A holonic ontology-based multi-agent system for the distributed scheduling and monitoring of industrial processes

This paper proposes a middleware architecture that will serve as a framework for the management of industrial processes in an industrial grid environment. The proposed system is based on agents making possible the effective distribution of intelligence, thus fortifying the survivability of the industrial network, while a central point of monitoring and control allows agent monitoring. By the term 'agent' we denote a piece of software which overviews a well-defined part of an industrial process, abides to rules, accepts inputs, issues reports and easily adapts to environmental changes. Utilizing key technologies like web services, workflow allocation and ontologies, a scalable and highly adaptable architecture is achieved.