Experiments in Distributed Control of Autonomous Mobile Robots

Abstract

Distributed control is known to produce simple and cost effective solutions for many control applications. This is particularily interesting for controlling autonomous mobile robots, where the designer faces severe constraints in what concerns energy consumption, and the overall cost of the mobile unit. This paper presents a couple of experiments aimed to illustrate two different approaches in the implementation of distributed control of autonomaous mobile robots. In the first experiment, the robot is integrated in a flexible assembly line, and the control tasks are shared between the on board electronics and a ground based PLC. In the second experiment, the robot communicates with a number of “neural beacons” deployed in the environment, and the resulting system acts as a distributed neural network capable to learn the control actions needed to guide the robot along predefined paths.