Driverless cars have been just around the corner for nearly a decade.
Yet after ambitious promises by CEOs of automakers and startups and buy-in from investors to the tune of more than $200 billion, there is not yet one fully autonomous vehicle on public roads in the U.S. today. And it isn’t even all that close.
Ford Motor Co. CEO Jim Farley said as much recently when the automaker pulled the plug on Argo AI, a once-promising startup on which it bet $1 billion five years ago, when then-Ford CEO Mark Fields predicted driverless cars would be widespread by now.
“Profitable, fully autonomous vehicles at scale are a long way off,” Farley told investors.
While not necessarily surprising, Argo’s unwinding has had a chilling effect on the AV tech sector and thrust the future of fully autonomous driving further into uncertainty. But that does not mean AV work has stalled, said Reuben Sarkar, president and CEO of the American Center for Mobility near Ypsilanti, the state’s public-private nerve center for AV testing.
“In the venture community, you might see an autonomous winter where one decision not to invest in Argo leads to other companies being more skeptical,” Sarkar said. “But in almost every innovation cycle I’ve seen, you have a flurry of new companies, you have lots of new activity out there, and eventually there’s a consolidation.”
In the past decade, investors have poured about $330 billion into 2,000 mobility companies, with two-thirds directed at AV technology and smart mobility, according to McKinsey & Co.
The driverless car future hasn’t panned out, but those watching closely caught on early that the splashy announcements and grandiose projects were a fantasy, said Hayder Radha, director of the Connected & Autonomous Networked Vehicles for Active Safety department at Michigan State University.
“Back in 2017, there was a lot of enthusiasm for AV technology, and there was a big rush from many of the major players, obviously from the Big Three, specifically Ford and GM,” Radha said. “I think after awhile, the reality really started to set in, that we are really not that close to developing truly self-driving, safe cars within this short time.”
The reality has been challenging to reckon with in Michigan, the country’s disputed AV tech capital alongside Silicon Valley. However, the work done by the private and public sectors, and the millions of dollars invested by the state into AV-related projects, has not been in vain, said Trevor Pawl, the state’s chief mobility officer.
“I’m not going to say it’s a surprise where we’re at,” Pawl said. “Complex problems defy simple solutions, and autonomous driving is one of the hardest challenges that you can face in the world today.”
The concept of fully driverless cars captures all the attention, but that’s no longer fueling AV tech development, Pawl said.
Companies have shifted away from the cash-burning moonshot of full autonomy to making incremental improvements to advanced driver-assistance systems. Features such as hands-free driving on select freeways, lane centering and adaptive cruise control have been deployed by Tesla, GM, Ford and other automakers.
“I would say that much of the innovation and real-world deployment has shifted modes,” Pawl said. “The other change is it’s gone feature by feature versus full Level 5 based autonomy.”
Most of the action in AV tech is now happening at Levels 1-3, and it’s happening in the shadow of electrification, which has become the industry’s biggest priority. The shift has forced companies and the state to rethink their approach to advancing mobility.
The American Center for Mobility is a prime example. Sarkar took over in 2020, when it was becoming clearer that driverless cars weren’t going mainstream anytime soon. His mission has been to make sure the center — pumped with more than $200 million by the state and founding investors, including Ford, Toyota, Hyundai and Visteon since it opened in 2018 — doesn’t turn into a boondoggle.
That’s meant big changes to its business case. Now that automakers have hit the brakes on full autonomy, most testing done on the track is for driver assistance features, Sarkar said. Additionally, the center is broadening its offerings to meet changing industry demands, including EV charger testing and hydrogen fueling, which is gaining traction in the clean energy race.
“The Argo AI announcement, I don’t think that’s the tipping point where things are now going to change,” Sarkar said. “I think that’s the result of just a repositioning of where people are putting their dollars.”
Reilly Brennan, founding partner at San Francisco-based Trucks Venture Capital, said Argo’s rise and fall exemplifies the fickle nature of the business.
“We’ve probably made more investments in (AV companies) than just about anybody, which probably means I’ve made more mistakes than anybody,” Brennan said.
Around 30 percent of Trucks’ $70 million seed-stage fund is in AV companies. It’s found some winners — Aptiv’s $450 million acquisition of NuTonomy was a windfall for investors, and John Deere’s $250 million acquisition of Bear Flag Robotics was a lucrative exit. It has also had losers, including Starsky Robotics, which sputtered out in 2019.
The way Brennan sees it, AV tech development will go on after Argo, and investors will continue to put their money on the next big thing, as they always have.
“The dream that Mark Fields sketched out when he was CEO of Ford, I don’t think that happened as he described it, but that’s totally fine because our vision of what’s actually going to happen with technology is virtually always wrong,” Brennan said.
Ford’s decision to cut off Argo does not mean the company is backing away from AV technology, MSU’s Radha said. In fact, the automaker said it plans to absorb some impacted employees. Automakers cannot afford to not continue investing in AV tech and driver safety, Sarkar added.
Ford is a lead investor in Washington, D.C.-based Cavnue LLC, which aims to develop the first connected and autonomous vehicle corridor on I-94. Earlier this year, the startup closed a Series A round of $130 million.
Mark de la Vergne, vice president of project development for Cavnue, said he doesn’t see investor interest waning — just shifting.
“We’ve expanded our scope as a company to different use cases around autonomous and automated technologies,” he said. “We’re obviously still very bullish on what folks are doing at these companies and the gains that they’re making.”
Cavnue, which focuses on Level 2-4 autonomy with communication between the car and infrastructure, will launch the pilot corridor with MDOT in 2023 or 2024, de la Vergne said.
Other AV startups are also making progress in the pursuit of full autonomy. Waymo, backed by Google and Magna, recently announced that it will expand its robotaxi offerings in the Phoenix area. The company operates a factory in Detroit where it upfits vehicles including the Chrysler Pacifica, Jaguar I-PACE and Class 8 trucks. Waymo has worked on hundreds of vehicles in Detroit and brought more than 100 jobs to Michigan, according to a spokesperson.
Meanwhile, Ann Arbor, Mich.-based May Mobility, which closed a $111 million Series C round this year, said it is aiming to launch fully autonomous shuttles by next year after running tests in Ann Arbor, Grand Rapids and elsewhere in the U.S. and Japan.
That might suggest that driverless cars are closer to being a reality, but experts say it will be at least another five years before Level 4 AVs are deployed widely. The projection sounds a lot like projections five years ago.
While it seems AVs are indefinitely stuck on the cusp of a breakthrough, Pawl said he expects it to follow a trajectory similar to EVs.
“There was also a lull in EV technology, and look at where we are today,” he said. “It was sort of a dark side of the mood period. Everyone knew we were going to go to electric but there needed to be some work done behind the curtains. I think AVs are sort of entering now into that natural next stage.”
Whereas EVs are positioned as a solution to global warming, AVs are the answer to solving traffic deaths, Pawl said. Fatalities on Michigan roadways hit a 16-year high in 2021 at 1,131, according to the state. More than 10,000 people have died in crashes in the past 10 years.
For that reason, Pawl doesn’t think the pursuit of fully driverless cars will stall out.
“To me the step isn’t necessary a groundbreaking technology that’s introduced as much as it is a symphony of different smart infrastructure technologies, autonomous features, that begin to move cities and states toward zero fatalities,” he said.