Automatic Discovery of Semantic Manufacturing Grid Services

Services in manufacturing grid (MG) are dynamically aggregated from different organizations. So it is essential for MG that services can be discovered efficiently. Automatic semantic and IR based approach of service discovery within a semantic manufacturing grid (SMG) architecture is studied in this paper. In SMG, services and job requirements are semantically described by a merged ontology of grid service ontology and manufacturing ontology. Service discovery is formalized as matchmaking between a job and advertised services. All kinds of degree of match (DoM) are formally defined considering both semantic match and document similarity. Then, a hybrid semantic and IR based matchmaking algorithm is presented. A SMG prototype system used for collaborative MEMS manufacturing has been built, in which a semantic service registry named UDDIe4SMG is built by extending UDDI and specially designed for efficient service discovery. Experiments show that the automatic service discovery approach in SMG works well and has good performance.