A Ford/Argo AI vehicle drives before the company is shut down, in the big bad story of the year
As we enter 2023, these were the big stories in self-driving cars for 2022. You an also check out my summary of early 2022 for more details on that quarter, but these are the game-changers for this year.
There’s no doubt this was a year of huge ups and downs. A year where some gave up and others doubled down. A year where the stock market and VC funding took a massive nosedive and companies mostly shrunk, but also a year of great technical progress.
Usually I do this as a countdown to the biggest story, but I’ll spoil it right now and say the big story was the incredible contrast of the good and bad news. So I’m instead going to tell you the good, the bad and the not particularly ugly.
First, the bad news.
While there’s been talk of a robocar winter for some time, ever since automotive OEM players missed on their invented predictions of having cars in 2020, things hit a low note in 2022. The events I am about to list triggered a spate of articles in some press declaring the death of the robocar. It became fashionable to now predict such cars were decades away, and that untold billions had been spent in the market frenzy with little to show for it.
One company that dealt with some of the bad news was one of the biggest newsmakers of the year, GM’s Cruise unit. Cruise had their vehicles out on the street each night, but didn’t make a fan of the city officials. Cruise had a spate of incidents where their cars would just get “stuck,” frozen on the roads with blinkers on, sometimes several of them at the same time. Some people got angry, or pretended to be angry, but the reality is that while Cruise clearly had some strange bugs, if blocking traffic in the middle of the night is a calamity, we would have banned human drivers long ago.
Cruise got a lot of attention and went viral in a bad way when one of their cars was pulled over by police for not having its lights on. That was strange enough, but a strange result of Cruise’s special programming for just such events took it viral. The car is programmed to attempt to pull off into a safe space when it is determined (by remote operators or the car) that police are pulling it over. But it decided that only after it had stopped and police walked up to it – and then it darted across the intersection to pull over in a better spot, seemingly trying to make a break for it.
For decades a very common question of people first learning about robocars was “what happens when police want to write it a ticket?” The police didn’t write a ticket in this case, but the car had a protocol for being pulled over. It displayed a phone number in big print on its screen and the police phoned it, talked to Cruise operators and resolved the problem. A remote operator had mistakenly turned off the lights. If it had been a software bug, Cruise would have fixed it quickly and it would have never happened again. When police write a ticket for a human driver, it doesn’t do much to stop other drivers from doing the same thing. Not so with robots.
The Fire Department was more upset when an unusual situation had a Cruise vehicle stopping in the oncoming lane – as it was programmed to do – when a fire truck approached. Problem was it stopped in a place opposite a double parked truck, and the road was blocked, and the Cruise could only back up into the intersection, which it didn’t want to do. The Cruise crew were about to fix it when the truck moved on. The Fire Department didn’t like even this short delay at all, but in reality I think that robocars will make the roads much better for emergency crews than they are now.
Cruise had a much more serious problem with their first injury accident. They were trying to turn left and started their turn while a car was coming at them in the right turn lane very fast. The car didn’t turn and tried to go straight through. The Cruise car froze – a mistake – and the oncoming car clipped it on the corner. Injuries were not serious but Cruise somehow arranged for the silence of all involved so we don’t know a lot more. Police didn’t put the Cruise car at fault – the other car was speeding and in the wrong lane – but it could definitely have done better. Cruise issued a “recall” to fix this – but it was just a software update. A recall doesn’t make sense when the carmaker owns the cars.
The real bad news was business news. At the start of the year, Cruise fired their CEO, Dan Amman, over a dispute with GM CEO Mary Barra over whether the company would go public and whether it would focus only on self-driving or feeding useful tech to GM. Founder Kyle Vogt returned to the role of CEO.
But there was much worse business news than that. Casualties were left on the side of the road as shuttle companies Optimus Ride and Local Motors shut their doors, along with Amazon’s Scout delivery robot project. We saw significant layoffs at high-flyer Nuro – even though they started manufacturing, and also Hyundai and Aptiv’s Motional joint venture. Trucking company tuSimple saw major chaos in the C-suite and boardroom, and ended the year by laying off ¼ of their staff. Almost no company was unscathed by layoffs. The many companies in the space who went public by SPACs saw massive devaluations, and lidar pioneer Quanergy, which I helped get started, went bankrupt in December.
We also law Audi shut down their Artemis unit – all the car OEMs except GM and Tesla have been scaling back their self-driving efforts in some way, which is part of what inspired the declarations of doom in the press.
We also saw leaks that Apple was scaling back their project, with a potential release in 2026 but with pedals and a wheel, like a regular car. Apple, of course, has made no comment.
But no question the biggest and baddest story was the shutdown of Argo AI, a startup funded by Ford and later VW. Argo was one of the big players, with billions invested, though it was not one of the leaders. Ford simply announced they were shutting it down. They rehired about 700 of the staff into Ford and reabsorbed the IP.
That OEMs were getting cold feet was no surprise. They never wanted a fast robocar revolution and are content to work on ADAS. What was more surprising is that nobody wanted to buy Argo for a song. It was reported that Amazon, which already owns Zoox, was interested in it for delivery vans but nothing happened. If they shopped it to anybody else, nobody bit in this slow market.
