F1Tenth
At the intersection of all things robotics.. And also Formula 1 racing…
In Spring 2023, I was lucky enough to work with three fellow UPenn ROBO grad students: Irene Grace-Kolsen, Yu-Chia Shen, and Zhuolon (Alan) Zhao, and together we formed the team "Autonomous Anonymous". We spent the semester writing and testing software for a small (~1 foot long) autonomous car, and ultimately competed in three timed races.
From “Hello World” to SLAM to obstacle-proof path planning…
Our initial objective was mastering ROS2, running it within an Ubuntu/Linux Virtual Machine, and managing/using it through direct terminal bash commands. Then we implemented Automatic Emergency Braking (AEB) as a safety protocol moving forward. Next for path planning/following and controls, we implemented RRT*, pure pursuit, MPC, Follow the Gap, and Wall Following, and used different approaches across the three races. On the perception side, we localized the car using a particle filter, and mapped out both required racetracks using the slam_toolbox package; these map-based methods covered both Hokuyo LiDAR and RealSense camera sensory input, and strengthened our performance in the latter two races.
Final Project Presentation
Code Repository
Final Race Performance
In this clip, our car demonstrates its “overtake” algorithm, passing our opponent’s car.
Afterwards, it resumes its typical feedback control strategy, namely execution of pure pursuit path following.
All the while, the perception stack continues running, most notably LiDAR particle filter for localization.
Coming Soon
This page is currently being developed, and will be available shortly! Thank you for your patience. I hope the above resources give a preliminary sample of our exciting work.
