Nov 02, 2022
Growing Doubts About Fully Autonomous Driving
Ryunsu Sung
Autonomous driving startup Argo AI, founded in 2016, is shutting down. The company, which raised more than $1 billion (about 1.4 trillion won), will have parts of its business absorbed by Ford (NYSE: F) and the Volkswagen Group (ETR: VOW3).
At an all-hands town hall meeting held last Wednesday, employees were told that some of them would receive job offers from the two automakers. Others are expected to be laid off.
In its official statement, Argo AI said:
In coordination with our shareholders, the decision has been made that Argo AI will not continue on its mission as a company. Many of the employees will receive an opportunity to continue work on automated driving technology with either Ford or Volkswagen, while employment for others will unfortunately come to an end.
In other words, after discussions with its shareholders, the company decided not to continue operating Argo AI as a standalone business, and stated that many of its employees will go on to develop autonomous driving technology under Ford or Volkswagen.
In its third-quarter earnings report released last Wednesday, Ford announced that it had made a strategic decision to reallocate resources away from autonomous driving systems for robotaxis and the like, and toward developing “advanced driver-assistance systems.”
The company also reported a quarterly net loss due to a non-cash impairment charge of $2.7 billion (about 3.8 trillion won) related to the dissolution of Argo AI.
The most likely reason Argo AI is being wound down is that it failed to secure new funding.
Ford’s CEO laid out the company’s stance on autonomous driving as follows.
But things have changed, and there’s a huge opportunity right now for Ford to give time — the most valuable commodity in modern life — back to millions of customers while they’re in their vehicles. It’s mission-critical for Ford to develop great and differentiated L2+ and L3 applications that at the same time make transportation even safer.
In short, Ford now believes that, for its customers, developing intelligent driver-assistance systems at the L2 and L3 levels creates more value than pursuing fully autonomous driving technology.
Commercializing autonomous vehicle (AV) technology is proving to be far more difficult than many expected. Even Tesla’s (TSLA) Full Self-Driving beta, probably the best-known “autonomous” system among the general public, is classified as L2 — a driver-assistance system. In other words, it is not full autonomy.
According to the article, Austin Russell, founder and CEO of Luminar Technologies (LAZR), which makes the LiDAR sensors used in autonomous driving systems, briefly explains why he considers himself a skeptic on autonomy.
It may sound ironic for someone working in this industry to say this, but it’s true that I’ve called myself the chief autonomous vehicle skeptic. Our company’s biggest bet was that it would take a very, very long time for true autonomy to arrive — the kind of autonomy where robotaxis drive around on their own, picking people up with no driver. The problems we had to solve to get to full autonomy were extraordinarily complex, and people underestimated that challenge by tens, hundreds, even thousands of times. Even today, full autonomy is still out of our reach. There is no commercial robotaxi service in operation that is truly based on full autonomy. That doesn’t mean full autonomy is impossible. Someday it will be, but early on the dominant assumption was that we’d get there in just a few years if we were willing to wait.
Everyone was saying that by around 2021 we’d be welcoming our robot overlords with open arms, they’d be driving us everywhere, and private car ownership would be a thing of the past. Obviously, that didn’t happen. That’s exactly why we at Luminar Technologies focused from the beginning on the multi-trillion-dollar incumbent auto industry. The goal is not to eliminate the driver, but to empower the driver. That’s how you build safer cars and save drivers time. It makes an enormous difference. From a go-to-market perspective, we made the right call, and that’s why we’re still here. We’ve achieved more commercial success than any other LiDAR company, and probably more than most autonomous driving companies as well. We’ve won most of the game.
What Austin Russell seems to be arguing is that today’s obsession with full autonomy is economically irrational.
Tens of trillions of won — if not more — have been poured into developing autonomous driving technology, yet no company has succeeded in commercializing it at scale. Even companies like Waymo and Cruise, which are running commercial pilot programs for geo-fenced robotaxis, frequently face complaints from passengers and other drivers that their vehicles “can’t drive.”
Even Tesla, which sells its driver-assistance software under the name Full Self-Driving in beta form, is still stuck at L2 — the level of mere driver assistance.
Ford’s decision to scale back its investment in a full-autonomy startup and focus instead on improving existing driver-assistance features suggests that
the market’s enthusiasm for fully autonomous driving is starting to cool.
Newsletter
Be the first to get news about original content, newsletters, and special events.