General Motors (NYSE:GM) said that it will host a webcast for investors on Nov. 30 in which the company will "share our vision for an autonomous future."
Having watched GM for years, I think this is likely to be a significant event. I think GM has some big news to announce -- such as how and when it will begin to bring autonomous vehicles to market -- in much more detail than it has so far.
I don't know exactly what to expect, of course. But in a presentation to Wall Street analysts last week, CEO Mary Barra dropped some hints that give a fairly clear picture of how GM will bring autonomous vehicles to market -- and some tantalizing hints as to when.
GM is very close to mass-producing self-driving cars
The big takeaway from everything Barra has said about autonomous vehicles in recent months is this: GM isn't dabbling. The company plans to make a lot of self-driving vehicles in the near future, and it -- perhaps uniquely -- will soon have the ability to do so. Here's how Barra put it last week:
Electric self-driving cars will save millions of lives, and it significantly accelerates the transition to sustainable entry to the world of zero emissions, but only when they're deployed at large scale. So our focus working with [GM's self-driving subsidiary, Cruise Automation] is to make sure we're positioned to rapidly deploy self-driving cars at scale.
Barra explained that the key to deploying self-driving cars "at scale" is to have a safe self-driving system, a vehicle that's well-suited for it, and the ability to manufacture those vehicles in large numbers.
GM has the vehicle. Cruise Automation has designed its self-driving system to be deeply integrated into the Chevrolet Bolt EV, GM's pioneering long-range electric car. The self-driving Bolt differs in significant ways from the "regular" Bolt -- Barra said that about 40% of the Bolt's parts are different in the self-driving version -- but it can be built on the Bolt's regular assembly line. That means GM can crank them out by the thousands as soon as its suppliers are ready.
When will that be? It's not clear, but Barra emphasized that many of the unique parts in the autonomous Bolt are proprietary GM pieces. GM recently acquired a company called Strobe that had developed small, low-cost lidar sensors; those lidar (light detection and ranging) units will presumably be integrated into the self-driving Bolt quickly, if they're not already.
Long story short: More so than just about any self-driving contender, GM owns key parts of the autonomous-vehicle supply chain, including the car itself (and the factory that makes it), the self-driving "brain" (developed by Cruise), and the one key sensor that isn't a commodity, the lidar units. That means that GM can probably start making self-driving vehicles by the thousands as soon as its software is ready.
That's one clue as to what GM might be preparing to tell us next week. Here's another: GM is using a challenging urban environment to "train" its self-driving system more quickly than it could in a suburban setting.
What GM will do with all those self-driving cars
"Although our focus is on building the product and having the product because it's the most complex problem to solve, getting that done first, we are also building a commercialization strategy that will allow us to unlock maximum value," Barra said during her presentation. "We believe this is an opportunity to participate in the biggest business since the creation of Internet, and we believe GM is in position to lead this opportunity."
But what is that "opportunity" that Barra is determined to lead? It appears to be ride-hailing in urban environments. It's not clear whether GM will partner with Lyft, in which it has a stake, or if it will instead launch its own proprietary service via its Maven car-sharing subsidiary.
Barra said that GM is making a point of racking up lots and lots of self-driving test miles in cities, specifically in San Francisco. That's something that key self-driving rivals like Alphabet's (NASDAQ:GOOG) (NASDAQ:GOOGL) Waymo unit haven't really focused on. She said a test vehicle operating in San Francisco "learns" in a minute what it would take an hour to learn in a suburban environment like Scottsdale, Arizona (where GM also tests).
The GM/Cruise system is believed to be "Level 4," meaning that it will be limited to areas that have been carefully mapped. Obviously, Cruise has San Francisco well-mapped at this point, and if GM launches a proprietary self-driving ride-hailing service, it will be no surprise if it begins in that city.
But interestingly, Barra noted that GM is working on mapping other cities, and said, "It will be 6 months or less to get ready to launch in other cities."
The upshot: The big question that GM will probably answer next week
Barra has said that GM will begin rolling out its self-driving vehicles in "quarters, not years." It seems pretty clear that those first self-driving Bolts will be put into ride-hailing service in U.S. cities, starting in San Francisco. It also seems clear that GM is close to being ready to mass-produce those cars.
How close? That's the big question. I expect that -- the timing of the launch -- to be the biggest news that comes out of next week's presentation.