Learning Feasible Transitions for Efficient Contact Planning

Rikhat Akizhanov, Victor Dhedin, Majid Khadiv, Ivan Laptev
Proceedings of the 7th Annual Learning for Dynamics \& Control Conference, PMLR 283:431-442, 2025.

Abstract

In this paper, we propose an efficient contact planner for quadrupedal robots to navigate in extremely constrained environments such as stepping stones. The main difficulty in this setting stems from the mixed nature of the problem, namely discrete search over the steppable patches and continuous trajectory optimization. To speed up the discrete search, we study the properties of the transitions from one contact mode to another. In particular, we propose to learn a dynamic feasibility classifier and a target adjustment network. The former predicts if a contact transition between two contact modes is dynamically feasible. The latter is trained to compensate for misalignment in reaching a desired set of contact locations, due to imperfections of the low-level control. We integrate these learned networks in a Monte Carlo Tree Search (MCTS) contact planner. Our simulation results demonstrate that training these networks with offline data significantly speeds up the online search process and improves its accuracy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v283-akizhanov25a, title = {Learning Feasible Transitions for Efficient Contact Planning}, author = {Akizhanov, Rikhat and Dhedin, Victor and Khadiv, Majid and Laptev, Ivan}, booktitle = {Proceedings of the 7th Annual Learning for Dynamics \& Control Conference}, pages = {431--442}, year = {2025}, editor = {Ozay, Necmiye and Balzano, Laura and Panagou, Dimitra and Abate, Alessandro}, volume = {283}, series = {Proceedings of Machine Learning Research}, month = {04--06 Jun}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v283/main/assets/akizhanov25a/akizhanov25a.pdf}, url = {https://proceedings.mlr.press/v283/akizhanov25a.html}, abstract = {In this paper, we propose an efficient contact planner for quadrupedal robots to navigate in extremely constrained environments such as stepping stones. The main difficulty in this setting stems from the mixed nature of the problem, namely discrete search over the steppable patches and continuous trajectory optimization. To speed up the discrete search, we study the properties of the transitions from one contact mode to another. In particular, we propose to learn a dynamic feasibility classifier and a target adjustment network. The former predicts if a contact transition between two contact modes is dynamically feasible. The latter is trained to compensate for misalignment in reaching a desired set of contact locations, due to imperfections of the low-level control. We integrate these learned networks in a Monte Carlo Tree Search (MCTS) contact planner. Our simulation results demonstrate that training these networks with offline data significantly speeds up the online search process and improves its accuracy.} }
Endnote
%0 Conference Paper %T Learning Feasible Transitions for Efficient Contact Planning %A Rikhat Akizhanov %A Victor Dhedin %A Majid Khadiv %A Ivan Laptev %B Proceedings of the 7th Annual Learning for Dynamics \& Control Conference %C Proceedings of Machine Learning Research %D 2025 %E Necmiye Ozay %E Laura Balzano %E Dimitra Panagou %E Alessandro Abate %F pmlr-v283-akizhanov25a %I PMLR %P 431--442 %U https://proceedings.mlr.press/v283/akizhanov25a.html %V 283 %X In this paper, we propose an efficient contact planner for quadrupedal robots to navigate in extremely constrained environments such as stepping stones. The main difficulty in this setting stems from the mixed nature of the problem, namely discrete search over the steppable patches and continuous trajectory optimization. To speed up the discrete search, we study the properties of the transitions from one contact mode to another. In particular, we propose to learn a dynamic feasibility classifier and a target adjustment network. The former predicts if a contact transition between two contact modes is dynamically feasible. The latter is trained to compensate for misalignment in reaching a desired set of contact locations, due to imperfections of the low-level control. We integrate these learned networks in a Monte Carlo Tree Search (MCTS) contact planner. Our simulation results demonstrate that training these networks with offline data significantly speeds up the online search process and improves its accuracy.
APA
Akizhanov, R., Dhedin, V., Khadiv, M. & Laptev, I.. (2025). Learning Feasible Transitions for Efficient Contact Planning. Proceedings of the 7th Annual Learning for Dynamics \& Control Conference, in Proceedings of Machine Learning Research 283:431-442 Available from https://proceedings.mlr.press/v283/akizhanov25a.html.

Related Material