Decentralized Role Assignment in Multi-Agent Teams via Empirical Game-Theoretic Analysis

TL;DR

This paper introduces a novel empirical game-theoretic approach enabling robots in a team to autonomously assign roles dynamically without communication, improving collaboration and collision avoidance in manipulation tasks.

Contribution

It formulates the role assignment as a dynamic game and develops a distributed controller based on empirical game theory for multi-agent teams.

Findings

Effective role assignment without communication demonstrated in simulations.

Robots successfully manipulate objects collaboratively while avoiding collisions.

The method adapts dynamically to changing team configurations.

Abstract

We propose a method, based on empirical game theory, for a robot operating as part of a team to choose its role within the team without explicitly communicating with team members, by leveraging its knowledge about the team structure. To do this, we formulate the role assignment problem as a dynamic game, and borrow tools from empirical game-theoretic analysis to analyze such games. Based on this game-theoretic formulation, we propose a distributed controller for each robot to dynamically decide on the best role to take. We demonstrate our method in simulations of a collaborative planar manipulation scenario in which each agent chooses from a set of feedback control policies at each instant. The agents can effectively collaborate without communication to manipulate the object while also avoiding collisions using our method.