crowds, deftly accounting for delivery trucks, cyclists, and joggers.
The self-driving car managed to safely handle one of those potentially risky what-if scenarios being drawn up by Argo’s engineers. While making a right turn near a take-out restaurant, a young woman looking at her phone stepped directly into the path of the Argo car. The robot hit the brakes so quietly and quickly, the jaywalker didn’t even look up.
Salesky likes to debrief visitors who go out on a test ride, and he lets out a joyous laugh after learning how his creation handled the messy real world. “I know,” he says of the bustling neighborhood near his office. “It’s like f---ing Armageddon out there!”
The same might be said of the pitched battle for driverless dominance. Despite starting from behind -- or maybe because of it -- Salesky likes his chances. “It’s just so rare in life when you can completely change the landscape of transportation,” he says. “We’re going to make this happen.”