You are here
Q&A: TG chats to Mercedes’ autonomous driving chief
A self-driving Benz by 2020, with no Tesla-esque mishaps. Can it really be done?
Christoph von Hugo is the manager of driver assistance systems at Mercedes. But he should really have a cooler job title, like emperor of the autobots. He’s at the forefront of Mercedes’ efforts to lead the autonomous driving chase, fighting not just Tesla, but Google, and possibly even Apple. Here’s what he had to say at the 2016 Paris Motor Show.
TG: You’ve already got steering pilot, radar cruise control, drowsiness assistant – where’s the new ground to break for the next big tech leap?
CvH: Two things: safety and reliability. Today’s driver assistance systems are there to help in rare cases when the driver is inattentive. They’re not there in 100 per cent of cases. Now we move to autonomous driving and we have to be able to cover 100 per cent of scenarios. Taking that step from say, 90 per cent reliability to 100 per cent is big. Even 98 per cent reliability is not sufficient.The message is we don’t need to do anything different, but everything needs to be much more reliable than today so we can tell the driver with confidence ‘hey, you don’t need to put your hands on the steering wheel any more’. That means improving the reliability of sensors, of actuators, of electrical systems.
Today if your power steering fails, it’s designed in a way that you have your hands on the steering wheel and you can handle the extra force. If we tell a person you don’t have to hold the steering wheel, then the autonomous steering has to work 100 per cent of the time. Granted, I don’t know if anyone has ever had a power steering fail, but in one of 10,000 cases that might happen. And that one in 10,000 case would not be acceptable in an autonomous vehicle.
TG: Have we been too quick to call cars autonomous, in the wake of the Tesla Autopilot crash, and imperfect Mercedes and Volvo self-driving cruise?
CvH: I don’t think so. It’s good to have an open discussion, and to tell people where we’re heading and the benefits. Of course it’s important not to mix that up and have too much hype. This has been an issue with EVs. Six to seven years ago, some people were saying that EVs would be the only cars on the road, then that wasn’t true, and that makes investment hard to come by. We have to work on expectation management.
Today’s systems have already improved safety a lot, but you have to make sure specific drivers know what their technology can do and how to use it. Our steering pilot option has gone down very well and we’ve had no trouble since 2013 with two billion miles covered in the US alone. No-one’s car drove off the road.
TG: With all this in mind, is it plausible we’ll see a fully self-driving car by the end of the decade?
CvH: Certainly you won’t see a Mercedes without a steering wheel, but we will have a highly autonomous car by the end of the decade.