Abstract
This paper describes a simple method which enables control of shape and internal movements as well as migration of a herd of autonomous mobile robots in a field with various types of obstacles. In order to facilitate such collective behaviors, this paper firstly introduces an individual law of motion for each robot employing simple virtual forces considering realization on actual robots. Then, the authors propose a method where a supervisor can control the collective behaviors of the herd in realtime only by manipulating parameters and a reference position signal which are broadcast to all the robots. The behaviors by the method is analyzed theoretically, and confirmed by simulation and by experiment using Child Machines of Super Mechano Colony Experimental System.
