All Botz The University of Auckland RoboSoccer Team

This paper provides general information about research at the University of Auckland into autonomous agents in highly dynamic environments, in particular in the RoboCup environment. The first part of this paper describes the design of our vision server and our camera calibration method and shows some of our experiences in developing our RoboCup entry. This may help other interested parties in avoiding similar pitfalls in the future. The second part of the paper describes our general architecture, which consists of a distributed, cooperative agent architecture with powerful path planning and controller components. The path planner is an anytime planning algorithm that supports reactive as well as strategic reasoning of the agent. The controller uses a fuzzy logic approach. Lastly, the paper describes our work on the design of a case-based reinforcement learner.