On May 19-20, 2022, Berkeley ROAR AI Racing team joined forces with partners from UCSD and University of Hawaii, and participated in the 2022 evGrandPrix autonomous racing competition in Indianapolis. The joint team, codenamed “Pacific AI Racing”, won the 2nd place prize by racing full-scale autonomous driving Go Karts against several other University teams that included Georgia Tech, Kennesaw State University, The University of Wisconsin, and Upenn.
evGrandPrix autonomous racing is a collegiate and high school robotics competition that challenges the overall drive-by-wire Go Kart chassis design and advanced vehicle perception systems. The traditional human driver is replaced with an on-board computer taking input from a variety of sensors placed throughout the kart. Combining the power of a suite of RGB and depth cameras, Berkeley ROAR team provided efficient yet reliable methods for obstacle detection and track tracking. The core development team from Berkeley consisted of EECS MS and MEng graduate students, and three Berkeley students Adam Curtis, CK Wolfe, Michael Wu participated in the competition on site. “It was a great experience to meet, collaborate, and connect with so many talented engineers from across the country. It was truly a unique and rewarding opportunity to learn from each other and exchange ideas. Next year, the team will be shooting for first place.” said Wolfe.
Berkeley ROAR project was founded by Dr. Allen Y. Yang and Prof. Shankar Sastry in 2020, focusing on democratizing AI research and education by offering cutting-edge low-cost reference hardware and software AI racing systems from software simulation, 1/10 scale RC car racing hardware, to full-scale Go Kart racing solutions. The project is open source for open collaboration with academia and industry at: https://roar.berkeley.edu/. The program also offers its own unique ROAR competition series for students from high school to graduate research.