Multi-agent scheduling in a human-robot collaborative warehouse

Bram Renting

PhD at Leiden University, Delft University of Technology

In cooperative multi-agent environments, agents can be interdependent in completing tasks. We consider environments where agents schedule future interactions with others they depend on to perform the task. More specifically, our project focuses on human-robot warehouses where humans pick products from shelves and robots transport goods. Proper scheduling of meetings between robots and humans is essential to use agents’ time effectively by keeping both of them from waiting for each other, which is challenging. We attempt to solve this problem using multi-agent negotiation, where agents that represent humans and robots negotiate with others to set future meetings. Lessons learned in this project can also be useful in closely related domains, such as agent-assisted meeting scheduling and scheduling of energy usage in communities of energy prosumers.

Keywords: Multi-agent scheduling, reinforcement learning, negotiation, cooperative AI

Scientific area: Multi-agent systems

Visiting period: April-May 2024 at University of Edinburgh