The self-driving Chevrolet Bolt is being tested by Cruze Automation in San Francisco. The vehicle has several Velodyne lidar sensors Michael we
Self-driving research and development conducted by General Motors Co.NYSE:NYSE:GM) is covered Of late, at least in the eyes of the public and the media, with activities involving the migration from internal combustion engines – ICE – to BEVs.
From an investment perspective, GM’s autonomous efforts so far have largely been a story about vast development costs and overcoming technical hurdles — and little or nothing in terms of revenue or profitability. A change is coming as GM succeeds in launching a commercial robotaxi service in San Francisco, albeit on a very small scale.
A modest start
A San Francisco-based startup founded in 2013, Cruze was acquired by GM in 2016, for a reported (though unconfirmed) sum of about $500 million. The company, which employed about 40 people when it was purchased, now employs about 3,000 people and is tasked with independent technology development and marketing for GM. GM Honda Motor Co. (HMC) took on additional investors and now owns 80% of the venture, which was eventually valued at around $30 billion.
Other enterprises, including Alphabet ( GOOG ) ( GOOGL )-owned Waymo, have been piloting and commercializing autonomous robotaxis, mostly in densely populated areas, Cruz said, a first for commercial use in a major U.S. city. California’s first license to offer self-rides in exchange for a fee was released in June.
According to Vogt, who presented at a Goldman Sachs-sponsored conference last week, customers seem generally happy with the service:
“We’ve peaked at over 70 concurrent AVs — driverless AVs operating concurrently. We’ll probably double or triple that by the end of the year. And it’s always in use. In the first half of the year, I’m proud of that statistic, like 28-day retention.” , so the returners after one month are at about 50%, and to put that into perspective, that’s as good as a mature ride-hailing company today, and we’ve only been a few months into this service.
CEO, Co-Founder and CTO of Cruise
The planned release in San Francisco was not without problems. On June 3, a Cruze vehicle collided with another vehicle while making a left turn at an intersection, resulting in minor injuries. Cruz said the other vehicle was speeding and then programmed to make left turns unexpectedly at a traffic signal. Also, that month, several Cruz Aves blocked traffic at the intersection for two hours.
Rough places
According to Vogt, the company is smoothing out the “rough spots” of the technology. For the first time in eight years, he said, technology is not an obstacle in the way of market reforms. The new hurdle is the unproven ability of Cruise to increase ridership in other cities and expand the number of vehicles it rides — with the goal of profitability.
For now, Cruz is using Chevrolet Bolt EVs for autonomous operations. The Cruze aims to transition from a traditional passenger vehicle like the Bolt to GM’s purpose-built people mover (Origin) — a strange transport that lacks a steering wheel and other traditional controls.
Cruze Base Prototype (Cruise)
“The starting point is now a real vehicle,” Vogt told a Goldman Sachs conference, “and we’re currently testing it on closed courses. Here it’s driving autonomously. It’s been a huge effort to bring this new vehicle. It’s got a new sensing and computing platform, it’s hard work.” They’ve been working on it. It’s obviously got new doors and screens. It’s got low-cost architecture, and it’s been proven using a first-in-class approach to testing, not millions of miles.”
According to Vogt, the original is designed for low cost and lasts for 1 million miles, compared to a vehicle like the Bolt that needs to be replaced after 200,000 or more miles. Cruz said he is working on low-cost sensors and custom semiconductor chips. Eventually, the startup could be sold to private consumers.
“The start is now a real vehicle, and we are currently testing it on closed courses,” he said. “It’s actually driving autonomously here. It’s been a huge effort to come up with this new vehicle. It’s got a new navigation and computing platform, which[we’ve]been working hard on. It’s got new doors and clear screens. It’s got a low-cost architecture, and it’s proven to be millions.” Miles is using a first approach, not an experiment.
Raise up
As for migration beyond San Francisco, Vogt says the software developed for the Bay Area is more compatible with other cities. The permit, which took 33 months in California, will take weeks to begin operations in Cruise’s next markets, Phoenix and Austin, before the end of the year.
“What I’m most excited about is going from zero footprint, no maps, no infrastructure on the ground — to the first profitable driverless ride in 90 days. This is something people think could take years,” said Cruise of Walmart (WMT) in Phoenix. An investor with a size of $2.75 billion – works on a package offer.
BEV technology’s rollout to the mainstream market faces an uncertain timetable, dependent on a much better charging infrastructure and questions about consumer acceptance of vehicles that initially cost more than standard ICE models. GM says it will sell 1 million BEVs (including China) by 2025 and aims for an all-BEV fleet by 2035. Much of that in the U.S. could depend on government incentives and developing a reliable supply network for batteries and minerals, he said. They ask to produce. Just how profitable BEVs will be for GM and other automakers remains a mystery that can only be solved another time.
Now that the Cruise San Francisco pilot is migrating to real-time, two more markets are about to launch, and the financial viability of autonomous technology could soon be assessed. The enthusiasm of GM’s top executives shouldn’t sway investors too much — the numbers will be available soon. But if the Cruze’s Phoenix and Austin prove fruitful, GM will no doubt accelerate its plans and spending.
So far I have been cautious about GM and remain so, rating it as holding. But I’m keeping a close eye on signs that autonomous technology continues to deliver on the promise it made a decade ago, an industry worth billions—and perhaps trillions—that will become a reality once our cars learn to drive themselves safely.
