Big Data-Oriented Product Infant Failure Intelligent Root Cause Identification Using Associated Tree and Fuzzy DEA

Abstract

Infant failure analyzing is an effective approach to improve production quality continuously. The root causes of infant failure have always been a puzzle to manufacturers. To satisfy the increasing demand for the fuzzy root cause analysis of product infant failure in the era of big data, a novel root cause identification approach based on the associated tree and fuzzy data envelopment analysis (DEA) is presented for product infant failure. First, to decrease fuzziness with regard to the mechanism of infant failure, the associated tree is adapted to guide the analysis process for possible root causes based on axiomatic domain mapping. Second, considering the fuzzy mechanism and massive data, the fuzzy DEA technique is adopted to cluster all the potential factors of functional parameters, physical parameters, and process parameters from big data regarding product life cycle. Third, the ranking method of decision-making unit efficiency in fuzzy DEA is used to model and rank the weight of each node in the established associated tree of infant failure. Finally, a case study of root cause identification for a typical infant failure of the vibration and noise of a washing machine is presented to demonstrate the feasibility and validity of the proposed method.