The research objective of this award is to create a software framework based on ontology representation that facilitates data exchange between various product development software systems. The development of methods to enable ontology-based interoperability ensures that the semantics of product data are preserved, and when necessary translated. The main focus is on determining maps between concepts across different product development software while addressing issues of external heterogeneity (differences across systems), internal heterogeneity (heterogeneous components of product data) and uncertainty (one-to-many maps). An ontology will be developed to represent knowledge-based inputs, e.g., Function, Assembly and Tolerances using the W3C standards, OWL and SWRL. Concepts from information theory will be coupled with probabilistic approaches to determine semantic similarities using available data. Fuzzy-set based approaches will be applied to determine semantic similarities for concepts that cannot be compared using crisp probabilistic approaches. A semantic interoperability test bed will be developed to validate our techniques and those developed by other researchers in this area. If successful, the results of this research will enable stakeholders in product development to express and exchange their intent instead of being restricted by the syntactic translators that lead to a loss of the intent. The use of W3C standards will facilitate the development of open-source software toolkits. The test bed will provide a suite of benchmarks, which can be used to evaluate competing techniques on a level playing field.. The ability to interoperate correctly will stimulate development of next-generation product development software. For small-business suppliers, this means that they can invest in just a few software tools that are the best for the nature of their job. This will enable engineers, managers, and technical professionals to be more effective, productive, and competitive.
