A holistic FMEA approach by fuzzy-based Bayesian network and best–worst method

Abstract Failure mode and effect analysis (FMEA) is a risk tool widely used in the manufacturing industry. However, traditional FMEA has limitations such as inability to deal with uncertain failure data including subjective evaluations of experts, absence weight values parameters, not considering conditionality between events. In this paper, we propose holistic overcome these limitations. The proposed approach uses fuzzy best–worst (FBWM) method weighting three parameters FMEA, which are severity ( S ), occurrence O detection D find preference modes according . On other side, it Bayesian network (FBN) determine probabilities modes. Experts use procedure using linguistic variables whose corresponding expressed trapezoidal numbers, parameter constructed BN. Thus, FBN expert judgments set theory addresses uncertainty includes robust probabilistic logic capture dependence As demonstration approach, case study was conducted an industrial kitchen equipment facility. results have also been compared existed methods demonstrating its robustness.