If you’re going to build an autonomous, electric, drifting automotive research vehicle, why not do it with some style?
That was the thinking of Chris Gerdes, a professor of mechanical engineering, when he and his students at Stanford decided to transform a vintage 1981 DeLorean into their newest high-performance test bed for researching the physical limits of autonomous driving.
Alongside partners the Revs Program at Stanford and Renovo Motors the team unveiled the latest edition to Stanford’s research fleet. Nicknamed MARTY – short for Multiple Actuator Research Test bed for Yaw control – the car is already proving to be an excellent vehicle for student-driven research.
“We want to design automated vehicles that can take any action necessary to avoid an accident,” Gerdes said. “The laws of physics will limit what the car can do but we think the software should be capable of any possible maneuver within those limits. MARTY is another step in this direction, thanks to the passion and hard work of our students. Stanford builds great research by building great researchers.”
In addition to paying homage to “Back to the Future,” the name MARTY provides insight to the types of research to be carried out by the team, which has been driven primarily by Jonathan Goh, a mechanical engineering graduate student in Gerdes’ Dynamic Design Lab (DDL).
Electronic stability control, or ESC, is a feature in most modern cars that ensures that the car stays within the boundaries of stable handling, by applying brakes to certain wheels, or even cutting engine power when needed.
“In our work developing autonomous driving algorithms, we’ve found that sometimes you need to sacrifice stability to turn sharply and avoid accidents,” Gerdes said. “The very best rally car drivers do this all this time, sacrificing stability so they can use all of the car’s capabilities to avoid obstacles and negotiate tight turns at speed. Their confidence in their ability to control the car opens up new possibilities for the car’s motion. Current control systems designed to assist a human driver, however, don’t allow this sort of maneuvering. We think that it is important to open up this design space to develop fully automated cars that are as safe as possible.
“One of the lab’s central paradigms is that autonomous cars must be able to handle all operating regimes, not just the simple one imposed by ESC. Learning how to program a car like MARTY to autonomously make the decision to trade the easiness of stability in order to drive with the fluidity and precision of a professional driver is at the heart of the lab’s research into figuring out how to use all of a car’s capabilities to create self-driving systems that will control the car more safely in all circumstances.
“When you watch a pro driver drift a car, you think to yourself that this person really knows how to precisely control the path and angle of the car, despite how different it is from normal driving,” Goh said. “The wheels are pointed to the left even though the car is turning right, and you have to very quickly co-ordinate throttle and steering in order to keep the car from spinning out or going the wrong way. Autonomous cars need to learn from this in order to truly be as good as the best drivers out there.”
Eventually, the car will be taught to race around a track utilizing this drifting technique to negotiate tight turns around obstacles when required. Already the car can autonomously lock itself into a continuous, precisely circular donut at a large drift angle, a significant feat of controls engineering accomplished by Goh. This is the first step on the path to a self-driving car that can deal with even the most extreme of situations. “The sublime awesomeness of riding in a DeLorean that does perfect, smoke-filled donuts by itself is a mind-bending experience that helps you appreciate that we really are living in the future,” Goh said.
MARTY was built in collaboration with Renovo Motors, an automotive start-up based in Silicon Valley that specializes in building advanced electric vehicle technology. Working closely together gave the Stanford team early access to a brand new platform derived from Renovo’s electric supercar that delivers 4,000 pound-feet from on-motor gearboxes to the rear wheels in a fraction of a second – allowing precise control of the forces required to drift.
Because the systems are all managed by a central API, the integration process was very rapid and Goh and his colleagues were able to get the car back on the road just a few months after pulling the original gasoline engine. Building off of the Renovo platform enabled the Stanford team to focus their development on subsystems and algorithms most important to the research goals.
“Stanford is a world leader in autonomous vehicle research, so partnering with them is an amazing opportunity. Having brilliant, hard-working students embedded here in our facility, conducting research and collaborating on MARTY is a great experience,” said Christopher Heiser, CEO and co-founder of Renovo Motors. “Using our platform, the team built a working version of MARTY in a matter of months and moved into research very quickly after that. I think we’re demonstrating how collaboration between Silicon Valley and universities can really work.”