Argo AI, the autonomous driving startup backed by Ford and Volkswagen, is shutting down. The news came as part of Ford’s third-quarter earnings report and was first reported by TechCrunch. It’s a big shakeup in the autonomous vehicle space.
Argo AI was founded in 2016 by former staff of Google and Uber’s self-driving car programs. After securing $1 billion in investment from Ford in 2017 (and a further $2.6 billion from Ford and VW a few years later), it quickly became one of the most promising operators of autonomous vehicles. Just this year it expanded its testing cities to cover Miami, Austin, Palo Alto, Detroit, Pittsburgh, Washington D.C., and internationally in Munich and Hamburg, Germany. It had even partnered with Lyft to offer driverless (with a safety driver) ride-hailing, and Walmart to offer driverless grocery delivery.
Given the traditional car makers supporting it and how seemingly advanced its testing was, Argo AI’s sudden shutdown is a bit of a shock. This wasn’t some billionaire making bold claims that never materialized and are now under criminal investigation—it was a serious self-driving project from industry veterans.
The issue for Argo AI seemed to be the timeline (and funding) for developing fully capable autonomous vehicles. In its third-quarter earnings report, Ford said that when it initially invested in Argo AI, the company had planned to bring a Level 4 advanced driver-assistance system (ADAS) to market by 2021—essentially a car capable of totally driving itself in some specific situations, though with human override still possible. That didn’t happen and Argo AI was unable to attract additional investment earlier this year.
On top of that, Ford CEO Jim Farley says in the earnings report that “profitable, fully autonomous vehicles at scale are a long way off.” Ford (and VW) have apparently lost faith in Argo AI’s ability to deliver a self-driving car program in any sort of reasonable timeframe, or at least without losing much more capital.
This doesn’t mean that Ford and VW are giving up on any form of driver automation, but they are refocusing their priorities. Farley says that although he optimistic that there will be L4 ADAS systems in the future (and that Ford will be able to buy one instead of having to build it), Ford will focus on developing lower levels of ADAS—specifically L2+ and L3 which include things like automatic lane following cruise control, automated lane changing, and other highly automated features that fall short of full self-driving. According to its press release, VW is still developing L4 ADAS systems, just not with Argo AI.
What this high profile casualty means for the rest of the self-driving industry remains to be seen. There are still a number of companies testing autonomous cars in the US and globally, including Pony.ai, Aurora, Cruise (owned by General Motors), Motional (owned by Aptiv and Hyundai), Waymo (owned by Google-parent company, Alphabet), and Zoox (owned by Amazon). All, though, are operating geographically limited trials generally overseen by human drivers; none of them are yet profitable or operating at scale. And all are facing the same technical challenges that Argo AI failed to overcome.
With venture capital investments falling across the board, we could be facing a tough few years for full driverless car dreamers. TechCrunch has gone as far as declaring that “it’s time to admit self-driving cars aren’t going to happen,” at least not in the near future and not with any kind of ubiquity.