MARLAS: Multi Agent Reinforcement Learning for cooperated Adaptive Sampling

TL;DR

This paper introduces MARLAS, a multi-agent reinforcement learning approach enabling cooperative, scalable, and robust adaptive sampling by robot teams to efficiently explore environmental phenomena within resource constraints.

Contribution

MARLAS is a novel decentralized RL method that encodes neighbor estimates and communication, improving multi-robot adaptive sampling performance and robustness.

Findings

Outperforms baseline sampling techniques in simulations and real experiments.

Demonstrates scalability with team size and region complexity.

Shows robustness to communication and robot failures.

Abstract

The multi-robot adaptive sampling problem aims at finding trajectories for a team of robots to efficiently sample the phenomenon of interest within a given endurance budget of the robots. In this paper, we propose a robust and scalable approach using Multi-Agent Reinforcement Learning for cooperated Adaptive Sampling (MARLAS) of quasi-static environmental processes. Given a prior on the field being sampled, the proposed method learns decentralized policies for a team of robots to sample high-utility regions within a fixed budget. The multi-robot adaptive sampling problem requires the robots to coordinate with each other to avoid overlapping sampling trajectories. Therefore, we encode the estimates of neighbor positions and intermittent communication between robots into the learning process. We evaluated MARLAS over multiple performance metrics and found it to outperform other baseline multi-robot sampling techniques. Additionally, we demonstrate scalability with both the size of the robot team and the size of the region being sampled. We further demonstrate robustness to communication failures and robot failures. The experimental evaluations are conducted both in simulations on real data and in real robot experiments on demo environmental setup.