This paper presents the Knowledge Base approach of the Skill Orchestration Agent (SkillOA), a modular rule-based architecture for dynamic skill-based control logic in Industry 4.0 environments. Building upon the Capability-Skill-Service (CSS) model, SkillOA introduces a three-layered knowledge structure and a domain-specific reasoner based on OWL-DL axioms and SPARQL rules. The system enables automated inference of executable skill sequences from declarative service requests while satisfying technical, functional, and temporal constraints. SkillOA thus provides a scalable, runtime-capable knowledge framework for adaptive, agent-based production systems.

Innovative engineering methods are needed to enhance the adaptability and agility of industrial control procedures and concurrently manage their rising complexity. Service-oriented Architecture (SOA) constitutes a promising paradigm to meet these challenges. To apply the rather abstract SOA principles to industrial automation, a model-driven engineering method is presented in this paper. Therefore, a reference architecture is introduced providing a framework to structure and define services for control tasks in a convenient way. The application of the engineering method is supported by an implementation concept. It uses SysML during the design phase and the process modeling and control language Grafchart for the execution of service-oriented control procedures.

The growing competition between manufacturers demands a higher versatility of factory automation technology. Component based automation principles propose an opportunity to address these challenges by encapsulating the functionality of mechatronic components in an abstract way. Service-oriented approaches depict a promising possibility to realize such architectures. Service discovery and orchestration are the key functionalities of such systems. However, standard web service technologies alone are not suited for the creation of highly-flexible automation systems. We describe a semantic service discovery and orchestration system, which is based on different semantic service technologies, and provide a concept towards the creation of adaptive production processes based on our experiences during practical implementation.

Industrial production and supply chains face an increased demand for mass-customization and tightening regulations for the traceability of goods, leading to higher requirements concerning flexibility, adaptability and transparency of processes. Technologies for the ¿Internet of Things¿ such as smart products and semantic representations pave their way into future factories and supply chains to fulfill these challenging market demands. In this paper a backend-independent approach for information exchange in open-loop processes based on so-called digital object memories (DOM) is presented. By storing order-related data via smart labels on the item, relevant life cycle information is attached to the product itself. This way, information handover via several stages of the value chain with potentially different stakeholders including manufacturer, distributor, retailer, and end customer has been realized. To summarize first best-practice experiences regarding memory structure and content, a prototype implementation based on a scenario of processing dietary supplements in an adaptive process is illustrated.