Show me what you want: Inverse reinforcement learning to automatically design robot swarms by demonstration

TL;DR

This paper introduces Demo-Cho, a method that uses inverse reinforcement learning and demonstrations to automatically design control software for robot swarms, eliminating the need for explicit objective functions.

Contribution

Demo-Cho is the first approach to automatically generate swarm control software solely from demonstrations without predefined objectives.

Findings

Successfully designed control software for four different missions

Works in both simulation and physical robot experiments

Eliminates the need for mission-specific objective functions

Abstract

Automatic design is a promising approach to generating control software for robot swarms. So far, automatic design has relied on mission-specific objective functions to specify the desired collective behavior. In this paper, we explore the possibility to specify the desired collective behavior via demonstrations. We develop Demo-Cho, an automatic design method that combines inverse reinforcement learning with automatic modular design of control software for robot swarms. We show that, only on the basis of demonstrations and without the need to be provided with an explicit objective function, Demo-Cho successfully generated control software to perform four missions. We present results obtained in simulation and with physical robots.