Computer-aided inspection data analyser

First part inspection, in a discrete-parts manufacturing environment, is used all over the world to assure quality and stay competitive in this global market. A major limiting factor for the wide use of this method is the expensive and time consuming process of analysing the inspection data to trace the error sources that might have caused the part to be dimensionally inaccurate. Automation of this process can cut both the time and cost involved. The deterministic dimensional errors produced during milling operations, which is the focus of this work, are identified and classified as machine, fixture, and tool related. A methodology by which the error sources can. be qualified using solely the inspection report is developed using both an algorithmic and a heuristic approach. The PC-based analyser has been successfully validated by processing several sample prismatic parts through the flexible manufacturing cell at the University of Maryland.