A probabilistic approach to reliability assessment of inventory management strategies in an MRP environment with random yield

Abstract Although there are several strategies for planning production or replenishment in a material requirement planning (MRP) environment, they are all designed to calculate planned stock parameters. Other metrics, such as more precise probabilistic estimates of safety stocks, shortages and the reliability of the entire inventory management system, are rarely considered. Improving the accuracy of probabilistic estimates of these indicators is necessary to enhance the reliability of describing production processes under conditions of uncertainty. The article proposes calculated dependencies that allow determining the safety stock level based on analytical calculations rather than practical experience or arbitrary assumptions. A general algorithm was developed for selecting an inventory management strategy under random yield that takes safety stocks and random yield into account. This algorithm includes estimates of efficiency (order costs, on‐hand stock holding costs and safety stock holding costs) and material planning reliability indicators in the form of failure‐free probability during the planning period. Analysis of the results revealed that more precise probabilistic consideration of safety stocks in MRP under random yield leads to a change in the production planning or replenishment strategy.