Smart Takes

Developing the brain of a driverless car

Developing the brain of a driverless car

Posting in Technology

Researchers from Stanford University are monitoring the brains of race car drivers to improve computer algorithms for driverless cars.

Vintage car enthusiasts and racing fans will gather tomorrow for the Rolex Monterey Motorsports Reunion races in Monterey, California. And scientists will be on the sidelines, too.

A team of researchers with Stanford University's Revs Program will be scanning professional racing driver Brian Redman’s brain in an effort to improve computer algorithms used in driverless cars.

Redman -- who participated in a similar study for Stanford earlier this year -- will be outfitted with biological sensors to record his brain waves, heart rate and body temperature as he zooms around the track. His car, a 1996 Ford GT40, will also be equipped with devices to measure the motion, acceleration and speed of the automobile.

Then the mechanical and biological data will be compared and analyzed to figure out which driving tasks demand a lot of mental effort and which ones are more reflexive. Researchers say this is important information for determining how to build a driverless car.

"We need to know what the best drivers do that makes them so successful," Stanford University associate professor of mechanical engineering Chris Gerdes said in a statement. "If we can pair that with the vehicle dynamics data, we can better use the car's capabilities."

Research in driverless cars has been on the rise. To date, Google has the only driverless car licensed for test drives on the roads in Nevada. But, as Time magazine reports, there is a catch: “Two people must be present in the vehicle — one behind the wheel and one in the passenger seat.”

Gerdes said his research will help to determine how best to create a cooperative system between humans and vehicle.

He offered this example for clarification: Imagine a human is driving down the road and they see a ball. Based on human experience and the ability to process visual scenes, a human would also be on the lookout for a kid. A car, however, would most likely be only concerned with the immediate obstruction.

By studying human brain activity while driving along with mechanical data on a car’s performance, researchers will be able to find the optimal method of sharing control.

“There is a possibility of using strengths of human and strength of machine together in driving,” Gerdes said.

via Stanford University

Share this

Amy Kraft

Weekend Editor

Contributing Editor Amy Kraft is a freelance writer based in New York. She has written for New Scientist and DNAinfo and has produced podcasts for Scientific American's 60-Second-Science. She holds degrees from CUNY Graduate School of Journalism and the University of Illinois at Chicago. Follow her on Twitter. Disclosure