ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations

Samuel Pfrommer, Mathew Halm, Michael Posa
Proceedings of the 2020 Conference on Robot Learning, PMLR 155:2279-2291, 2021.

Abstract

Common methods for learning robot dynamics assume motion is continuous, causing unrealistic model predictions for systems undergoing discontinuous impact and stiction behavior. In this work, we resolve this conflict with a smooth, implicit encoding of the structure inherent to contact-induced discontinuities. Our method, ContactNets, learns parameterizations of inter-body signed distance and contact-frame Jacobians, a representation that is compatible with many simulation, control, and planning environments for robotics. We furthermore circumvent the need to differentiate through stiff or non-smooth dynamics with a novel loss function inspired by the principles of complementarity and maximum dissipation. Our method can predict realistic impact, non-penetration, and stiction when trained on 60 seconds of real-world data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v155-pfrommer21a, title = {ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations}, author = {Pfrommer, Samuel and Halm, Mathew and Posa, Michael}, booktitle = {Proceedings of the 2020 Conference on Robot Learning}, pages = {2279--2291}, year = {2021}, editor = {Kober, Jens and Ramos, Fabio and Tomlin, Claire}, volume = {155}, series = {Proceedings of Machine Learning Research}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v155/pfrommer21a/pfrommer21a.pdf}, url = {https://proceedings.mlr.press/v155/pfrommer21a.html}, abstract = {Common methods for learning robot dynamics assume motion is continuous, causing unrealistic model predictions for systems undergoing discontinuous impact and stiction behavior. In this work, we resolve this conflict with a smooth, implicit encoding of the structure inherent to contact-induced discontinuities. Our method, ContactNets, learns parameterizations of inter-body signed distance and contact-frame Jacobians, a representation that is compatible with many simulation, control, and planning environments for robotics. We furthermore circumvent the need to differentiate through stiff or non-smooth dynamics with a novel loss function inspired by the principles of complementarity and maximum dissipation. Our method can predict realistic impact, non-penetration, and stiction when trained on 60 seconds of real-world data.} }
Endnote
%0 Conference Paper %T ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations %A Samuel Pfrommer %A Mathew Halm %A Michael Posa %B Proceedings of the 2020 Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2021 %E Jens Kober %E Fabio Ramos %E Claire Tomlin %F pmlr-v155-pfrommer21a %I PMLR %P 2279--2291 %U https://proceedings.mlr.press/v155/pfrommer21a.html %V 155 %X Common methods for learning robot dynamics assume motion is continuous, causing unrealistic model predictions for systems undergoing discontinuous impact and stiction behavior. In this work, we resolve this conflict with a smooth, implicit encoding of the structure inherent to contact-induced discontinuities. Our method, ContactNets, learns parameterizations of inter-body signed distance and contact-frame Jacobians, a representation that is compatible with many simulation, control, and planning environments for robotics. We furthermore circumvent the need to differentiate through stiff or non-smooth dynamics with a novel loss function inspired by the principles of complementarity and maximum dissipation. Our method can predict realistic impact, non-penetration, and stiction when trained on 60 seconds of real-world data.
APA
Pfrommer, S., Halm, M. & Posa, M.. (2021). ContactNets: Learning Discontinuous Contact Dynamics with Smooth, Implicit Representations. Proceedings of the 2020 Conference on Robot Learning, in Proceedings of Machine Learning Research 155:2279-2291 Available from https://proceedings.mlr.press/v155/pfrommer21a.html.

Related Material