| The tradeoffs between push, e.g., MRP, and pull, e.g., just-in-time (JIT) or kanban, manufacturing control strategies are well documented in the literature. Hybrids of these two strategies are shown to be effective in comparison to pure push and pure pull. This thesis analyzes and compares six different hybrid push/pull manufacturing control strategies, as well as pure push and pure pull, for various manufacturing environments.; Five existing hybrid strategies are explained and a sixth is introduced. Simulation is used as a tool to compare these strategies for a fairly simple example from the literature. Work-in-process inventory as well as shortage (of finished goods inventory to meet demand) are used to rank the strategies. Perturbations in the environment are performed to get a feel for the robustness of the various strategies. The two most promising strategies are analyzed further. First, experimental design analysis shows which factors in the environment have the greatest impact on performance. Then, these two strategies are placed in an environment where they must compete for demand.; These strategies are compared in a similar manner for a more complex example. A portion of an actual consumer electronics assembly line is simulated. The objective in this example is to maximize throughput. Similarities and differences exist between the two examples. Results suggest that no strategy dominates all others under all the conditions analyzed, although CONWIP is shown to be a robust, consistent, and practical approach. In many cases, no significant difference in performance of selected strategies can be proven. In addition, the manufacturing objective and other factors within the environment, as well as strategy, are shown to have an impact on a system's performance. This study is significant for manufacturing managers who are dissatisfied with MRP or JIT and are seeking an alternative. |
