An Evaluation of Forecasting Methods that Could be Used in the Brazilian Air Force Uniform Distribution Process
Abstract
Every year the Brazilian Air Force (BAF) spends the equivalent of approximately 15 millions of dollars for uniforms. These purchases come from a tight budget, are executed through public procurement processes, and are tied to Brazilian acquisition regulations, which are often very strict. For this reason, lead times are unpredictable. It can take anywhere from one month to a year to replenish an item. The purpose of this research is to analyze the forecasting process performed at a BAF military organization named Sub-directorate of Supply (SDS) with the intent of building an algorithm comprised of a selection of forecasting models in order to help SDS optimize its inventory investments. With this in mind, monthly sales, prices, and inventory records from January of 2010 to July of 2015 were extracted from a database and converted to a standard spreadsheet format. Several forecasting models were evaluated and applied to randomly selected items from the database to create the algorithm. In the final analysis, it was concluded that two models precisely depicted the behavior of sales in BAFs stores. These two models were then utilized to develop the forecasting tool that may prove valuable in future BAF uniform purchasing decisions.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 24, 2016
- Accession Number
- AD1054003
Entities
People
- Leandro V. De Oliveira
Organizations
- Air Force Institute of Technology