Strategic buffer management utilizing time based order replenishment in a multi-level physical distribution environment
Watson, Kevin James
MetadataShow full item record
The purpose of this dissertation is to develop a more effective planning and control heuristic for the multi-item, multi-echelon distribution environment. This environment is characterized by multiple products flowing through a distribution system containing at least one source location and one or a number of receiving locations with the most complex of these systems allowing for transshipment of inventory between receiving locations. The first step in creation of the improved distribution planning heuristic was a review and critique of the relevant literature. Problems are identified with the current state-of-the-art multi-item, multi-echelon inventory control techniques, namely Distribution Resource Planning. Second, a field study was undertaken to determine the characteristics of this demand environment and to study the use of a DRP system in a realistic setting. Third, an improved heuristic for multi-item, multi-echelon inventory control was developed based on the logic of the Theory of Constraints to address the problems identified in part one. Finally, a simulation was developed in order to compare the results obtained by applying DRP and the improved heuristic to the demand environment established in part two. The simulation models a multi-echelon physical distribution system containing one manufacturing facility, two distribution centers, and six retail locations with up to 111 distinct stock keeping units. The simulation model also exhibits seasonality in creation of customers and allows partial lost sales. Simulation results show that the TOC heuristic is able to produce significantly higher levels of customer satisfaction while reducing the total cost of distribution and reducing the amount of demand variability received at the manufacturing facility.