Interoperability of formal semantics of production data across product development systems.

Abstract

There are several shortcomings of current product development systems regarding their suitability to support a collaborative, but distributed design environment. This dissertation addresses problems in the interoperability of product information between different application software. First, there is a need to incorporate information about material gradation along with solid model data to represent and exchange heterogeneous objects. An information model to represent 3-D heterogeneous objects by extending ISO 10303 is developed. The data planning model is then extended to represent the two-dimensional (2-D) slice information for process planning. Further, designers who collaborate across geographic, cultural and temporal differences use different software to enhance the product from their own perspective. Collaboration in such an environment demands a translation of semantics across the different application software. In order to enable semantic interoperability, this work employs an interlingua approach and presents building blocks to develop a Product Semantic Representation Language (PSRL) after analyzing existing and next-generation product representations. The extensibility of the PSRL makes it a suitable basis to integrate various application domains. Effective exchange of information is not possible without determining semantic maps. After a systematic analysis of the problem of determining semantic maps, this dissertation presents a procedure to determine exact semantic equivalences through an effective use of the formal logical interpretation provided by the PSRL. The absence of a formal basis to define correct interacting ontologies, each of which contains all the concepts or their valid interpretations leads to inexactness in the interpretation of the meanings. This dissertation presents a detailed analysis of the corresponding challenges. It analyzes desired features of similarity of product information and utilizes a composition approach to determine semantic similarities. Using the semantic maps determined at the higher level, enhanced ISO 10303 can be used to effectively translate product information at the physical level, while ensuring that the semantics are preserved or translated.