Tesla’s stock fell greatly, and they got new investigations by government agencies. They got a poorly ordered recall on the ability to do rolling stops at empty intersections and investigations into hitting emergency vehicles and other Autopilot accidents. Impressively, there are not reports of serious FSD accidents. They also had Dan O’Dowd, the wealthy CEO of Green Hills Software decide to spend a lot of his money on a campaign against Tesla FSD, running full page ads in the New York Times and pretending to run for U.S. Senate to do political ads. That didn’t seem to dent Tesla much, but the biggest wounds came from Elon Musk deciding to buy Twitter, and to seemingly destroy it and his formerly stellar reputation in the process. This took his eye off the ball at Tesla and scared away at least some customers.
But there was good news
Even with all the layoffs, shutdowns and market haircuts, many players pushed forward. In particular, the teams that were startups and tech companies rather than car companies. I include Cruise here, which began as a startup and is mostly run as one.
Several of the trucking and delivery companies had big improvements this year. Starship, where I’m a stockholder, hit the 4 million paid autonomous delivery milestone and expanded to a whole raft of new service locations, mostly universities. Gatik, a “middle-mile” delivery company is now doing regular runs with nobody in the truck between depots in Arkansas for Wal*Mart and in Toronto for Loblaws. Regular commercial service with no safety driver aboard in a big heavy truck is nothing to sneeze at. While the long-haul companies that went public by SPAC have seen major hits, Einride, which makes trucks that can’t even hold a driver, just closed an additional $500 million investment round. Kodiak, a long-haul company led by Don Burnette who worked on the early Waymo team, just announced a big military contract. Meituan, a Chinese company similar to Nuro, has reported good growth due to Covid lockdown deliveries.
MobilEye was bought by Intel for $16 billion several years ago. Late in 2022 they went public at a valuation not too different from that, which at first seemed like bad news. They popped the first day and MBLY is up over 75% while the rest of the market has been down or flat.
While Tesla has certainly had its challenges, their FSD prototype, mistakenly labeled as a beta, is definitely showing improvement. While I still can’t get it to do a significant drive without interventions, the distance between interventions is improving, and the system is getting better at turns. I still wouldn’t expect any of Elon Musk’s regular “next year, and I mean it this time” predictions to come true, and all his attention is on Twitter.
News from China was of continued growth, and opening up a few more cities to no-safety-driver robotaxi service. There are now over a dozen Chinese cities with robotaxi service. Baidu showed off their custom robotaxi design – it still has a wheel
due to the law but it’s designed so it can be removed later. They say it will cost just $37,000 to make. Waymo also announced it would use a new robotaxi built by Geely in China.
Zoox is growing. Amazon is building them a factory and spending $1 Billion per year on them. They are hiring when the rest of Amazon is cutting back with the downturn.
But the big story was the progress and expansion of the field’s top two teams, namely Waymo and Cruise, with a lot of it coming near the end of the year.
In 2019, Waymo launched service in the Phoenix suburb of Chandler with no safety driver in the vehicle. This year, Cruise began service at night in San Francisco to a subset of the public, and Waymo joined in later with 24 hour service (with safety drivers in heavy weather.) Cruise got permission to charge money as well.
Near the end of 2022 we saw big improvements. Waymo started taking the public, and expanded their no-safety-driver territory to the whole city, though the NE quadrant including downtown is only for employees. Waymo also expanded to downtown Phoenix, and opened up to public use there, and this this month doubled the service area. As a plus, they will also take you to Phoenix Airport’s Skytrain that runs to all terminals. Waymo will soon charge money in SF. I rode in their car in SF, and while no single experience can tell you a system is good, it can quickly reveal if it’s bad, and the ride was smooth and well performed. Waymo now has 700 vehicles in operation across its fleet.
Cruise announced in October that in 90 days they would open in both Austin and Phoenix, and they pulled it off, though it’s still only at night, and the Phoenix area is also the easy suburb of Chandler and the Austin area is small. Even so, it shows they can expand and do it fairly quickly.
Waymo has also said they will soon operate in Los Angeles, and just got their DMV permit. There are now many cities around the world where people can hail a robotaxi ride, sometimes with no safety driver in the vehicle. This map shows where they are – mostly in the USA and China, but growing regularly.
Not so Ugly
This year I began a YouTube channel. You might enjoy some of the videos there, which all come with text articles for those who prefer that. This includes a 2 part series on why you aren’t riding in a robocar quite yet, plus a review of Tesla FSD that gave it an “F” – and still does, even though it’s better. I also covered why Tesla should get maps, and why they might be putting in imaging radar. There was also a lot of coverage of electric vehicles and the charging business, as well as some on the Air Travel overload, e-VTOL aircraft and much more. Plus plenty about Tesla, as seems to be always the case.
Tune in and read in 2023 for what should be an interesting year. With a bit of luck, we’ll bounce back from the market downturn, though not without losing a few companies on the way. The business models will get more refined and that will inspire a new round of startups if they can get funding. We’re getting close to the time where a real commercial robotaxi service arrives, scaled to make money, at least on a gross margin basis. Then the land rush can begin. The same will happen in China in the next couple of years.
Stay connected with us on social media platform for instant update click here to join our Twitter, & Facebook
We are now on Telegram. Click here to join our channel (@TechiUpdate) and stay updated with the latest Technology headlines.
For all the latest Technology News Click Here
