A Large Group Decision-Making Model for FMEA With Distributed Partial Evaluation Information and Parallel Communities

Failure mode and effect analysis (FMEA) is a popular group-oriented method to improve reliability and quality through testing and eliminating hidden failure modes in products, processes and services. As the complexity of real-world industrial systems increases, different components with multidisciplinary and cross-functional natures may have a large number of failure modes, which should be assessed by a large number of FMEA members in a distributed setting. Moreover, due to limited understanding, an FMEA member in the group may only provide partial evaluation information for a part of failure modes according to their areas of expertise. In this study, we introduce a large group decision-making model to solve FMEA problems considering partial evaluation information. To reduce the dimension of the large group, an inclusion measure-based group division method is used to classify members into parallel communities. This process can achieve a match between members and their corresponding failure modes required to be evaluated. Then, to manage conflicts within communities, a hierarchical consensus mechanism combining type $\alpha $ consensus and type $\gamma $ consensus is introduced. A case study about risk evaluation for floating offshore wind turbines is provided to verify the effectiveness of the proposed model.