Autonomous Engineer

Dan Williams, director of advanced driver assistance systems and autonomy in the Indiana plant of Germany’s ZF Group, is working to develop self-driving trucks where “the vehicle can control steering and acceleration/braking, but the driver remains behind the wheel.” (Image by Roy Molter)

Dan Williams, ’85 ENG, MS ’87 ENG, realizes the irony: He got into the automotive industry 34 years ago because he loves to drive, but his pioneering work in autonomous automotive technology could take him—and all drivers—out from behind the wheel. As director of Advanced Driver Assistance Systems and Autonomy for ZF Group (ZF Friedrichshafen AG), Williams holds 36 patents for computer-controlled chassis systems. Here, he weighs in on the advancements and the future of self-driving cars.

How close are we to realizing fully autonomous driving?

In commercial vehicles, we’ve achieved driver assistance, such as adaptive cruise control, and we’re moving to the next stage where the vehicle can control steering and acceleration/braking, but the driver remains behind the wheel. Passenger vehicles have reached that next stage.

In what type of environment would a self-driving car work best?

A repetitive route that is well mapped, so that the autonomy has a good idea of where it is on the route and what kind of interactions to expect. Being on the highway, for example, is a fairly predictable situation. It’s more difficult to go into an urban environment where you have pedestrians and drivers who aren’t necessarily following the traffic rules.

People are looking for autonomy to change the idea of ownership, enabling transportation-as-a-service: an autonomous taxi fleet. When you want to go someplace, you just get on your iPhone.

What challenges still need to be resolved for self-driving cars to become everyday transportation?

In the short term, the industry is putting in a lot of time into getting the self-driving system reacting correctly in less-than-ideal conditions and rare occurrences in the environment. Something as commonplace as the inconsistency from state to state in how dashed lines cross highway exit lanes affects a car’s lane-keeping system.  Longer term, there will need to be a lot of changes in infrastructure and in the way we use, maintain and service cars.

What would be the greatest benefit of widespread adoption of self-driving cars?

Their potential to improve safety. A huge percentage of accidents are caused by driver error. Autonomy has the potential to make our highways a lot safer than they are.