Case Study: UC Berkeley Research achieves real-time autonomous race car optimization with Julia Computing

UC Berkeley researchers use Julia to optimize model predictive control for the Berkeley Autonomous Race Car

UC Berkeley Research used Julia Computing’s Julia language and the JuMP optimization package to tackle the challenge of planning and controlling the Berkeley Autonomous Race Car (BARC), an autonomous RC car that must handle drifting, obstacle avoidance, and other complex maneuvers with high precision at high speed. The team needed a fast, clean way to build optimization problems and support real-time decision-making from sensor data.

With Julia Computing, UC Berkeley researchers implemented real-time model predictive control for autonomous race car path planning using open source solvers. The solution improved their ability to translate research code into practical applications, and Professor Francesco Borrelli noted that Julia’s new features and ARM support made deployment to real-world systems easier, enabling safer and more effective autonomous driving research.

UC Berkeley Research