Distilling contact planning for fast trajectory optimization in Robot Air Hockey
(2025)- Abstract
- Robot control through contact is challenging as it requires reasoning over long horizons and discontinuous system dynamics. Highly dynamic tasks such as Air Hockey additionally require agile behavior, making the corresponding optimal control problems intractable for planning in realtime. Learning-based approaches address this issue by shifting computationally expensive reasoning through contacts to an offline learning phase. However, learning low-level motor policies subject to kinematic and dynamic constraints can be challenging if operating in proximity to such constraints is desired. This paper explores the combination of distilling a stochastic optimal control policy for high-level contact planning and online model-predictive control... (More)
- Robot control through contact is challenging as it requires reasoning over long horizons and discontinuous system dynamics. Highly dynamic tasks such as Air Hockey additionally require agile behavior, making the corresponding optimal control problems intractable for planning in realtime. Learning-based approaches address this issue by shifting computationally expensive reasoning through contacts to an offline learning phase. However, learning low-level motor policies subject to kinematic and dynamic constraints can be challenging if operating in proximity to such constraints is desired. This paper explores the combination of distilling a stochastic optimal control policy for high-level contact planning and online model-predictive control for low-level constrained motion planning. Our system learns to balance shooting accuracy and resulting puck speed by leveraging bank shots and the robot's kinematic structure. We show that the proposed framework outperforms purely control-based and purely learning-based techniques in both simulated and real-world games of Robot Air Hockey. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/c5f36547-9b67-49b9-8eb5-c0c5d9f7f2d8
- author
- Jankowski, Julius
; Marić, Ante
; Liu, Puze
; Tateo, Davide
LU
; Peters, Jan
and Calinon, Sylvain
- publishing date
- 2025
- type
- Chapter in Book/Report/Conference proceeding
- publication status
- published
- subject
- host publication
- Robotics : science and systems : online proceedings - science and systems : online proceedings
- pages
- 9 pages
- language
- English
- LU publication?
- no
- id
- c5f36547-9b67-49b9-8eb5-c0c5d9f7f2d8
- alternative location
- https://www.roboticsproceedings.org/rss21/p115.html
- date added to LUP
- 2025-10-16 14:51:31
- date last changed
- 2025-11-03 16:17:54
@inproceedings{c5f36547-9b67-49b9-8eb5-c0c5d9f7f2d8,
abstract = {{Robot control through contact is challenging as it requires reasoning over long horizons and discontinuous system dynamics. Highly dynamic tasks such as Air Hockey additionally require agile behavior, making the corresponding optimal control problems intractable for planning in realtime. Learning-based approaches address this issue by shifting computationally expensive reasoning through contacts to an offline learning phase. However, learning low-level motor policies subject to kinematic and dynamic constraints can be challenging if operating in proximity to such constraints is desired. This paper explores the combination of distilling a stochastic optimal control policy for high-level contact planning and online model-predictive control for low-level constrained motion planning. Our system learns to balance shooting accuracy and resulting puck speed by leveraging bank shots and the robot's kinematic structure. We show that the proposed framework outperforms purely control-based and purely learning-based techniques in both simulated and real-world games of Robot Air Hockey.}},
author = {{Jankowski, Julius and Marić, Ante and Liu, Puze and Tateo, Davide and Peters, Jan and Calinon, Sylvain}},
booktitle = {{Robotics : science and systems : online proceedings}},
language = {{eng}},
title = {{Distilling contact planning for fast trajectory optimization in Robot Air Hockey}},
url = {{https://www.roboticsproceedings.org/rss21/p115.html}},
year = {{2025}},
}