Distributed Parallel Intelligent Agents for Critical Domain Robotic Control

G. Pitts, M. Eggen, B. Munsinger, and D. Nevin (USA)


intelligent control, agent, real-time, distributed, robotics


Autonomous intelligent robotic systems that can survive in dynamic environments have yet to be realized. The challenge stems from the need to process incoming information, make decisions, and learn from previous actions concurrently. Reliance upon one all encompassing system and one highly burdened processor, no matter how fast, makes this goal impossible. Recent advances in neurobiology have made it clear that living organisms use massively distributed processing systems to accomplish their goals. This paper presents a new methodology that takes advantage of the parallel nature of thought. By correctly dividing the tasks needed to control an autonomous intelligent robotics system among multiple clusters of distributed intelligent agents, real time action by a robotics system is attainable. In this paper, we present such a model.

Important Links:

Go Back