Expert System for Inspection Planning

Abstract

An expert inspection task planning system for Coordinate Measuring Machines(CMM) is developed in PROLOG. It is generative in nature and is bused on a feature-oriented computer-aided modelling system which has been developed in association with the planning system. The planning system takes into consideration the CMM characteristics, inspected part function and geometric properties, geometric tolerancing theory as well as the data extracted from the modelling system. New ideas regarding feature accessibility, feature representation in terms of the geometric elements used by CMM, inspection priority of datum features, measurement grouping according to the nature of geometric elements, checking all possible accessible features before changing a probe and/or part orientation, grouping tolerance checking tasks together based on corresponding datums, etc. and associated algorithms are proposed and implemented in this reseach. These concepts were verified and found to be feasible. Syntactic Pattern Recognition technique is used to identify a part directly from a feature-based CAD database from inspection planning point of view. The developed Feature-Oriented Modelling and Planning System, using artificial intelligence techniques, is discussed. Inspection knowledge representation and planning logic are described and illustrated with examples. Sample planning rules and inspection plan are also included. It is believed that this research can provide some useful contributions towards the development of automated inspection task planning, systems especially for coordinate measuring machines.