Cooperative Control of Multi-robots Using Mobile Agents in a Three-Dimensional Environment

This paper presents a framework for controlling multiple robots connected by communication networks. In the framework, each robot does not achieve just a single task based on initially set program, but performs several tasks depending on programs introduced by mobile software agents from outside. Since the mobile software agents can migrate to arbitrary robots by wireless communication networks, they can find the most suitably equipped and/or the most suitably located robots to perform their task. Using the mobility of the agents, we have proposed several approaches suppressing the total energy consumption of the multiple robots without sacrificing their efficiency. In these approaches, the target robots in the migration of software agents are restricted to those within the view of the camera of the source robot. Because it is necessary to identify the location and the IP address of the robots to migrate, and therefore, all the approaches are restricted by a condition that the source robot has to have a clear view without any obstacles so that the robot can see the target robots. In this paper, we propose a simple manner for controlling multiple robots in a three-dimensional environment by including aerial robots. The aerial robots provide bird´s-eyes to the robots on the ground. Furthermore, since the aerial robot hovers over the ground robot on which a software agent resides, the movement of the aerial robot is achieved by traversing ground robots following the agent. The controlling manner contributes to making it easy to control multiple robots with the aerial robots through reducing three-dimensional control environment to two-dimensional one. We have implemented the basic behaviors in the three-dimensional model. We show prospects and problems of our model through their experimental results.