Waymo CEO John Krafcik resigned yesterday, which seemed to surprise just about everybody.
Tekedra Mawakana and Dmitri Dolgov, previously the COO and CTO, will step into co-CEO positions. As of right now, Mawakana hasn’t yet updated her LinkedIn profile, which perhaps illustrates the suddenness of the announcement.
Like lots of otherfolks, I have a hard time making heads or tails of this move. Krafcik joined Waymo five years ago, in the wake of Chris Urmson’s departure. The presumptive logic was that Krafcik was a business leader (albeit with engineering chops) who could take Waymo beyond research and development, and into a business launch.
Krafcik did that, technically, by overseeing the launch of Waymo One in the Phoenix suburbs last year. But Waymo One is still small and limited, and it seems like the task of launching a product has just started.
Krafcik’s stated rationale, “I’m looking forward to a refresh period, reconnecting with old friends & family, and discovering new parts of the world,” sounds an awful lot like what a poorly performing or scandal-ridden leader might say in the wake of a forced resignation.
But Krafcik is widely admired in the autonomous vehicle industry, and Waymo is currently the only company to have launched a commercial driverless business of any sort. I’m skeptical he was forced out, although anything is possible.
Perhaps it’s simply the case that the road to fully autonomous vehicles is longer than any of us expected, Krafcik spent five years getting Waymo this far, and it’s time for a break. Like he wrote.
For the first fifteen years of my career in software, I barely used algebra. Web development requires plenty of abstract reasoning, but not higher math. Since working on autonomous vehicles, however, I constantly find myself revisiting topics I last touched in high school.
When on earth would you actually use this in “real life?” A few years ago I might have said never. Now I use it all the time.
Imagine I’m trying to calculate a vehicle’s acceleration, and the sensors tells me it’s accelerating 1 meter per second in the x-direction and 0.5 meters per second in the y-direction. What is the acceleration in the vehicle’s own frame of reference?
I’d use Pythagorean Theorem to find the answer. You can try it with an online calculator. What’d you get?
Yesterday I wrote that 2020 traffic deaths showed a big increase over previous years, probably due to distracted driver (cell phones). At the end of the post, I wondered whether advanced driver assistance systems (ADAS) are helping or hurting this situation.
I did some very light Internet searching today, and the most information I could find is this NHTSA report from 2016, which states:
“The data show that the Tesla vehicles crash rate dropped by almost 40 percent after Autosteer installation.”
In addition to this, I found some sporadic reports of Autopilot-related insurance discounts from Tesla owners and small insurance companies.
This isn’t a lot to go on — the data are dated (I forgot that “Autosteer” was a feature name), and mentioned off-hand in a NHTSA report that is really about AEB deployment. More importantly, there’s no data for ADAS options other than Tesla. Tesla is a small and controversial portion of the ADAS market.
Volkswagen has released a rendering of its 2025 self-driving car, which will be based on the ID. Buzz electric vehicle platform, which itself is a re-imagining of the iconic VW Bus. The autonomy software stack will come from Argo, which Volkswagen partially owns.
A major notable feature of this vehicle is that it is uni-directional with front-facing seats. In this respect, it resembles Waymo’s active driverless vehicles, which are based on the FCA Pacifica minivan platform, and the Jaguar I-PACE SUV platform. In contrast, both Cruise and Zoox have announced new vehicle platforms, specifically designed for autonomy, that are bi-directional, with no front or rear.
Just as the original automobiles were “horseless carriages”, designed to resemble horse-drawn carriages, it makes sense both that early self-driving vehicles would resemble traditional automobiles, and also that we will soon find form-factors more conducive to a driverless vehicle. What that more conducive form factor is, remains to be decided.
Reilly Brennan’s Future of Transportation newsletter included links to the pitch decks that Joby and Lucid used in their recent SPACs. In an alternate universe, these would have been the pitch decks presented to growth- and late-stage venture capitalists, and wouldn’t have been available to the public. Because both Joby and Lucid went public via reverse mergers with SPACs, their pitch decks are also public.
