Spare parts inventory control for an aircraft component repair shop

We study spare parts inventory control for a repair shop for aircraft components. Defect components that are removed from the aircraft are sent to such a shop for repair. Only after the component has been inspected does it become clear which specific spare parts are needed to repair it, and in what quantity they are needed. Market requirements for shop performance are reflected in fill rate requirements for the turnaround times for each component type. From a modeling perspective, the system is similar Assemble-to-Order systems. The inventory is controlled by independent (s, S) policies. We study the optimization of these policies. This problem is formulated as an integer program. We solve the LP relaxation with column generation, and develop an efficient method to solve the related pricing problem. This method is interesting in its own right because it works under more general conditions than existing methods. LP-based algorithms are developed to find integral solutions. These algorithms solve large-scale practical instances of the problem in minutes. We find that implementation of the algorithm at a repair shop improves cost efficiency, and allows for better alignment between inventory decisions and performance targets than traditional methods.