On enhancing communication of the manufacturing service capability information using reference ontology

Abstract

A manufacturing service capability (MSC) model is essential for correct communication of MSC information between dynamic supply chain (DSC) partners. MSC information elicits service details such as locations, specialisations, capacities, certifications, and software and material processing capabilities. Presently, as in the case of web portal-enabled supply chain communications, this information is communicated using models that provide limited semantic precision. These models are also proprietary to specific industry communities; hence, access to MSC information across communities is limited. This article describes deployment of the reference-ontology-based semantic mediation approach using Web Ontology Language (OWL). The mediation introduces shared semantics that enhances access to and precision of proprietary MSC models encoded in relational databases. Detailed deployment steps are discussed using a manufacturing sourcing use case. In the first step, we analyse three alternative conventions for encoding proprietary MSC models into OWL and show that the ontology-oriented encoding convention is most suitable for semantic mediation in OWL. In the deployment step, single- or multi-community deployment is possible. It is shown that the single-community deployment enables precise and greater access to MSC information within the community, while the multi-community deployment additionally enables interoperable access to MSC information across the communities.

Keywords:

• OWL
• ontology encoding convention
• semantic mediation
• semantic enhancement
• semantic enrichment
• manufacturing service
• service-oriented manufacturing

Notes

1. Since the semantics of MSC information is generally embedded in both MSC descriptions and their corresponding MSC schemas, for the purpose of discussion in this article we use the term MSC model to refer to the combination of both MSC description(s) and their schemas. As a simple example, a supplier’s MSC description may state the supplier to have the Wire EDM (electro discharge machining) capability, which is a piece of MSC information. Then, the corresponding MSC model schema that states the Wire EDM to be a kind of EDM capability is an additional piece of MSC information. More complex axioms and rules may be contained in the MSC model schema to encode MSC information.

2. In this article, we focus only on engineering type of information within the MSC information and do not analyse other types of information (e.g. supplier's location or contact information) necessary to fully support customer-supplier interaction.

3. In the deep hole drilling category, which is another category that we have investigated, only 3 out of 403, or less than 1%, of suppliers provide capability details. Upon discussions with the portal operator, a cited reason is that suppliers may not know how to describe their manufacturing capabilities, indicating that a richer manufacturing service model with guidance could help.

4. In other words, only assertion axioms and declaration axioms are utilised in this transformation. Other types of axioms particularly class expression axioms and property expression axioms are not utilised. See W3C (2009c) for definitions of types of axioms.

5. We have simplified the IRI output from the D2RQ for readability and ease of understanding.

6. In this and subsequent encoding illustration we left out, for brevity, some columns that are not related to service information such as the isConcrete column in the ProductOrServiceCategory table. The purpose of the isConcrete column is to support a business rule indicating which category can be used to categorise the capability. It has nothing to do with the service information.

7. The instance db:AlloySteel_1 has no real semantic meaning in the sense that there is no difference if other instances such as db:AlloySteel_2 and db:AlloySteel_3 were created and used.

8. The mediation relating to data unit conflicts is outside of the scope of this article. The approach to deal with this within the realm of manufacturing sourcing using the manufacturing service models is deferred to future research. Interested readers are referred to existing works by Bijan and Smith (2008) and Hodgson and Keller (2011) that discuss the handling of units in OWL.

9. The min value is defaulted to zero here. Portal-A engineer may default it to other values such as a part size known to be the smallest that can be handled by EDM machines in the market.

10. It should be noted that material hierarchy and definitions illustrated in this article have not been reviewed by experts and hence should not be used as reference. The purpose of these figures is to illustrate the semantic mediation and that material ontology modelled in OWL DL can be useful.

11. Other alternatives may exist such as adjusting the reference ontology. We believe that there is a need for further research on guidance to design, usage, and mapping of OWL object and data properties. We are continuing to look into these issues in our research.