| The success of CAD/CAM integration depends on the ability of these two functions to communicate with each other. Computer-aided process planning serves as the key to this problem. A significant amount of work that provided a basic structure for process planning analysis has been conducted. However, in order to achieve integrated computer-aided manufacturing, the computer-aided process planning function must be totally automated.; Recently, the field of artificial intelligence (AI) has advanced to the point where AI techniques are producing encouraging results both in academia and industry. Some of the results involve the design and implementation of expert systems, which are intelligent computer systems using domain-specific knowledge to achieve a high level of performance in the field which we would normally consider as requiring a human expert.; A significant amount of process planning knowledge has been extracted, refined, and formalized in this research. The domain knowledge and a pattern-directed inference mechanism with a concurrent rule-firing concept is implemented in PROLOG. The system contains a machined surface identifier, a knowledge base, a backward inference engine, a machining data selection module, an automated NC part program generator, and an effective relational data base manager. The system is able to generate process plans from workpieces designed using a wireframe CAD modeler. The NC instructions associated with the workpieces are also automatically created. Several symmetric rotational parts have been designed, planned, and produced using this system. The results look very promising. |
