Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks

Ekaterina Tolstaya, Fernando Gama, James Paulos, George Pappas, Vijay Kumar, Alejandro Ribeiro
Proceedings of the Conference on Robot Learning, PMLR 100:671-682, 2020.

Abstract

We consider the problem of finding distributed controllers for large networks of mobile robots with interacting dynamics and sparsely available communications. Our approach is to learn local controllers that require only local information and communications at test time by imitating the policy of centralized controllers using global information at training time. By extending aggregation graph neural networks to time varying signals and time varying network support, we learn a single common local controller which exploits information from distant teammates using only local communication interchanges. We apply this approach to the problem of flocking to demonstrate performance on communication graphs that change as the robots move. We examine how a decreasing communication radius and faster velocities increase the value of multi-hop information.

Cite this Paper


BibTeX
@InProceedings{pmlr-v100-tolstaya20a, title = {Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks}, author = {Tolstaya, Ekaterina and Gama, Fernando and Paulos, James and Pappas, George and Kumar, Vijay and Ribeiro, Alejandro}, booktitle = {Proceedings of the Conference on Robot Learning}, pages = {671--682}, year = {2020}, editor = {Kaelbling, Leslie Pack and Kragic, Danica and Sugiura, Komei}, volume = {100}, series = {Proceedings of Machine Learning Research}, month = {30 Oct--01 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v100/tolstaya20a/tolstaya20a.pdf}, url = {https://proceedings.mlr.press/v100/tolstaya20a.html}, abstract = {We consider the problem of finding distributed controllers for large networks of mobile robots with interacting dynamics and sparsely available communications. Our approach is to learn local controllers that require only local information and communications at test time by imitating the policy of centralized controllers using global information at training time. By extending aggregation graph neural networks to time varying signals and time varying network support, we learn a single common local controller which exploits information from distant teammates using only local communication interchanges. We apply this approach to the problem of flocking to demonstrate performance on communication graphs that change as the robots move. We examine how a decreasing communication radius and faster velocities increase the value of multi-hop information.} }
Endnote
%0 Conference Paper %T Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks %A Ekaterina Tolstaya %A Fernando Gama %A James Paulos %A George Pappas %A Vijay Kumar %A Alejandro Ribeiro %B Proceedings of the Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2020 %E Leslie Pack Kaelbling %E Danica Kragic %E Komei Sugiura %F pmlr-v100-tolstaya20a %I PMLR %P 671--682 %U https://proceedings.mlr.press/v100/tolstaya20a.html %V 100 %X We consider the problem of finding distributed controllers for large networks of mobile robots with interacting dynamics and sparsely available communications. Our approach is to learn local controllers that require only local information and communications at test time by imitating the policy of centralized controllers using global information at training time. By extending aggregation graph neural networks to time varying signals and time varying network support, we learn a single common local controller which exploits information from distant teammates using only local communication interchanges. We apply this approach to the problem of flocking to demonstrate performance on communication graphs that change as the robots move. We examine how a decreasing communication radius and faster velocities increase the value of multi-hop information.
APA
Tolstaya, E., Gama, F., Paulos, J., Pappas, G., Kumar, V. & Ribeiro, A.. (2020). Learning Decentralized Controllers for Robot Swarms with Graph Neural Networks. Proceedings of the Conference on Robot Learning, in Proceedings of Machine Learning Research 100:671-682 Available from https://proceedings.mlr.press/v100/tolstaya20a.html.

Related Material