Failure modes and effects analysis (FMEA) is widely recognized tool to identify the
potential failure risk for improving the reliability and safety of new product development. A major task of FMEA is to determine the risk priority number (RPN) of the each potential failure mode based on the assessments of three risk indexes: severity, occurrence and detectability. Unlike the conventional practice, these indexes are measured by discrete ordinal scales, but treated as numerical one, this study adopts the 2-tuple fuzzy linguistic approach to
represent the assessments of three risk indexes and develop a FMEA assessment model.
Applying the characteristics of the computing with words, 2-tuple fuzzy linguistic approaches composed a linguistic term and a numeric value assessed in [-0.5, 0.5） to overcome the drawback of conventional practice in FMEA. Besides, related with severity index, the reasonable assessments of occurrence and detectability indexes for potential failure modes are difficult to obtain due to lack information and uncertainty during the early stage of new
product development (NPD). This study employs the idea of quality function deployment to
determine the assessments of occurrence and detectability indexes using the
effect-cause-effect relationship among three risk indexes. Furthermore, three risk indices can be considered with different weights to obtain the more reasonable RPN. The outcomes of RPN to the corresponding potential failure modes are useful to decision maker for risk
management and arranging the limited resource to reduce or eliminate the serious potential
failure occurring in NPD. An illustrative example is used to demonstrate the applicability of the proposed model.
International Conference on Innovation and Management, 馬來西亞, 2011.7.12