Multiagent Architecture for Telerobotics

Abstract

We present a robot architecture, based on multiagent technology, for robot teleoperation on the Internet. Intelligent agents are derived from agent theory and they are used to model simple rational behaviors in elaborated distributed applications. In particular, they are structured as multiagent systems based on the deliberative belief-desire-intention architecture. At present, multiagent system technology is employed in many applications because it is considered as a new software paradigm able to solve complex problems. On the other hand, robot teleoperation has arisen from the necessity to execute operations where it is difficult to put a human being due to constraints such as cost, safety or time. Considering robot teleoperation on the Internet is a complex problem, we have proposed and developed a multiagent architecture, where information streams, including video data and voice commands, transmitted on the Internet, are processed and managed using intelligent agents to execute successful tasks.