Data-Driven System Reliability and Failure Behavior Modeling Using FMECA

• Published in: IEEE transactions on industrial informatics Vol. 12; no. 3; pp. 1253 - 1260
• Main Authors: Khorshidi, Hadi Akbarzade, Gunawan, Indra, Ibrahim, M. Yousef
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.06.2016; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Budgeting; Case studies; effects and criticality analysis; Failure; Failure modes; Genetic algorithm; Indexes; Maintenance management; Modelling; Optimization; Overall failure index; Particle separators; Probability distribution; Qualitative data; Reliability; Reliability modelling; Risk assessment; Sensitivity analysis; System reliability; Uncertainty; Universal generating function; Welding; Effects and criticality analysis; overall failure index (OFI); failure modes; universal generating function (UGF); genetic algorithm (GA); qualitative data; uncertainty; reliability modeling
• ISSN: 1551-3203; 1941-0050
• DOI: 10.1109/TII.2015.2431224

System reliability modeling needs a large amount of data to estimate the parameters. In addition, reliability estimation is associated with uncertainty. This paper aims to propose a new method to evaluate the failure behavior and reliability of a large system using failure modes, effects, and criticality analysis (FMECA). Therefore, qualitative data based on the judgment of experts are used when data are not sufficient. The subjective data of failure modes and causes have been aggregated through the system to develop an overall failure index (OFI). This index not only represents the system reliability behavior, but also prioritizes corrective actions based on improvements in system failure. In addition, two optimization models are presented to select optimal actions subject to budget constraint. The associated costs of each corrective action are considered in risk evaluation. Finally, a case study of a manufacturing line is introduced to verify the applicability of the proposed method in industrial environments. The proposed method is compared with conventional FMECA approach. It is shown that the proposed method has a better performance in risk assessment. A sensitivity analysis is provided on the budget amount and the results are discussed.