Avoiding pedestrian collisions in a future of automated cars

Posted: 

Originally published by Colleen Herr, Center for Automotive Research (CAR)


The Ohio State University Center for Automotive Research’s (CAR) Automated Driving Lab is enabling cars to “talk” to one another to avoid accidents with other vehicles. But what about collisions with pedestrians?

At the Automated Driving Lab, Research Professor Bilin Aksun-Guvenc stands alongside the DASH car, a fully-electric autonomous research vehicle.
At the Automated Driving Lab, Research Professor Bilin Aksun-Guvenc stands alongside the lab
That’s one of the problems Research Professor Bilin Aksun-Guvenc, Department of Mechanical and Aerospace Engineering, seeks to solve. It is a serious challenge, with 5,376 pedestrians killed in the U.S. in 2015 alone — and one of increasing urgency as automated vehicles are expected to be ready for series production by 2020. When you consider the needs of people with blindness/visual impairment or other mobility challenges, the problem becomes even more complex.

“When we think about automated driving, we need to be sure to think about the elderly, the disabled, people without cars and who have mobility issues,” says Aksun-Guvenc. “Once you become older, it can take longer to cross a crosswalk. There needs to be some indication for cars to understand those situations.”

Building on recent robotics research that prevents mobile robots from colliding with people in highly populated areas like malls, Aksun-Guvenc created a “socially acceptable collision avoidance” (SACA) method that protects pedestrians and cyclists from vehicles and respects humans’ personal space. An algorithm determines a collision-free path, making very fast calculations to solve collision avoidance problems in real time.

The SACA technology taps into the interconnectivity made possible by pedestrians’ smart phones, enabling cars and phones to communicate to keep people safe. A smart cane extension is also under development through a collaboration with Professor John Lannutti and his research team at Ohio State’s Department of Materials Science and Engineering. In addition to giving feedback to its user—for example, letting them know if they have stepped outside of a crosswalk—the connected cane would also broadcast location information so that cars know a person with a visual impairment is nearby. Once the connected cane technology has reached the proof-of-concept stage, the research partners hope to test it at the Ohio State School for the Blind.

Aksun-Guvenc and her research team have tested the SACA method in simulations and in real-time test drives on Ohio State’s campus and the Ohio State-affiliated Transportation Research Center in East Liberty, Ohio. They will soon begin testing at either Easton Town Center, the large suburban shopping area on Columbus’ east side, or the new planned autonomous vehicle pilot route on West campus, where Aksun-Guvenc and her colleagues are also piloting a Smart Shuttle proof-of-concept program.

The opportunity to see the results of her work and its impact on real people is what motivates Aksun-Guvenc to continue.

“We don’t do it for the sake of technology, but to help people. We want to make sure driving is safe, that no pedestrian or passenger in the car is hurt. And we want to improve mobility for everyone."

Category: Faculty