Inventing a Car That Drives ItselfOctober 13, 2017
As Principal Engineer for Autonomous Driving at MBRDNA in Silicon Valley, Dominik Nuss has a goal that’s easy to understand: Along with his team, he’s helping to invent a car that can drive itself. Nuss and his team provide the algorithms that help autonomous cars head down streets on their own.
“I’m doing something challenging but with a clear application — even if the engineering behind it can’t be explained easily.” Nuss adds, “Having a clear goal like this is really satisfying. A lot of engineers at other companies work on projects that are really vague. But there’s nothing vague about what I do: I’m helping to put a new kind of car on the road that will change the way we live.”
Working in a Field That Changes Nearly Every Day
While he was studying autonomous driving as a graduate student at Ulm University, Nuss won a competition for building a small remote-controlled vehicle that could drive on a test track. Upon getting his Ph.D. in May 2015, Nuss began his career the following June with MBRDNA in Silicon Valley.
Nuss says there’s always something new happening at MBRDNA, because the company sees itself as a technology company aimed at improving the automobile. “There could be a new development in machine learning or eye-tracking detection — anything, really — and it happens constantly here.”
Nuss loves the variety of expertise required to take his field to the next level. “We need a huge variety of people. We need physicists, electrical engineers, mathematicians, computer engineers — every kind of technology person out there.” He says the MBRDNA team is a lot like a top professional soccer team. “When you put together a soccer team, you need a lot of specialists: You need someone who can run fast, someone great at scoring, maybe someone who is tall enough to block a shot and then someone else who can watch the goal.”
Similarly, the AD team at MBRDNA has specialists who are experts in their field — from laser technology to web development to camera detection. The team also has generalists on-hand who have a broader perspective. Teams win games, but there’s a big difference between playing soccer and developing a car that can drive itself: there’s no playbook.
“We’re breaking new ground in this field. There’s no product we can buy right off the shelf that will solve our problems. Every day, we deal with a puzzle — and it’s a big challenge.”
Every Day is Different
Nuss doesn’t have an average working day: He could be test-driving a vehicle or working alongside internal and external specialists. He says the job for everyone at MBRDNA is the same, in that they must all push the limits of technology — after all, an autonomous vehicle must detect a complex environment like a city street filled with bike riders, other cars and pedestrians. It must also safely speed down a highway — and be always ready to stop quickly.
It isn’t easy to get a car to drive itself. “You have to bring all of the technology together,” he says. “The on-board cameras can tell you what’s in front of the car, but they’re not very good at estimating the speed of another car in the next lane — that’s reserved for the radar, which will tell you how fast that car is going. My team brings it all together to make sense of the world around the car.”
Nuss says that’s what he likes the most about the job: Nothing is spelled out. “We need to create solutions on our own, but also have to work together. It’s not a loner’s job. Here, it takes a lot of experts, willing to work together.”
What do all of these different experts have in common? “I’d say we’re all extremely motivated. And that we’re all dreamers. We really do want to change the world. Just imagine if your grandmother could get around on her own with an autonomous vehicle. You’ve changed the world, right there.”
Ready for some #Urban_Automated_Driving? Take a ride with us in one of our self-driving vehicles.
Posted by Mercedes-Benz Research & Development North America, Inc. on Wednesday, September 27, 2017