A hybrid FMEA approach for risk analysis of cold chain logistics and distribution system in developing countries: the case of China

The rapid development of cold chain infrastructure presents new challenges, making the risk analysis of the cold chain logistics and distribution system (CCLDS) critical for ensuring the quality and safety of temperature-sensitive goods. As a powerful risk management tool, failure mode and effect analysis (FMEA) can assess the risk levels of failure modes, but its limitations restrict its effectiveness. Therefore, an improved FMEA model is proposed. Specifically, the decision-making trial and evaluation laboratory (DEMATEL) is first utilised to quantify causal relationships among potential failures, followed by the evaluation of failure risks using interval-valued intuitionistic fuzzy sets for more precise expert assessments. Moreover, the objective weights of risk factors and experts are determined by the anti-entropy weight method and grey relational analysis, while the subjective weights are obtained through the improved G2 weight method and consensus decision based on the deviation. These are then integrated into comprehensive weights using the deviation maximisation method. Furthermore, cloud models are used for failure modes risk ranking, and k-means clustering is employed for risk classification. Finally, the proposed FMEA model is validated through a representative Chinese CCLDS case, demonstrating its improved failure-mode identification, reliable risk prioritisation, and a transferable framework for developing-country cold chain logistics.