Learning and Development for Physical Animats: Environmental Complexity Control for Vision-Based Mobile Robot

This paper discusses how an animat can develop its internal representation according to the complexity of the interactions with its environment through its capabilities of sensing and acting. The complexity should be carefully increased so that the animat can develop its internal representation so as to accomplish a task at hand. A method for controlling the complexity is proposed for a vision-based mobile robot of which task is to shoot a ball into a goal avoiding collisions with a goal keeper. The central issue is how to control the behavior of the goal keeper. The motion speed of the goal keeper is scheduled for e cient development of the learner that attempting at coping with new situations by nding a new axis in its state space. Simulations and the preliminary experiments by real robots are shown and, other issues for the complexity control are discussed.