The hype around full-on autonomous driving has retreated about as fast as my hairline, though at least my hairline was something real that many people could enjoy. The latest victim is the Ford-and-Volkswagen-backed Argo AI, which was once of the most promising autonomous startups. It’s reportedly going to be broken up for parts and absorbed into the two parent companies as Ford reports a big loss on that investment.
As she often does, Kristen Korosec at TechCrunch has the details. It sounds like employees are getting a healthy severance, but the timing of this is extremely curious. This comes while GM-backed Cruise is trying to expand into other cities, though Tesla is reportedly under investigation for self-driving claims.
As TechCrunch notes:
Argo seemed to be gaining ground in the past year. The company’s self-driving Ford Fusion vehicles, and now Ford Escape Hybrids, were frequently seen testing on public roads in Austin, Detroit, Miami, Palo Alto and Pittsburgh, where it is headquartered. In the EU, Argo was using the all-electric Volkswagen ID. Buzz for its testing programs in Hamburg and Munich. Argo also has several pilot programs underway in Austin, Miami and Pittsburgh with Lyft, Walmart and 412 Food Rescue.
Just last month the company revealed an ecosystem of products and services designed to support commercial delivery and robotaxi operations. The products — a list that includes fleet management software, data analytics, high-definition mapping and cloud-based communication tools — stretches far beyond the self-driving system that allows a vehicle to navigate city streets without a human driver behind the wheel. Argo appeared to be telling the world it was open for business.
The answer for why it’s getting tossed aside appears to be Ford, Ford, Ford. Per a Ford press release:
“We’re asking ‘What’s best for customers?’ in everything we do,” said President and CEO Jim Farley. “Winning for customers is driving a re-founding of the company through Ford+, with high ambitions for quality, innovation, profitability and growth across all our businesses – making smart choices about how we deploy capital even as we learn and adapt.”
During the quarter, Ford concluded that the auto industry’s large-scale profitable commercialization of Level 4 advanced driver assistance systems will be further out than originally anticipated – but development and customer enthusiasm for benefits of L2+ and L3 ADAS warrant dialing up the company’s nearterm aspirations and commitment in those areas. In the third quarter, Ford made a strategic decision to shift its capital spending from the L4 advanced driver assistance systems being developed by Argo AI to internally developed L2+/L3 technology. Earlier, Argo AI had been unable to attract new investors. Accordingly, Ford recorded a $2.7 billion non-cash, pretax impairment on its investment in Argo AI, resulting in an $827 million net loss for Q3.
Yikes. Emphasis mine. An accounting terms, an “impairment” means that the value of something purchased is now less than it was (or what you said it was). If Ford is saying that they’re taking a $2.7 billion pre-tax impairment that means that they’d invested more than $2.7 billion in Argo AI (they initially put in $1 billion) OR they counted the investment as earnings (maybe when Argo said it was worth $7.5 billion).
Additionally, L3 autonomous driving is currently crap and L2+ autonomous driving isn’t a real thing?
Curious times at Ford.
UPDATE: Went back in their filings to answer the above question:
Ok, fun with math time. Here is Ford’s 2021 10K SEC filing. Here’s what it says:
Argo AI, LLC (“Argo AI”). On June 1, 2020, we completed a transaction with Volkswagen AG (“VW”) that reduced our ownership interest in the autonomous vehicle technology company Argo AI and resulted in Ford and VW holding equal interests that comprised a majority ownership of Argo AI. The transaction involved us selling a portion of our Argo AI equity to VW for $500 million and VW making additional investments in Argo AI, including contributing its Autonomous Intelligent Driving company. As a result of the transaction, we deconsolidated Argo AI, remeasured our retained investment in Argo AI at fair value, and recognized a $3.5 billion gain in Other income/(loss), of which $2.9 billion related to our retained investment in Argo AI. Our retained investment in Argo AI consists of an equity method investment and a preferred equity security investment, reflected on our consolidated balance sheets in Equity in net assets of affiliated companies and Other assets, respectively.
So, Ford said at one point they put in $1 billion. Ford sold some of their holdings for $500 million to Volkswagen, who reportedly put in $2.6 billion. Ford then said it gained $3.5 billion due to their retained investment in Argo. I’m assuming this was when the company was worth $7.5 billion (as determined, I assume, by the VW investment). There’s an unknown variable here because we don’t know if Ford put in more than $1 billion overall, but it sounds like Ford got to ride Argo as a gain and most of that money has evaporated in a quarter.
Top photo Argo AI
Interesting to me mostly because I’ve seen a lot of Argo Escapes driving around my part of Austin. The company is aptly named, because those vehicles are covered with huge sensor blisters—it’s made me wonder “how are they going to shrink all that down for a commercially viable vehicle?”. I’ve had some encounters with them on my bike—they are cautious to a fault.
Huh – now you’ve got me trying to figure out what Argo has to do with sensors(?) or blisters(?)
https://en.wikipedia.org/wiki/Argus_Panoptes
Well, that was an unexpected result. The technology to replace the inconsistent, haphazard input from organic drivers is definitely there.
Okay, I can’t keep my face straight anymore.
The U.S. is the best and worst place for trying to develop full autonomy, for the same reason: Our trash infrastructure. The plus side is that it forces autonomous systems and their developers to confront chaos. But that very chaos could mean full autonomy never gets there.
Even if we could improve the condition and consistency of our roads by orders of magnitude–with obsessive standardization and maintenance of road markings, signals and signage–there will always be construction zones. And weather. And the complete unpredictability of human-driven vehicles.
