Stanford Researchers Get Aggressive on Autonomous Technologies

A team of Stanford University students are developing autonomous vehicle technologies that focus on extreme driving maneuvers. The goal is to incorporate high-speed “drifting” and other aggressive racecar-like capabilities into self-driving cars to quickly take corrective actions and avoid imminent collisions.

The team, which is working with Campbell, Calif.-based Renovo Motors on the project, showcased the technology in a modified 1980s DeLorean sports car that performed tire-burning tight donut circles at Thunderhill Raceway Park near San Francisco. A video of the event was released Tuesday in conjunction with the 30^th anniversary of the iconic Back to the Future movie and its futuristic sequel that was set on Oct. 21, 2015.

Although not capable of time travel or flying like the plutonium-fueled movie version of the car, Stanford’s DeLorean can drive itself without inputs from a driver. The car is nicknamed “Marty” after Michael J. Fox’s movie character—and ostensibly short for multiple actuator research test for yaw control.

The Stanford team is led by Chris Gerdes, who heads the university’s Dynamic Design Lab, and Jonathan Goh, a mechanical engineering graduate student in the program. Renovo, a five-year-old startup company that is developing its own supercar, supplied Marty with a torque vectoring system that includes two electric motors and transmissions to quickly shift power between the rear wheels.

Gerdes says his team is developing algorithms that identify dangerous conditions and mimic the abilities of top rally car drivers to safely navigate them. This requires sacrificing some lateral stability to allow self-driving vehicles to quickly make tight turns and avoid potential obstacles. The researchers envision self-driving cars will eventually be capable of drifting around corners better than any human driver can, precisely coordinating steering, throttle, torque and braking to push the limits of grip when necessary.