The Joby deck has forty beautiful slides and a ton of information. I found it a little hard to parse, though, perhaps because I’m not used to navigating decks of companies at this stage.
I did not a few items:
Joby plans to start generating revenue in 2024, and to reach “scale” in 2026.
The pro forma financials call for revenue growth from $0 in 2023 to $2BB+ (!!) in 2026.
At scale, they anticipate the vehicle to cost $1.3MM.
The projected returns for investors seem maybe not that great? Again, I found this hard to parse, but I think they hypothesize that if they hit their 2026 goals, and if the stock market applies a 25–30x P/E multiple (!!), then investors today would see a 20% (annual?) return. Lots of caveats.
They forsee a robust market for intra-city transport.
They will focus on “meaningfully” penetrating each city before moving on to the next, and thus only anticipate penetrating 20 cities in the next ten years.
They project an average trip length of 24 miles, with an average “passenger load” of 2.3, and a price point of $3 per “seat mile.” Each vehicle will have 4 passenger seats, so the price point per “vehicle mile” is presumably $12. For a hypothetical 24 mile trip, the price would then be $288 total, but potentially as low as $72 per passenger, split four ways. They present this as “cheaper than Uber black for an individual.”
Geely, a Chinese automotive manufacturer that also owns Volvo, announced will launch hundreds, and perhaps thousands, of satellites, in order to support V2X and V2V communication.
The launches are a little ways down the road — the current press release touts breaking ground on the facility that will manufacture the satellites.
“Geely Technology Group knows how to start the Lunar New Year right — with important news regarding its future low-orbit exploits. On February 18th 2021, its Taizhou Facility was given its license to begin the commercial manufacturing of its satellites, which will be ultimately used for realizing Vehicle-to-vehicle (V2V) and Vehicle-to X-(V2X) communications to realize full autonomous self-driving.
The license, awarded by China’s National Development and Reform Commission, essentially means that the factory, located in Geely Group’s original hometown of Taizhou in Zhejiang Province, can begin production. When production begins, at present planned for October of this year, the facility will have an estimated production output of over 500 satellites per year.”
In an interesting twist that I hadn’t thought about until now, Geely categorizes these satellites as “new infrastructure.” There’s been a lot of talk in the automotive world about China’s ability to build infrastructure much faster than the US’s, and the advantages that may or may not bring. But I had always assumed this meant infrastructure on the ground. I hadn’t really thought about satellites as “infrastructure.”
The Geely press release is pretty sparse and focuses on V2X communication as the goal, but an article in SpaceNews suggests that the satellites may also foster an alternative and more accureate form of GPS / GNSS. That would make sense, as I typically think of satellites as being useful for receiving data on the ground, but not so much for sending data to the satellite. V2X would require two-way transmission, but navigation systems typically only require one-way reception of data.
GPS has been run more-or-less as an international public service by the US government for decades. Attempts to augment it have typically relied on ground-base supplementary broadcast stations, but those are hard to scale and are easily blocked by hilly terrain. If a private Chinese automotive company controls the next generation of navigation satellites, that would be a big change with potentially big implications.
I am spending most of today “in the field,” at The Villages, San Jose, where Voyage is preparing to launch a driverless robotaxi service for senior citizens.
This is my first trip to The Villages, and my first time riding in a Voyage vehicle. I love it!
I’ve been fine-tuning the brake control parameters for our 3rd-generation vehicles, largely from home (everyone works from home right now), using data collected in the field by our operations team. Today was an opportunity to ride along in the vehicle itself and see how well the brakes perform.
Riding in the vehicle is such a different and more visceral experience than sitting at a desk, working with data files on a computer. I could really feel what was comfortable and what wasn’t, in a way that normally gets relayed second-hand from our operations team.
And, since this was my first trip to The Villages, I got a much better understanding of the streets and the environment where we operate. Reviewing top-down maps or even vehicle sensor feeds just doesn’t provide the same level of context and being in the operational design domain (ODD).
This is a big part of what makes working on self-driving cars so much fun!