MAMPS: Safe Multi-Agent Reinforcement Learning via Model Predictive Shielding
Systems and Control
2020-01-01 v2 Artificial Intelligence
Multiagent Systems
Robotics
Systems and Control
Abstract
Reinforcement learning is a promising approach to learning control policies for performing complex multi-agent robotics tasks. However, a policy learned in simulation often fails to guarantee even simple safety properties such as obstacle avoidance. To ensure safety, we propose multi-agent model predictive shielding (MAMPS), an algorithm that provably guarantees safety for an arbitrary learned policy. In particular, it operates by using the learned policy as often as possible, but instead uses a backup policy in cases where it cannot guarantee the safety of the learned policy. Using a multi-agent simulation environment, we show how MAMPS can achieve good performance while ensuring safety.
Cite
@article{arxiv.1910.12639,
title = {MAMPS: Safe Multi-Agent Reinforcement Learning via Model Predictive Shielding},
author = {Wenbo Zhang and Osbert Bastani and Vijay Kumar},
journal= {arXiv preprint arXiv:1910.12639},
year = {2020}
}