Planning-based semantic web service composition in factory automation

• WSDL annotation conventions facilitate semantic web service description derivation.
• A software agent composes semantic web services to achieve production goals.
• Graphical user interfaces visualize domain model and workflow execution statuses.
• Preprocessing algorithms neutralize syntactic restrictions imposed by AI planners.

The Service Oriented Architecture (SOA) paradigm enables production systems to be composed of web services. In an SOA-based production system, the individual production devices provide web service interfaces that encapsulate the behavior of the devices and abstract the implementation details. Such a service-oriented approach makes it possible to apply web service orchestration technologies in the development of production workflow descriptions. While manual formulation of production workflows tends to require considerable effort from domain experts, semantic web service descriptions enable computer algorithms to automatically generate the appropriate web service orchestrations. Such algorithms realize AI planning and employ semantic web service descriptions in determining the workflows required to achieve the production goals desired. In addition, the algorithms can automatically adapt the workflows to unexpected changes in the goals pursued and the production devices available.