Managing Capacity Utilization with Advance Orders

Manufacturers that utilize capacitated flexible resources to produce a variety of products are often exposed to large mismatches between supply and demand due to some irregularities in their demand patterns. These irregularities arise from (1) uncertainty about the number of orders received from customers during a time period, (2) uncertainty about the quantity demanded for each order, and (3) uncertainty about the requested demand lead time for each order. The common advice in research and practice is to fully utilize capacity to keep a steady level of inventory, as long as there are some confirmed orders, regardless of the length of the requested demand lead times. This practice causes colossal inventory costs for the manufacturers. We develop an analytical model to solve this problem and apply the model to Kordsa Inc., a market leader that develops reinforcement technologies and produces composites for manufacturers in the auto, aerospace, construction, and infrastructure industries. Our analytical framework first captures the three irregularities by correctly characterizing the demand process with a multiplicative jump-diffusion process. It then employs the fast Fourier transform (FFT) to determine the capacity utilization rates given a service level target. Using the data from the company, we show that our approach increases inventory turnover by 91.3% over a practical benchmark of the material resource planning (MRP) model and by 21.3% over a theoretical benchmark of the intermittent demand model without sacrificing high service levels. Consequently, our model has been adapted into a decision tool used by the company