This project designs, builds, and validates a 1/10th-scale autonomous RoboRacer vehicle, capable of safely navigating an unknown indoor race track at competitive speeds. The vehicle must complete multiple laps autonomously without collisions while meeting performance requirements. The hardware platform is built on a Traxxas chassis equipped with an NVIDIA Jetson Orin Nano for onboard computation, a Hokuyo 2D LiDAR for environment sensing, a VESC 6 motor controller for throttle control, and a servo-based steering system, powered by a 3S LiPo battery through a custom power distribution board. A custom laser-cut deck was designed and fabricated to support a balanced, reliable hardware integration.

On the software side, the vehicle runs Ubuntu and ROS2 Humble with a modular node architecture spanning sensing, localization, planning, control, and safety monitoring. The autonomy stack combines centerline-based path following for trajectory planning, Pure Pursuit trajectory tracking with curvature-adaptive speed scaling, and Follow-The-Gap reactive obstacle avoidance for safety. A DriveMux arbitration node enables Follow-The-Gap to override nominal planners when unsafe conditions are detected, ensuring safe autonomous operation.