An Autonomic Architecture for Multi-Agent Self-Maintaining Robotic Systems

Abstract

Autonomic Computing (AC) systems have four maj or characteristics: self-healing, self-configuring, self-optimizing and self-protecting, combined refer to self-CHOP. This paper examines the self-healing properties of an autonomic system and how they can be applied to a multi-agent autonomous robotic system. This work defines an architecture that incorporates the maintenance of robots into the autonomic management of multi-agent systems. Related concepts from self-healing, prognostics and health management (PHM), and non-destructive evaluation (NDE) domains are discussed and integrated into the design of the architecture. A simple simulation applying the architecture is implemented using python and preliminary results are presented.