Sounds like Ford has figured this out. Better late than never, I guess.
Good on them for shutting it down but I hope they’re not putting all that money into those worthless lower level systems.
Oh thank God! Someone is being sensible at last!
WELL DONE FORD !
And straight away i take most of it back for writing the most confusing pile of horseshit i think ive ever seen.Could you make it any more obvious? All that does is make you look like you’ve screwed up
Irrational exuberance at it’s finest. Conclusion that is.
Here’s a brilliant idea as the entire nation’s freight railways are about to go on strike next month….
Take half of these self-driving auto needs and work out how to do self driving trains. It should require 1/100th of the effort.
Then after you figured out trains do the same for buses with fixed routes. That should take 1/10th of the effort.
Then after you figure out trains and buses then start to think about cars.
Even floating turntables make more sense than autonomous cars. Congratulations, Ford. Here’s your sobriety pin.
“remeasured our retained investment in Argo AI at fair value, and recognized a $3.5 billion gain in Other income/(loss), of which $2.9 billion related to our retained investment in Argo AI. Our retained investment in Argo AI consists of an equity method investment and a preferred equity security investment, reflected on our consolidated balance sheets in Equity in net assets of affiliated companies and Other assets, respectively.”
This here is basically corporate nonsense speak for ‘our valuation is based on flying pig level estimates of the value of the intellectual property and has no actual relation to real money, tangible goods, or even anything of value.’
Anything like software, contracts, trademarks, etcetera goes in the “Other assets” category. Which is a grab-bag of legitimate (“we have a contract we can exercise to sell 25,000 used office chairs at $250 each”,) bullshit (“our trademark is worth eleventy bajillion dollars,”) and a mix of the two (“we spent $500M and say the software is worth $500B”.)
That isn’t to say that Argo wasn’t pissing money down the drain like you wouldn’t believe. They absolutely were. They were shoveling stacks of cash into the dumpster fire. They were paying SWE2’s on average $180k+ excluding bonuses. SWE2! Sr. Systems Engineer – not even architect – was over $230k/year.
IN PITTSBURGH.
Worse, it was a fucking yet another white bros only hellscape. Which meant they were overpaying drastically for people who would only ever under-deliver. I mean, it’s pretty damn obvious since they’ve pulled the plug saying they can’t do the job.
So basically they’re saying “we’ve lost tons of money,except we totally havent because we say so” ?
Please show the nice social worker on the doll where Argo AI touched you.
If you know of a place that I can underdeliver for over $200k, let me know!
Eh. They’re saying that VW bought half the company for $3.5 billion, so we decided that the half we still own is also worth $3.5 billion.
I *think* the “preferred equity security investment” that Ford is counting as “other assets” means that Ford loaned Argo some money.
I cannot understand why the NTSA has allowed this tech to reach public highways in it’s present form. Too many different systems not integrated with one another, no “best practices” determined, over-hyped and under explained. People have already died and there are sure to be more victims in our future. For what? Bragging rights? How about the automakers put these investments in true drivers ed?
What I really don’t understand that everyone else testing on the streets has permits for that they applied for while Tesla has no permits and doesn’t expect to get them this year but yet they are running public beta tests wherever they choose. How is that possible?
It takes a smart man or company to have the courage to admit they’ve made a big, expensive mistake. As others have said, the reality is fully autonomous cars are off in the distant future. Right now it’s like investing in a science fiction project. Ford decided to pick the lesser of two evils and write off their investment to date, instead of pouring more good money after bad into a bottomless pit.
A saying in the investment world, the first one to try and break through a wall often ends up bloodied, bruised, beaten – and broke. It’s a lot safer and cheaper to follow them through the hole they made with their sacrifice (and money).
The only thing that would have been even smarter would have been if Ford didn’t flush $1 billion on that boondoggle in the first place.
This smells like “We’re not going to pay or the development of ‘levitation’ anymore, at the cost of hundreds and hundreds of millions. We’re going to wait for someone else to figure it out, then ste-, I mean, license it, copy it, whatever.”
My assumption is that Ford realizes it is too far off to be nothing but a money pit and decided to take the hit, do the write down, and just say “meh, we’ll partner later once someone figures it out.”
Personally I don’t see it being anything beyond a novelty until V2V, V2E, and maybe even V2P (or at least the phones on people) communication is a thing.
Maybe I should have read that second block quote better, or I wouldn’t have made the above post since my “assumption” is basically what the quote said. I’m dumb. Don’t work and internet comment at the same time kids.
So……about that “edit comment” button?
Beat me to it.
Yup just like they licensed hybrid tech from Toyota.
With so many economic cloud on the horizon, they probably have some opportunity to pick someone for the cheap. Intel is spinning off MobileEye (a previous Tesla supplier), there are components which will reach commodity level.
This is the next stage of the hype cycle coming to pass. Full autonomous operation is at least 5 years out (and maybe, like nuclear fusion power, always 5 years out).
The key statement in the article is: “Earlier, Argo AI had been unable to attract new investors.”
The burn rate is too high and Ford doesn’t want to keep pumping cash into the dream. Not something a public company wants to keep showing up on their books. The write-down is painful enough.
Well something I can address intelligently. We all know during the Covid years car manufacturers lost billions. It’s only now when they have everything up and running that they are making those billions back. So you have an asset like this maybe it has potential maybe not. But you feel it was a loss or its gone as far as it can what do you do? You shut it down. Take the billions of dollars in losses and apply them to your billions of dollars in profits and eliminate paying taxes. While taking anything valuable and putting it in your old business with out paying anything.