Summaries - Office of Research & Innovation
Research Summaries
Back Wholesale Inventory Optimization Model with Distribution Requirements Planning Simulation
Fiscal Year | 2019 |
Division | Graduate School of Operational & Information Sciences |
Department | Operations Research |
Investigator(s) |
Buss, Arnold H.
Salmeron-Medrano, Javier |
Sponsor | Naval Supply Systems Command Weapon Systems Support Philadelphia - N52 (Navy) |
Summary |
During fiscal year FY 2019, the Principal Investigators and other Naval Postgraduate School personnel will develop the Wholesale Inventory Optimization Model (WIOM) with Distributed Requirements Planning (DRP) Simulation. WIOM-DRP will incorporate two major modules: An optimization system that builds on the existing WIOM 4.10 development, but whose underlying structure will be substantially modified; and, a completely new DRP simulation system. In addition, WIOM-DRP will develop software architecture for a graphical user interface (GUI) module, that should enable planners to easily (a) manipulate and visualize input data and results, and (b) coordinate access to and work between the simulation and optimization modules. Development and computational implementation of the three modules is the object of the proposed work. Several reasons motivate the development: a) A large majority of items (including repairables and non-repairables) cannot be accurately forecasted as items with “consistent” demand (in the sense of low variability). Most items have either infrequent (intermittent, such as one or two units per year) demand, or frequent demand but with high variability. The consequences of making assumptions such as adherence to standard probability distributions results in inaccurate estimates of fill rates as rendered by those distributions. This effect is further aggravated by the use of closed-form formula (currently embedded in WIOM, with adjustments) that approximates fill rate, which has variable accuracy depending on the cycle distribution and the set reorder point. b) WIOM assumes a deterministic lead time, therefore ignoring valuable uncertainty information available for some items. In its parametric version, WIOM uses a lead-time demand (PPV) mean that is inputted as a function of the lead-time variance. This, to some extent, mitigates the issue. In its historical-demand version, WIOM disregards lead-time variance. c) Repairable items constitute the majority of wholesale items. WIOM assumes an oversimplified version of the actual procure-and-repair processes that take place in DRP for repairable items. Specifically, order quantity (Q) and PPV are inputted in WIOM as weighted averages of the procure and repair pipelines, both of which have different characteristics. |
Keywords | Inventory Optimization Simulation Weapon Systems |
Publications | Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal |
Data | Publications, theses (not shown) and data repositories will be added to the portal record when information is available in FAIRS and brought back to the portal |