Predictive Safety Network for Resource-constrained Multi-agent Systems

Meng Guo, Mathias Bürger
; Proceedings of the Conference on Robot Learning, PMLR 100:283-292, 2020.

Abstract

Coordinating multiple agents, such as mobile robots, with shared resources, such as common battery charging stations, is a highly relevant but still challenging decision problem. Traditionally, the motion and task planning of multi-agent systems are tackled by either designing ad-hoc decision rules or employing optimization tools. The former requires intensive manual tuning while the latter needs a static and accurate model of the complete system. Both approaches are prone to uncertainties in the robot motion and task execution. In this work, we propose a novel planning framework based on recent advances in deep reinforcement learning. The framework combines a centralized safety policy that acts on direct predictions of future resource levels and a decentralized task policy that optimizes task completions. The safety network is trained using supervised learning without extraneous supervision, while the task policy is trained using concurrent self-play. The whole framework follows a hierarchical structure to avoid the exponential blowup in the state and action space. We demonstrate significant improvements in a practical logistic planning problem for warehouse robots, compared with heuristic solutions, optimization tools and other reinforcement learning methods.

Cite this Paper


BibTeX
@InProceedings{pmlr-v100-guo20a, title = {Predictive Safety Network for Resource-constrained Multi-agent Systems}, author = {Guo, Meng and B\"urger, Mathias}, booktitle = {Proceedings of the Conference on Robot Learning}, pages = {283--292}, year = {2020}, editor = {Leslie Pack Kaelbling and Danica Kragic and Komei Sugiura}, volume = {100}, series = {Proceedings of Machine Learning Research}, address = {}, month = {30 Oct--01 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v100/guo20a/guo20a.pdf}, url = {http://proceedings.mlr.press/v100/guo20a.html}, abstract = {Coordinating multiple agents, such as mobile robots, with shared resources, such as common battery charging stations, is a highly relevant but still challenging decision problem. Traditionally, the motion and task planning of multi-agent systems are tackled by either designing ad-hoc decision rules or employing optimization tools. The former requires intensive manual tuning while the latter needs a static and accurate model of the complete system. Both approaches are prone to uncertainties in the robot motion and task execution. In this work, we propose a novel planning framework based on recent advances in deep reinforcement learning. The framework combines a centralized safety policy that acts on direct predictions of future resource levels and a decentralized task policy that optimizes task completions. The safety network is trained using supervised learning without extraneous supervision, while the task policy is trained using concurrent self-play. The whole framework follows a hierarchical structure to avoid the exponential blowup in the state and action space. We demonstrate significant improvements in a practical logistic planning problem for warehouse robots, compared with heuristic solutions, optimization tools and other reinforcement learning methods.} }
Endnote
%0 Conference Paper %T Predictive Safety Network for Resource-constrained Multi-agent Systems %A Meng Guo %A Mathias Bürger %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-guo20a %I PMLR %J Proceedings of Machine Learning Research %P 283--292 %U http://proceedings.mlr.press %V 100 %W PMLR %X Coordinating multiple agents, such as mobile robots, with shared resources, such as common battery charging stations, is a highly relevant but still challenging decision problem. Traditionally, the motion and task planning of multi-agent systems are tackled by either designing ad-hoc decision rules or employing optimization tools. The former requires intensive manual tuning while the latter needs a static and accurate model of the complete system. Both approaches are prone to uncertainties in the robot motion and task execution. In this work, we propose a novel planning framework based on recent advances in deep reinforcement learning. The framework combines a centralized safety policy that acts on direct predictions of future resource levels and a decentralized task policy that optimizes task completions. The safety network is trained using supervised learning without extraneous supervision, while the task policy is trained using concurrent self-play. The whole framework follows a hierarchical structure to avoid the exponential blowup in the state and action space. We demonstrate significant improvements in a practical logistic planning problem for warehouse robots, compared with heuristic solutions, optimization tools and other reinforcement learning methods.
APA
Guo, M. & Bürger, M.. (2020). Predictive Safety Network for Resource-constrained Multi-agent Systems. Proceedings of the Conference on Robot Learning, in PMLR 100:283-292

Related Material