Online Learning for Obstacle Avoidance

David Snyder, Meghan Booker, Nathaniel Simon, Wenhan Xia, Daniel Suo, Elad Hazan, Anirudha Majumdar
Proceedings of The 7th Conference on Robot Learning, PMLR 229:2926-2954, 2023.

Abstract

We approach the fundamental problem of obstacle avoidance for robotic systems via the lens of online learning. In contrast to prior work that either assumes worst-case realizations of uncertainty in the environment or a stationary stochastic model of uncertainty, we propose a method that is efficient to implement and provably grants instance-optimality with respect to perturbations of trajectories generated from an open-loop planner (in the sense of minimizing worst-case regret). The resulting policy adapts online to realizations of uncertainty and provably compares well with the best obstacle avoidance policy in hindsight from a rich class of policies. The method is validated in simulation on a dynamical system environment and compared to baseline open-loop planning and robust Hamilton-Jacobi reachability techniques. Further, it is implemented on a hardware example where a quadruped robot traverses a dense obstacle field and encounters input disturbances due to time delays, model uncertainty, and dynamics nonlinearities.

Cite this Paper

BibTeX
@InProceedings{pmlr-v229-snyder23a, title = {Online Learning for Obstacle Avoidance}, author = {Snyder, David and Booker, Meghan and Simon, Nathaniel and Xia, Wenhan and Suo, Daniel and Hazan, Elad and Majumdar, Anirudha}, booktitle = {Proceedings of The 7th Conference on Robot Learning}, pages = {2926--2954}, 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/snyder23a/snyder23a.pdf}, url = {https://proceedings.mlr.press/v229/snyder23a.html}, abstract = {We approach the fundamental problem of obstacle avoidance for robotic systems via the lens of online learning. In contrast to prior work that either assumes worst-case realizations of uncertainty in the environment or a stationary stochastic model of uncertainty, we propose a method that is efficient to implement and provably grants instance-optimality with respect to perturbations of trajectories generated from an open-loop planner (in the sense of minimizing worst-case regret). The resulting policy adapts online to realizations of uncertainty and provably compares well with the best obstacle avoidance policy in hindsight from a rich class of policies. The method is validated in simulation on a dynamical system environment and compared to baseline open-loop planning and robust Hamilton-Jacobi reachability techniques. Further, it is implemented on a hardware example where a quadruped robot traverses a dense obstacle field and encounters input disturbances due to time delays, model uncertainty, and dynamics nonlinearities.} }
Endnote
%0 Conference Paper %T Online Learning for Obstacle Avoidance %A David Snyder %A Meghan Booker %A Nathaniel Simon %A Wenhan Xia %A Daniel Suo %A Elad Hazan %A Anirudha Majumdar %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-snyder23a %I PMLR %P 2926--2954 %U https://proceedings.mlr.press/v229/snyder23a.html %V 229 %X We approach the fundamental problem of obstacle avoidance for robotic systems via the lens of online learning. In contrast to prior work that either assumes worst-case realizations of uncertainty in the environment or a stationary stochastic model of uncertainty, we propose a method that is efficient to implement and provably grants instance-optimality with respect to perturbations of trajectories generated from an open-loop planner (in the sense of minimizing worst-case regret). The resulting policy adapts online to realizations of uncertainty and provably compares well with the best obstacle avoidance policy in hindsight from a rich class of policies. The method is validated in simulation on a dynamical system environment and compared to baseline open-loop planning and robust Hamilton-Jacobi reachability techniques. Further, it is implemented on a hardware example where a quadruped robot traverses a dense obstacle field and encounters input disturbances due to time delays, model uncertainty, and dynamics nonlinearities.
APA
Snyder, D., Booker, M., Simon, N., Xia, W., Suo, D., Hazan, E. & Majumdar, A.. (2023). Online Learning for Obstacle Avoidance. Proceedings of The 7th Conference on Robot Learning, in Proceedings of Machine Learning Research 229:2926-2954 Available from https://proceedings.mlr.press/v229/snyder23a.html.

Related Material