Ontology-based microservices architecture for industrial assistant systems

Today's production environments provide only little standardization regarding human machine interaction. Vendors of machines and production systems usually are developing proprietary solutions for data storage, programming interfaces and especially for user interfaces. Machine customers are confronted with different visual styles and different perceptual experiences. The variety of interfaces and the closed nature of such systems cause high effort for development and maintenance of industrial assistance solutions. This paper proposes a system architecture with open interfaces as a base for modular and independent assistance software. Where applicable, data is represented using linked-data concepts while semantics of production data and descriptive information are represented using lightweight ontologies. This allows for combining data from disparate sources and gathering new information. The proposed concept is illustrated with examples from the manufacturing domain like maintenance, data analysis and harmonization of heterogeneous human-machine interfaces (HMI).