Integrating views of properties in models of unit manufacturing processes
Peter O. Denno, Duck B. Kim
This paper investigates the potential advantages and pitfalls of directly integrating knowledge expressed as predictive model equations with other process knowledge. The context of this integration is models of unit manufacturing processes. The method described uses metamodels and semantic web technology to relate equations, as objects, to downstream activities. The potential advantages of this include enhanced knowledge refinement and reuse, traceability, model verification, and agility in production activities. In an example usage, we apply the method to the development and downstream usage of predictive models of a selective laser sintering process. Use of equations as objects enables linking with supporting evidence, property definitions, and dimensions in an engineering notebook paradigm. Metamodel-based interpretation of the equations enables composition in trade-off studies and mapping to downstream process parameter optimization.
International Journal of Computer Integrated Manufacturing
modeling methodology, predictive models, design of experiments, unit manufacturing processes, metamodeling, ontologies