An enhanced failure mode and effect analysis method based on preference disaggregation in risk analysis of intelligent wearable medical devices

Abstract

Conducting a risk analysis on potential failure modes that may damage the performance of intelligent wearable medical devices is imperative since the failure of the devices could directly impact human health. Failure mode and effect analysis (FMEA) is an evaluative instrument for potential failure modes in risk management. This paper presents an enhanced FMEA technique grounded in preference disaggregation analysis considering the interrelationships between failure modes to improve the precision of risk analysis. First, the initial evaluation of failure mode occurrence is updated by the overall influence-strength matrix among failure modes. The matrix formation considers the indirect interrelationships between failure modes, the positive/negative effects of failure modes, and the initial strength of failure modes. Then, a preference disaggregation method is applied to derive the weights of risk factors and the overall utilities of failure modes from historical decision examples. Failure modes are categorized from the most severe to the least severe according to their utilities. Smart bracelets, as a type of intelligent wearable medical devices, apply artificial intelligence technology in health monitoring. Through an illustrative case study of smart bracelets, the efficacy of the proposed approach is validated.