Prediction of Manufacturing Resource Requirements for Mass-Customization Production

摘要

This research introduces various methods to predict manufacturing resource requirements based on sales forecast data for an industrial company -Gienow Windows and Door considering numerous variations of products in a mass-customization production environment. In this work, first the relations between resource requirements and sales at the levels of company, sales branches, and markets in sales branches are studied. Then linear regression and neural networks are employed to model the linear and non-linear relations between resource requirements and sales. These different resource requirement prediction methods are evaluated in terms of prediction accuracy.