Autonomous distributed action selection algorithm for multiple mobile robot in transportation task

Recently, as robots are required to have more advanced ability to achieve various tasks, due to limitation of a single mobile robot, the research on multiple mobile robots system under decentralized control attention. Particularly, it is very important to assign automated delivery tasks to multiple mobile robots. Multiple mobile robots under decentralized control are not controlled, but recognize surround environment using sensors and decide action by themselves. However, since there are many other robots, collisions with other robots occur many times. In order to solve this problem, explicit communication is used, but it needs communication cost. In this paper, we propose a method of behavioral decision without knowing other robots' state for multiple mobile robots system which is applied to sorting task. Experimental simulations are conducted for different environment and the change of number of robots and the task progress are discussed.