The goal of this project is to add learning capabilities to the navigation system being developped in the framework of the ARGOS Project (see description above). In particular we are investigating the application of reinforcement learning techniques in a multi-agent setting, in collaboration with the group of Tom Dietterich at Oregons State University.