Investigation of different forecasting methods and models using a Forecast Support System to forecast spare parts demand in the MRO industry

This paper examines the development and implementation of diverse forecasting methods and models within a tailored Forecast Support System (FSS) designed to address the distinctive challenges of spare parts demand in the Maintenance, Repair, and Overhaul (MRO) industry. The study underscores the intricacies associated with intermittent and lumpy demand patterns, which present considerable forecasting challenges. The research demonstrates the potential of a tailored FSS to enhance forecast accuracy, optimize inventory management, and reduce operational costs by implementing and evaluating multiple forecasting methods, including univariate time series, multivariate regression, classification, and clustering, as well as forecasting models. The findings highlight the necessity of selecting appropriate forecasting methods and models based on specific demand characteristics, thereby facilitating more informed decision-making and enhanced logistical performance of MRO service providers. Future work will focus on enhancing and refining the FSS, integrating additional forecasting methods and models, and extending the system to include safety stock dimensioning approaches, thereby further enhancing its utility in managing spare parts demand under uncertain conditions.