Reliability centered maintenance actions prioritization using fuzzy inference systems

Deploying an effective maintenance strategy across an organization stands out as an essential risk mitigation measure that plays a critical role toward improving the reliability and availability of production facilities. The purpose of this paper is to propose a simple, yet well-structured approach toward prioritizing maintenance actions as part of a reliability centered maintenance (RCM) implementation plan, and selecting the most important subset of those actions subject to time and budget constraints.

A comprehensive RCM actions prioritization methodology is proposed using four criteria: severity, benefit to cost ratio, customer satisfaction, and easiness of action implementation. The method utilizes fuzzy inference system (FIS) to incorporate subject matter experts’ feedback into the decision-making process. The output of the FIS, which takes the form of a numerical weight that assesses the relative importance of each criterion, is then fed into a binary integer program that selects the optimal maintenance actions out of a set of possible actions.

The implementation of the developed methodology is demonstrated using a real-life example of a hydraulic brake system circuit that is used in construction equipment. The computational results illustrate the validity of the proposed approach and indicate that the selection of which maintenance actions to carry out is impacted by the relative importance (i.e. weight) of the considered criteria.

The work presented in this paper provides the decision makers with a systematic procedure that helps in selecting the most relevant maintenance actions instead of making the selection in a complete ad hoc manner or based merely on subjective opinions.