A Multi-Agent based Control Architecture for Autonomous Robotic Systems (MACAARS)

To achieve a transport application requiring multiple Autonomous Robotic Systems (ARS), a distributed control architecture endowed with deliberative and reactive capabilities is necessary. In this paper, a Multi-Agent based Control Architecture for Autonomous Robotic Systems (MACAARS) is suggested allowing a group of cooperative ARS (agents) to carry out this application in dynamic environments. The agents are grouped in a given formation under a flat organization where each agent is consisted of a deliberative level, an interpreter level and a reactive level. They exchange information via a communication module. This architecture is characterized by its flexibility and its distributed nature. Copyright (C) 1999 IFAC.