ResearchML
F1Tenth Autonomous Racing
Perception and planning for a 1/10-scale autonomous race car at CMU's AIMS Lab.
- Result
- 5th / 17
- Event
- IEEE IV 2026
- Stack
- ROS 2 / C++
Overview
Competed in the 28th Roboracer Autonomous Racing Competition at IEEE IV 2026 in Detroit, finishing 5th of 17 teams. The work was conducted at CMU's AIMS Lab under the advisement of Prof. John Dolan.
Approach
Developed an MPC controller combined with a Barrier State Control approach to enable safe, high-speed navigation in competitive race conditions. Work is extending toward an imitation-learning architecture to achieve faster onboard reaction times by distilling expert driving behavior directly into the planning stack.
Specifications
- Lab
- AIMS / DRIVE Lab, CMU
- Advisor
- Prof. John Dolan
- Control
- MPC + Barrier State Control
- Platform
- 1/10-scale autonomous vehicle