Trajectory Planning and Control near the Limits: an Open Experimental Benchmark on the RoboRacer Platform

Researchers introduced an open benchmark for near-limit trajectory planning and control on the 1:10 RoboRacer platform.

The framework combines time-optimal raceline generation, online velocity replanning, tracking controllers, and a neural inverse-dynamics model.

Tests on two tracks showed improved tracking, fewer steering oscillations, and faster lap times.

Code, data, videos, and results were released to give the robotics community a shared testbed for comparison.