Simulation-based Multi-Criteria Evaluation of Cost-Risk-Effectiveness in Prognostic Maintenance Operations: A Case Study from Railway Industry

Maintenance of physical assets plays a key role in a company's long-term profitability and has increasingly become an important element of business in the last decades. However, many organizations are under pressure to reduce their costs associated with maintenance operations while keeping high level of safety and reliability. Currently, the of maintenance programs in industries varies from corrective repair or replacement (through which the assets are restored to working condition in case of a failure) to prognostic programs (with the aim of detecting anomalies as early as possible to prevent a failure). To achieve the aim of reducing cost and increasing efficiency, the asset owners must determine what maintenance program should be adopted for each group of assets and also know when, where and how the maintenance procedures must be conducted. Simulation models are one of the most powerful techniques to assess the efficiency/effectiveness of maintenance programs. A simulation model has the capability to take into account system parameter uncertainties as well as to explore the feasibility of novel maintenance strategies. This paper presents a simulation-based multi-criteria approach to assess the effectiveness of various maintenance programs adopted for industrial plants and public facilities (e.g. preventive maintenance (PM), risk-based maintenance (RBM), condition-based maintenance (CBM), etc.) to save cost and reduce risk. A maintenance program with the best performance is chosen for execution by means of Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). A case study of passenger door systems of the Class 380 trains on the Scotland’s railway transport network is provided to demonstrate how to use the model. Based on the results, CBM was found to be superior to other maintenance programs for the train passenger door units, followed by mileage-based PM.