9612067 Lu This award provides funding for research to develop a model of the responsive process planning system in agile manufacturing with a focus on automated determination of the sequence of machining operations in agile manufacturing environments for axisymmetric machined parts. An approach using case-based reasoning to support learning capability will be employed. The methodology to accomplish various issues in case-based operation sequencing will be formalized. In particular, the following issues will be tackled. First, a feature-based representation will be used for operation sequencing case representation. Then, a method called model-based similarity measurement, based on a domain-specific hierarchical model, will be applied for case indexing and case retrieval. Furthermore, the adaptation of an operation sequencing case including case modifier, plan simulator, and plan repairer will be developed. The system model will be implemented to demonstrate the learning functions of the system in response to rapidly changing markets and manufacturing environments. The generated process plans will be evaluated. If successful, this research will lead machine learning into the realm of practical application in agile manufacturing and provide a way to guide a computer-aided process planning system towards the level of human intelligence. It also offers to capture best practices through industry interaction and thus prevent the loss of knowledge through the attrition of experienced personnel. Furthermore, the proposed process planning system, with enhanced intelligence through learning, should respond rapidly to fast changing demands and enable the realization of the agile manufacturing concept of shorter time-to-market and the increase of product quality. It is expected that this model could be extended to other activities, such as product design, metal forming planning, and assembly planning, in the industry.
