Fuzzy inference systems for probabilistic failure assessment in asset-centric environments

Purpose Asset reliability is the likelihood of equipment performing as expected under normal conditions for a specific period. In today's competitive manufacturing landscape, accurately predicting unexpected failures of critical assets is essential to prevent safety incidents, minimize downtime and reduce costs. Design/methodology/approach This paper proposes a methodology to calculate an asset's probability of failure (PoF) effectively. The methodology identifies factors influencing equipment failure and uses the analytic hierarchy process (AHP) to assign weights to these factors. A fuzzy inference system (FIS) with linguistic terms, membership functions and expert-defined rules processes inputs such as operating conditions, maintenance history and environmental variables to estimate PoF. Sensitivity analysis validates the impact of each factor and weight. Findings The study presents a methodology for calculating the probability of failure (PoF) by integrating expert knowledge, historical data and environmental factors. It demonstrates enhanced accuracy and decision-making effectiveness across various industries. Originality/value The research improves decision-making by integrating expert insights into failure calculations and tailoring risk assessments to specific industries. It highlights the value of sensor functionality, and environmental considerations in optimizing equipment failure prediction and mitigating risks.