Enabling Multi-Robot Collaboration from Single-Human Guidance
This addresses the challenge of enabling efficient collaboration in multi-agent systems, such as robots, by reducing the need for extensive human expertise, though it is incremental in leveraging existing human guidance concepts.
The paper tackles the problem of learning collaborative behaviors in multi-agent systems by proposing a method that uses only a single human's guidance, where agents learn by dynamically switching control and incorporating a theory-of-mind model. The result is a 58% improvement in success rate on a hide-and-seek task with just 40 minutes of human input, and the approach transfers to real-world multi-robot experiments.
Learning collaborative behaviors is essential for multi-agent systems. Traditionally, multi-agent reinforcement learning solves this implicitly through a joint reward and centralized observations, assuming collaborative behavior will emerge. Other studies propose to learn from demonstrations of a group of collaborative experts. Instead, we propose an efficient and explicit way of learning collaborative behaviors in multi-agent systems by leveraging expertise from only a single human. Our insight is that humans can naturally take on various roles in a team. We show that agents can effectively learn to collaborate by allowing a human operator to dynamically switch between controlling agents for a short period and incorporating a human-like theory-of-mind model of teammates. Our experiments showed that our method improves the success rate of a challenging collaborative hide-and-seek task by up to 58% with only 40 minutes of human guidance. We further demonstrate our findings transfer to the real world by conducting multi-robot experiments.