This thesis improves the method of determining inventory levels for commodities (provisions, high usage load list consumables, and ships store merchandise) managed by the Sixth Fleet on station AFS. Historical demand generated by ships deployed to the Sixth Fleet is used to develop two models, the Lognormal Model and the Point Estimate Model. Improvement is achieved by considering each item's variance in demand. The Lognormal Model computes sample standard deviations for each item and provides the more accurate results. The Point Estimate Model uses regression to estimate a standard deviation for groups of items. Although the Point Estimate Model is easier for hands-on users to understand it is no easier to implement. The two models are compared against current procedures using a second set of actual Sixth Fleet data to simulate six months of inventory activity. Satisfied customer demands are improved by five percentage points (from 93% to 98%) and end of the month contingency inventory reserves are improved by 30 percentage points (from 65% to 95%). Keywords: Inventory analysis; Logistics planning; Statistical distributions; Mathematical models; Theses. (edc)
| Identifer | oai:union.ndltd.org:nps.edu/oai:calhoun.nps.edu:10945/27269 |
| Date | January 1989 |
| Creators | Bennett, Jeffery P. |
| Contributors | Henderson, NA, NA, Operations Research |
| Publisher | Monterey, California. Naval Postgraduate School |
| Source Sets | Naval Postgraduate School |
| Language | en_US |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0208 seconds