An integrated multi-criteria decision-making approach for the risk assessment in the automotive parts industry

Failure modes and effects analysis (FMEA) is a structured method frequently utilized to recognize and eliminate possible failure modes. In the overall FMEA procedure, failure modes are evaluated and prioritized based on the risk priority numbers (RPN), calculated by multiplying three parameters: frequency (F), non-detection (ND), and severity (S). However, the conventional FMEA method fails to cope with the uncertainty in complex systems. To address the challenge of uncertainty, an adaptive neuro-fuzzy nnference System (ANFIS) is used as a dynamic selected model where these factors are represented as components of a fuzzy set that has been fuzzified utilizing the suitable membership functions. In addition, to enhance decision-makers' confidence a new hybrid model for acquiring a more logical ranking of failure modes is suggested. Essentially, two parts of the evaluation procedure are explained: determining the weights of risk parameters using the Fuzzy Analytic Hierarchy Process (FAHP) and ranking the failure modes using a mathematical rough set theory and 'Technique for Order Performance by Similarity to Ideal Solution' (TOPSIS) method. Key results demonstrate that the proposed model resolves the shortcomings in failure mode categorization by introducing additional parameters, such as cost and treatment duration, which improve the precision of risk prioritization and deliver a robust preventive-corrective plan. The proposed rough TOPSIS approach identified FM4 (Burr/chip contact with eyes while grinding) as the most critical failure mode with a score of 0.312, followed by FM3 and FM8. This systematic approach categorizes failure modes into clear risk levels (intolerable, major, moderate, and acceptable), facilitating more effective decision-making. A comparison with existing methods validates the reliability and adaptability of the suggested approach for real-world applications.