ABSTRACT
Decentralized autonomous robots can effectively perform certain tasks with their collective behaviors. In this chapter, we will address the problem of how to
acquire a collective goal-directed, task-driven behavior in a group of distributed robots. The specific task that we will consider here is for the individual robots to perform coordinated movements that will push a box toward a desired goal location. The difficulty of this task is that the movements of the robots with respect to the box cannot follow an explicit, global plan and control command, due to certain modeling limitations as well as planning costs. In such a case, it is important that the individual robots locally modify their movement strategies. At the same time they must create a desired collective interaction between the robot group and the box, which will successfully bring the box to the goal location.
