Forecasting Electronic Repair Parts Consumption for AMC Electronics Depots.
ARMY MATERIEL COMMAND TEXARKANA TX INTERN TRAINING CENTER
Pagination or Media Count:
The primary objective of the paper is to discuss the design of a forecast system for repair parts support of overhaul activities in electronics depots. Four major stages in the design process were data, models, methods, and error measurement. Approximately six years of actual repair parts consumption were used as a data base. Both time-series and probability models were employed to provide mathematical representations of overhaul activities. Using accuracy and computational simplicity as criteria, the following forecasting techniques were selected maximum likelihood, exponential smoothing, and quadratic. Finally, relative error was used as a measure of accuracy in comparing the techniques. A computer program was written to handle the tedious job of data manipulation and screening, application of the repetitive techniques, and error measurement. Since the quadratic scheme produced the lowest error mean and variance, it was recommended for implementation. Author
- Logistics, Military Facilities and Supplies