A decision-making framework for optimal maintenance management: An integrated simulation-mathematical programming-expert system approach

In this study, an integrated simulation- mathematical programming- expert system approach is proposed to develop an optimal solution for maintenance management in a complex public infrastructure system. As the first step, a model of maintenance system is made through a computerized simulation with respect to the collected data on maintenance programs and unexpected failures. Then, the Taguchi method is adopted to develop different scenarios for each of which a precise measurement of cost and reliability based on customer dissatisfaction rate (i.e. modified reliability) is made. Regression method is applied to observe the association of components' maintenance workload with cost and modified- reliability. After that, the best solution for maintenance management is achieved by developing a mathematical model, with the aim of reliability improvement and cost savings. Finally, after identifying the most significant components in terms of cost and modified- reliability by sensitive analysis, the maintenance actions taken during the program are determined using an expert system. Results demonstrated that the optimization model Outdo Taguchi scenarios achieved the optimum one as it met both targets simultaneously. Having determined the significant components of the system and enhanced the methodology by signaling maintenance specialists to devote additional attention to them, it can be especially helpful in assisting administrators in lowering the accident rate arising from the shortcomings in maintenance plans.