For the first time, Caltech’s autonomous race car competed on a U.S. road course—and made a bold impression. In the July 24 leg of the Indy Autonomous Challenge (IAC) at Laguna Seca’s WeatherTech Raceway, known for its demanding layout and the notorious “Corkscrew” turn, the Caltech Racer clocked some of the fastest opening laps before ultimately spinning out and crashing.
The IAC provides identical drive-by-wire AV-24 cars to university teams, challenging them to differentiate based on software performance in perception, planning, and control. Caltech only joined the competition in late 2024, yet has leveraged deep expertise from its Center for Autonomous Systems and Technologies (CAST) and partnerships with NASA’s JPL and Beyond Limits to build a sophisticated autonomy stack. During the race, the team’s aggressive race line strategy and adaptive motion planning delivered first and second laps that topped the field—before an error in “The Corkscrew” forced a spin and triggered a red flag under competition rules, dropping Caltech’s best times.
Though the finish was bittersweet, the result is still a milestone. Caltech demonstrated that its algorithms can perform under real-world stress, blending audacious pacing with fail-safes to respond when vehicle dynamics push limits. Team leads say the result strengthens their resolve to improve machine learning and adaptability for future races. As one commentator noted: “they had a car that was way, way, way competitive.”
