A novel methodology for analyzing variation risk introduced by the manufacturing process in microsystems

This paper describes a novel methodology for analyzing the variation risk introduced by manufacturing processes in microsystems using a Key Characteristic (KC) method and statistical analysis applied to microelectromechanical systems (MEMS) fabrication tolerances. The KCs of the microsystems are identified based on the product requirements and then propagated to the subsystem KCs and associated manufacturing processes. Statistical approaches are given for estimating the variation of the product KCs, the variation contribution of subsystem KCs and the associated manufacturing processes. The KC variation risk is predicted and the manufacturing processes that contribute most to the variation are identified. This methodology is applied to an innovative design for a micro CMM (co-ordinate measuring machine) probe. One product KC and its subsystem KCs are identified and a systematic view of KC propagation structure is demonstrated. The variation in the product KC is predicted and the variation contribution of the MEMS probe manufacturing processes is presented. Variation risk analysis results can be used to implement a variation risk mitigation strategy.