Date of Award

9-1995

Document Type

Thesis

Degree Name

Master of Science

First Advisor

Terrance Pohlen, PhD

Second Advisor

Lee Lehmkuhl, PhD

Abstract

Current Air Force demand forecasting systems, DO41 and REALM, assume reparable demand is solely flying hour driven. The purpose of this study is to evaluate the relationship between demands, flying hours, and number of sorties at the work unit code level and improve reparable demand forecasting. A three phase methodology is used to analyze the demand, flying hour, and sortie relationship. The first phase uses multiple regression to determine a relationship at various work unit code levels. Multiple regression provides limited correlation between demands, flying hours, and sorties. The second phase uses Poisson regression to evaluate the integer, count nature of the demands variable used in the analysis. Poisson regression also exhibits poor correlation between demands, flying hours, and sorties. The third phase fits a Poisson process to the data and produces better results than multiple or Poisson regression. However, the Poisson process performs poorly in estimating future demands at the work unit code level, based on historical flying hour and sortie demand rate occurrences. Despite research at the work unit code level, the study results support previous demand forecasting research, which has been unable to demonstrate an accurate demand, flying hour, and sortie relationship.

AFIT Designator

AFIT-GLM-LAL-95S-9

DTIC Accession Number

ADA300683

Comments

The authors' respective Vita pages are omitted.

Presented to the Faculty of the Graduate School of Logistics and Acquisition Management of the Air Force Institute of Technology.

Download

Share

COinS