Dynamic Multi-Team Racing: Competitive Driving on 1/10-th Scale Vehicles via Learning in Simulation

Peter Werner, Tim Seyde, Paul Drews, Thomas Matrai Balch, Igor Gilitschenski, Wilko Schwarting, Guy Rosman, Sertac Karaman, Daniela Rus
Proceedings of The 7th Conference on Robot Learning, PMLR 229:1667-1685, 2023.

Abstract

Autonomous racing is a challenging task that requires vehicle handling at the dynamic limits of friction. While single-agent scenarios like Time Trials are solved competitively with classical model-based or model-free feedback control, multi-agent wheel-to-wheel racing poses several challenges including planning over unknown opponent intentions as well as negotiating interactions under dynamic constraints. We propose to address these challenges via a learning-based approach that effectively combines model-based techniques, massively parallel simulation, and self-play reinforcement learning to enable zero-shot sim-to-real transfer of highly dynamic policies. We deploy our algorithm in wheel-to-wheel multi-agent races on scale hardware to demonstrate the efficacy of our approach. Further details and videos can be found on the project website: https://sites.google.com/view/dynmutr/home.

Cite this Paper

BibTeX
@InProceedings{pmlr-v229-werner23a, title = {Dynamic Multi-Team Racing: Competitive Driving on 1/10-th Scale Vehicles via Learning in Simulation}, author = {Werner, Peter and Seyde, Tim and Drews, Paul and Balch, Thomas Matrai and Gilitschenski, Igor and Schwarting, Wilko and Rosman, Guy and Karaman, Sertac and Rus, Daniela}, booktitle = {Proceedings of The 7th Conference on Robot Learning}, pages = {1667--1685}, year = {2023}, editor = {Tan, Jie and Toussaint, Marc and Darvish, Kourosh}, volume = {229}, series = {Proceedings of Machine Learning Research}, month = {06--09 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v229/werner23a/werner23a.pdf}, url = {https://proceedings.mlr.press/v229/werner23a.html}, abstract = {Autonomous racing is a challenging task that requires vehicle handling at the dynamic limits of friction. While single-agent scenarios like Time Trials are solved competitively with classical model-based or model-free feedback control, multi-agent wheel-to-wheel racing poses several challenges including planning over unknown opponent intentions as well as negotiating interactions under dynamic constraints. We propose to address these challenges via a learning-based approach that effectively combines model-based techniques, massively parallel simulation, and self-play reinforcement learning to enable zero-shot sim-to-real transfer of highly dynamic policies. We deploy our algorithm in wheel-to-wheel multi-agent races on scale hardware to demonstrate the efficacy of our approach. Further details and videos can be found on the project website: https://sites.google.com/view/dynmutr/home.} }
Endnote
%0 Conference Paper %T Dynamic Multi-Team Racing: Competitive Driving on 1/10-th Scale Vehicles via Learning in Simulation %A Peter Werner %A Tim Seyde %A Paul Drews %A Thomas Matrai Balch %A Igor Gilitschenski %A Wilko Schwarting %A Guy Rosman %A Sertac Karaman %A Daniela Rus %B Proceedings of The 7th Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2023 %E Jie Tan %E Marc Toussaint %E Kourosh Darvish %F pmlr-v229-werner23a %I PMLR %P 1667--1685 %U https://proceedings.mlr.press/v229/werner23a.html %V 229 %X Autonomous racing is a challenging task that requires vehicle handling at the dynamic limits of friction. While single-agent scenarios like Time Trials are solved competitively with classical model-based or model-free feedback control, multi-agent wheel-to-wheel racing poses several challenges including planning over unknown opponent intentions as well as negotiating interactions under dynamic constraints. We propose to address these challenges via a learning-based approach that effectively combines model-based techniques, massively parallel simulation, and self-play reinforcement learning to enable zero-shot sim-to-real transfer of highly dynamic policies. We deploy our algorithm in wheel-to-wheel multi-agent races on scale hardware to demonstrate the efficacy of our approach. Further details and videos can be found on the project website: https://sites.google.com/view/dynmutr/home.
APA
Werner, P., Seyde, T., Drews, P., Balch, T.M., Gilitschenski, I., Schwarting, W., Rosman, G., Karaman, S. & Rus, D.. (2023). Dynamic Multi-Team Racing: Competitive Driving on 1/10-th Scale Vehicles via Learning in Simulation. Proceedings of The 7th Conference on Robot Learning, in Proceedings of Machine Learning Research 229:1667-1685 Available from https://proceedings.mlr.press/v229/werner23a.html.

Related Material