Article ID: | iaor20012276 |
Country: | United States |
Volume: | 29 |
Issue: | 2 |
Start Page Number: | 479 |
End Page Number: | 497 |
Publication Date: | Mar 1998 |
Journal: | Decision Sciences |
Authors: | Kanet J.J., Sridharan V. |
Keywords: | scheduling, production: MRP |
Current research and practice in the field of production and inventory planning is dominated by hierarchical, material requirements planning-based methods. The objective of this research is to determine the merits of a major redirection for the design of manufacturing planning systems – a direction that more directly exploits the capabilities of modern computer technology. We test the relative performance of two fundamentally different classes of manufacturing planning systems: (1) those that make no use of sequencing and scheduling information in the planning of material deliveries (for example, traditional MRP-based systems), and (2) those that use sequencing and scheduling information to plan delivery of materials. We first derive an analytical expression for the expected benefit (in terms of flow time reduction) of using scheduling information for a single machine system. We then describe and test a simple heuristic for predicting the improvement that could be realized in more complicated multimachine systems, under certain conditions. Results from controlled simulation experiments, over a wide range of operating environments, suggest that although the value of scheduling information is influenced by the operating environment, it could be substantial.