| On 2023-02-14 at 15:00:00 (Brussels Time) |
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 seminar based on my paper for ICRA 2023, 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. Additionally, I'll introduce MEDIRL, a more advanced inverse reinforcement learning technique I'm currently using to address more classes of robot swarm missions.
