Towards Agent-Based Smart Factories: A Subject-Oriented Modeling Approach

This paper proposes subject-oriented modeling as an enabling technology for agent systems to be used in Smart Factories. This modeling approach is consistent with the agent-oriented view of processes as emerging from the interactions and local behaviors of autonomous entities (or "subjects"). While the subject-oriented methodology has been successfully applied in practice, to date its scope has been limited to enterprise-level business processes involving human agents. This paper lays the conceptual foundations for extending this scope to include both human and artificial agents, across all levels of industrial control and spanning entire value chains. Using a generic representation of agents, derived from the Function-Behavior-Structure (FBS) ontology, the paper develops a framework for the flexible assignment of agents to processes at runtime. This allows for the dynamic reconfiguration of process instances to adapt to changes in the production environment.