Self-configurable Manufacturing Industrial Agents (SMIA): a standardized approach for digitizing manufacturing assets

Abstract

• SMIA self-configures automated Digital Twins from semantically enriched AAS models. • The approach integrates AAS, CSS and MAS into an operational software architecture. • The extended CSS model distinguishes physical and agent capabilities in assets. • With open source engineering, modularity is combined with user-level extensibility. • It is validated in a robotic logistics scenario with autonomous agent negotiation. The integration of heterogeneous industrial assets into flexible and distributed manufacturing environments remains a fundamental challenge in the evolution of Industry 4.0. Although standardization efforts – such as the Asset Administration Shell (AAS) and the Capability-Skill-Service (CSS) model promoted by the Plattform Industrie initiative, the OWL ontology language promoted by W3C, and the FIPA Agent Communication Language promoted by IEEE – offer structured frameworks for interoperability, their combined application still presents unresolved implementation challenges. This paper introduces SMIA (Self-configurable Manufacturing Industrial Agents), a dual-layer solution that operationalizes these standards by converging semantic modeling and distributed software execution. First, SMIA proposes a methodology for characterizing proactive AAS by enriching their descriptions with an OWL ontology based on the CSS model. Second, it provides a software toolchain that automatically generates Digital Twins –implemented as FIPA-compliant industrial agents using SPADE – from these enriched AAS descriptions. The resulting agents can react to external events, coordinating with peers and executing domain-specific skills within a standardized I4.0 communication framework. Built following an Open Source Software Engineering approach, SMIA leverages mature tools such as Eclipse BaSyx, OWLReady2 and SPADE, showcasing a replicable and extensible approach to the adoption of industrial standards in practice. A preliminary validation in a robotic logistics scenario demonstrates its feasibility and adaptability.