Full metadata record
DC FieldValueLanguage
dc.contributor.authorHosang, Jürg-
dc.contributor.authorHeitz, Christoph-
dc.contributor.authorMönkeberg, Sigrid-
dc.description.abstractA simple stochastic model is proposed to forecast the demand for injection molding products. The model is based on individual orders, as characterized by their arrival time and the amount ordered. Both properties of the orders are described as random variables, the order arrivals by their rate and the quantities ordered by their mean and spread. Demand data for 383 products were analyzed, i.e. their model parameters estimated from the data. Based on visual inspection and quantitative analyses, it was concluded that 334 products could be described very well with the proposed model. The demand for some products is dominated by a few customers who regularly order large quantities. It was suggested to exclude these orders from the analysis and to treat them separately, e.g., by settling contracts with the relevant customers. For 36 products there were not enough data to estimate the model parameters and/or decide on the validity of the model. For another 13 products, there were systematic deviations of the data from the model. Many of the latter cases could be attributed to products which are demanded by only very few customers. For conclusion, the vast majority of the products could be described successfully with the simple approach proposed which is also suited for implementation in an operational forecasting system. Equations were developed which allow the demand to be predicted along with its uncertainty for forecasting horizons of arbitrary length. Today, smoothing techniques, based on time-aggregated demand data are widely used for predictions. One main advantage of our model over such approaches is that the problem is avoided of choosing an adequate period over which demand is to be aggregated. As a consequence, high-demand and low-demand products can be modeled equally well. Moreover, our model is not limited to one-period forecasts, i.e., one-aggregation-period-ahead predictions and its parameters are theoretically well defined and easy to estimate.de_CH
dc.rightsLicence according to publishing contractde_CH
dc.subjectStatistical analysisde_CH
dc.subjectCustomer behaviourde_CH
dc.subjectStochastic modellingde_CH
dc.subject.ddc658.5: Produktionssteuerungde_CH
dc.titleDemand forecasting for inventory control and production planningde_CH
dc.typeKonferenz: Paperde_CH
zhaw.departementSchool of Engineeringde_CH
zhaw.organisationalunitInstitut für Datenanalyse und Prozessdesign (IDP)de_CH
zhaw.conference.detailsIPLnet 2002 Workshop: From research to applications, National Network of Competence of the Swiss Universities of Applied sciences on "Integrated Production and Logistics", Saas Fee, 10-11 September 2002de_CH
zhaw.publication.reviewNot specifiedde_CH
zhaw.title.proceedingsFrom research to applications : proceedingsde_CH
Appears in collections:Publikationen School of Engineering

Files in This Item:
There are no files associated with this item.
Show simple item record
Hosang, J., Heitz, C., & Mönkeberg, S. (2002). Demand forecasting for inventory control and production planning. From Research to Applications : Proceedings.
Hosang, J., Heitz, C. and Mönkeberg, S. (2002) ‘Demand forecasting for inventory control and production planning’, in From research to applications : proceedings. Yverdon-les-Baines: IPLnet.
J. Hosang, C. Heitz, and S. Mönkeberg, “Demand forecasting for inventory control and production planning,” in From research to applications : proceedings, 2002.
HOSANG, Jürg, Christoph HEITZ und Sigrid MÖNKEBERG, 2002. Demand forecasting for inventory control and production planning. In: From research to applications : proceedings. Conference paper. Yverdon-les-Baines: IPLnet. 2002
Hosang, Jürg, Christoph Heitz, and Sigrid Mönkeberg. 2002. “Demand Forecasting for Inventory Control and Production Planning.” Conference paper. In From Research to Applications : Proceedings. Yverdon-les-Baines: IPLnet.
Hosang, Jürg, et al. “Demand Forecasting for Inventory Control and Production Planning.” From Research to Applications : Proceedings, IPLnet, 2002.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.