GM’s Strategy for the Autonomous Car

President Dan Ammann talks about what the company is doing and what the timetable is

Will auto makers or tech companies lead the way to the era of the autonomous car?

Dan Ammann is helping chart General Motors Co.’s course. The company’s president for nearly four years, he spoke with Jamie Heller, The Wall Street Journal’s business editor, about the company’s strategy of developing the whole autonomous-car ecosystem and whether the new era will benefit auto makers. Here are edited excerpts of their conversation.

HELLER: About a decade ago, GM’s CEO was talking about leading the way in autonomous cars. And history intervened—the financial crisis, GM’s struggles. Why should we believe now is the time for GM to succeed in robot cars?

AMMANN: No. 1, the technology is here today to enable this to happen. No. 2, we think that with the advent of transportation as a service, and the change in consumer behaviour that we’re seeing with ride sharing and so on, that now is the time for autonomous technology to move to a point where it can be deployed in commercial scale.

HELLER: So it’s a technology play, and Google’s been out there with Waymo, and has logged many more hours of test-driving on public roads than GM. Why do you think you can succeed as the leader?

AMMANN: A demonstration-level capability or small-scale deployment doesn’t have the impact on the world that we think this technology can. Our goal is to get to the point where we can launch a truly driverless car, with all of the safety validation and everything that needs to go with that, at large commercial scale.

We’re making sure we have control of all of the capabilities necessary to do that, from the advanced software-engineering capability for the self-driving brain of the car, all the way through the integration of that into all of the sensors on the vehicle, the electrical architecture, the automotive-grade safety validation, and then the ability to reliably build these cars at large scale once we believe we’re ready to do that.

That all lets us think about this as one big integrated system, rather than just working on one piece of it or another. So we can move quickly and iterate the entire system continuously through development.

HELLER: You played with some partnerships, and there seem to be a lot of frenemy relationships between Detroit and Silicon Valley. Is the new message “We got this”?

AMMANN: The message is that this technology will change the world for the better in a really significant way. It will have a huge impact on safety and our roads. It will have a huge impact on the consumer experience. It will create time for people.

So we think therefore we have an obligation to work as fast as we can, to bring it to commercial reality as quickly and as safely as we can. So that’s what we’re working on.

HELLER: You just said in The Wall Street Journal that you’re now going to be testing in lower Manhattan.

AMMANN: We’re currently testing in San Francisco, in Scottsdale, Ariz., and in Michigan. There’s a different rate of learning and a different experience operating in each of these environments.

As you would expect, downtown San Francisco is a significantly more complex environment than Scottsdale, for example. As a rule of thumb, our cars experience about as much in one minute of San Francisco driving as they do in one hour of Scottsdale driving, just given the number of interactions with other players, between pedestrians, cyclists, other vehicles, and things going on.

In the more complex environment, our cars can learn more quickly, and we learn more quickly about the problems we need to solve. Now we’re taking that a step further, to Manhattan, which most of us would agree is the most complex and difficult driving environment in the U.S.

HELLER: Is that just a technical thing, or is there also a messaging and branding element in taking it to New York, where Wall Street is?

AMMANN: It’s all about getting the technology ready as fast as we can. Until we’re testing in the most complex environments, which are the same environments in which we ultimately think the technology should be deployed, we won’t get the level of performance that we need.

Attracting talent

HELLER: A lot of people here talk about competing for talent. How are you attracting the people you need to GM?

AMMANN: When we began to see the change in the transportation landscape a few years ago, we saw we were missing some of the specific software-engineering capability that we needed, particularly in the self-driving area. So we learned what was out there, and we ended up acquiring a company called Cruise Automation, which was a very small but very talented engineering team.

And with that team we not only bought capability at the time, but we bought the ability to recruit more talent. People see what we’re doing, and often reach the conclusion that if you want to come to the place where your technical contribution will actually be commercialized in large scale, then this is the place to come.

HELLER: What’s happening on the regulatory front? How much of an impediment or help is that going to be?

AMMANN: It’s still being formulated, but I think the promise of what the technology can do will help from a regulatory perspective. Every year 40,000 people are killed in traffic accidents in the U.S. Regulators will tell you 95% of those fatalities are caused by human error.

So if we can make a significant impact by replacing the human driver with a better driver, then it almost becomes an obligation to deploy this technology for the betterment of society.

Industry impact

HELLER: Let’s say you succeed and we’re all getting driven around by robots. Is this going to be good or bad for the car industry long term? Will we just need fewer cars?

AMMANN: We think by making transportation more accessible, lower cost, safer, we actually create the opportunity for more vehicle miles to be travelled. We think transportation becomes more readily available in general to people. So total vehicle miles travelled or total passenger miles travelled probably goes up.

While autonomous cars are seen out on the road in testing, what we still haven’t seen is what we’re aiming to achieve, which is a truly driverless car deployed in large commercial scale. But once you’ve deployed, the technology will only continue to improve. So we view that on-off switch of the beginning of commercial deployment as just the beginning of a really interesting journey, and changing that total customer experience.

HELLER: The big question is when that on-off switch gets flipped.

AMMANN: We and others are working on this as fast as we possibly can.

HELLER: You said the last prediction was five years from 2016. Are we still in that timetable?

AMMANN: I think inside of that time frame we will see some pretty interesting developments.”


Photo DAN AMMANN | ‘I think the promise of what the technology can do will help from a regulatory perspective.’