DISPO 4.0 | Simulation-based optimization of stochastic demand forecast of intermittent material in the capital goods industry

This research introduces a digital planning method tailored to intermittent demand, leveraging simulation-based optimization to select and parameterize forecasting techniques for individual items. Accurate demand forecasting is crucial for optimizing inventory and order quantities, especially in spare parts planning, where demand is often intermittent. Despite the recognized value of digital solutions in this area, practical implementation of optimized forecasting tools for intermittent demands remains limited. To address this gap, the proposed approach offers a practical methodology for systematically selecting and configuring forecasting techniques specifically suited to irregular demand patterns. The method combines a rule-based heuristic with a static simulation of demand time series and metaheuristic-based optimization to calibrate forecasting parameters. The outcome is an automated, item-specific forecast tailored to the characteristics of intermittent demand. The approach is validated through two practical case studies from the capital goods sector, demonstrating its effectiveness in enhancing forecast accuracy and improving intermittent item planning.