Learning Cooperative Control Policies for Multi-Robot Systems

In this talk, I discuss how we leverage machine learning methods to generate cooperative policies for multi-agent / multi-robot systems. I describe how we use Graph Neural Networks (GNNs) to learn effective communication strategies for decentralized coordination across various multi-agent tasks. In particular, I will discuss recent work on the role of agent heterogeneity and its impact on transferring learned cooperative policies to real-world experimental settings.