Time Series Forecasting of Airlift Sustainment Cargo Demand
Abstract
Forecasting demand for airlift of sustainment cargo is an important function for logistics planners. For the civil reserve air fleet participants (CRAF), having a useful long-range forecast enables them to make business decisions to maximize profit and manage their fleets. Because the DoD relies on CRAF for much of its steady-state and wartime surge requirements, it is important for these civilian enablers to stay financially healthy in what has become a difficult market. In addition to the CRAF airlines, DoD schedulers also benefit from somewhat shorter-term forecasts of demand, as accurate forecasts help them allocate aircraft type and determine route frequency for airlift of sustainment cargo. Time series forecasting is a method applied in many circumstances, to include forecasting of aviation service demand. It does not require the modeler to attribute causation, but rather uses historical data of a univariate series to predict future values. This paper applies a variety of time-series techniques to historical sustainment demand data from Iraq and Afghanistan AORs, ultimately choosing a single technique to develop into a prediction model for future demand in each AOR. The resulting models show excellent goodness-of-fit values and are successfully validated against a reserved portion of data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2012
- Accession Number
- ADA566289
Entities
People
- Daniel S. Deyoung
Organizations
- Air Force Institute of Technology