Weekly Brief: AVs Depend on Evolution not Revolution

Commentary By Andrew Tolve

---

10th May 2021

General Motors will sell self-driving cars to the general public by 2030, the carmaker’s CEO Mary Barra declared on a quarterly-earnings call last week.

Then, she hedged. “Later in the decade, I believe, and there’s a lot to still unfold, but I believe we will have personal autonomous vehicles,” she said. Nothing could better capture the uncertainty looming over the autonomous vehicle revolution than Barra’s equivocation. Gone are the days of unchecked confidence, blind optimism and promised deployments.

Here is the new reality in which AVs have failed to go mainstream or even prove a real use case, despite billions in venture capital and carmaker investment. Start-ups are folding. Valuations are plummeting. The industry is rapidly consolidating. In the last six months alone, Aurora has acquired Uber’s Advanced Technology Division, Cruise has acquired Voyage and Toyota has acquired Lyft’s Level 5.

Niche AV areas 

Consolidation doesn’t mean that the end is nigh – far from it. Self-driving cars may still be the future for trucking, ride-hailing and the consumer segment and could usher in a whole new paradigm of living and moving in the world. See Nathan Eddy’s article for more on why Driverless trucks are worth the investment and how autonomy is likely to drive productivity gains across the freight industry. The same could be argued for the consumer segment but, if or when, any of this will come to fruition is now anyone’s guess.

GM has some solid ground to stand on in making a prediction for this decade. It owns Cruise, one of the leaders in the AV space, and already has widespread testing underway for ride-hailing and last-mile delivery applications. It also has Super Cruise, its advanced driver assistance system that allows hands-free driving on more than 200,000 miles of highways in North America. Super Cruise is comparable to Tesla’s Autopilot but has a superior safety record and includes infrared eye-tracking technology that ensures a higher level of driver engagement.

Super Cruise will appear in 22 models come 2023, Barra announced last week. She expects Super Cruise to eventually enable hands-free transportation in 95% of driving scenarios. For the other 5%, that’s where the LiDAR and self-driving tech of Cruise comes in. “I’ve always said we have kind of a revolutionary and an evolutionary strategy around driver assistance all the way to full level four, level five autonomy,” Barra said.

This dual-pronged approach could place GM ahead of Tesla as the front runner for delivering self-driving cars to the consumer segment. Elon Musk has written off LiDAR as too expensive and impractical for use in consumer cars. That means all of his chips are pot committed to what Barra calls an evolutionary approach. Tesla is trying to expedite that approach by beta testing its “Full Self-Driving” system on customers in its Early Access Program. Musk has promised full self-driving by the end of 2021 but early reports from the field suggest that the technology isn’t anywhere close to true autonomy.

Tesla admission

Even Tesla’s own team has admitted as much in private. Last week an exchange between Tesla and the California DMV was made public. In it, Tesla’s director of Autopilot software said that “Elon’s (claim about autonomy) does not match engineering reality.”

Other than Tesla, GM’s biggest threat on the consumer front could be the world’s top-selling carmaker, Toyota. With its acquisition of Level 5, Toyota now has a foothold in Silicon Valley and has made no secret about its desire to integrate self-driving technology into everyday cars.

Ultimately no single carmaker is likely to win this race. The most probable scenario is that advanced driver assistance systems become increasingly robust across the industry, gradually introducing autonomy to the masses, until it’s so pervasive that we all forget that we were ever waiting for a revolution to begin with.