Demand Modeling With Forecast Evolution: An Application to Production Planning

The value of demand forecast information has been the subject of numerous studies in the inventory literature, but has had limited application in production planning. This paper integrates ideas from forecast evolution and inventory theory to plan work releases into a production facility in the face...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on semiconductor manufacturing Vol. 28; no. 3; pp. 374 - 384
Main Authors Albey, Erinc, Norouzi, Amirhosein, Kempf, Karl G., Uzsoy, Reha
Format Journal Article
LanguageEnglish
Published IEEE 01.08.2015
Subjects
Computational modeling
demand modeling
forecast updating/evolution
Planning
Predictive models
Production planning
Semiconductor device modeling
Stochastic processes
forecast updating/evolution
demand modeling
Production planning
simulation
chance constraints
linear programming
rolling horizon
Online AccessGet full text

Cover

More Information
Summary:The value of demand forecast information has been the subject of numerous studies in the inventory literature, but has had limited application in production planning. This paper integrates ideas from forecast evolution and inventory theory to plan work releases into a production facility in the face of stochastic demand. The model, which is solved on a rolling horizon basis, manages the inventory, backorder, and shortfall levels at each planning epoch using chance constraints to represent desired service levels. The evolution of demand information over time is modeled using the Martingale Model of Forecast Evolution. The model is implemented using historical data from a major semiconductor manufacturer. The value of forecast evolution is quantified using a benchmark model that does not consider forecast evolution.
ISSN:0894-6507
1558-2345
DOI:10.1109/TSM.2015.2453792