Evaluation of Demand Prediction Techniques
Abstract
Reparable spares of aircraft components constitute an important management item for the Air Force, amounting to a computed budget requirement of $4.077 billion in FY85. Allocation of this investment a cross items is critical to the readiness and sustainability of weapon systems. Proper allocation, in turn, depends on solving the statistical problem of estimating item demand rates and variances. Using historical Air Force data, we have compared the performance of various estimating procedures, including the one used in the Air Force Recoverable Consumption Item Requirements (D041) System, which computes item Peacetime Operating Stock (POS) requirements. To be consistent with Air Force orientation toward weapon system management, aircraft availability, an aspect of readiness, served as the measure of performance. The major conclusions were that the mean demand should be estimated using exponental smoothing and the variance- to-mean ratio as a power function of the mean demand. When compared to current AFLC policy, the proposed techniques showed a reduction in backorders of over 50 percent for the F-16 and A-10 weapon systems.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 1987
- Accession Number
- ADA178547
Entities
People
- Craig C. Sherbrooke
Organizations
- LMI