Abstract
A reliable demand forecasting method is required to improve the overall planning of remanufacturing. Drawing on a Holt-Winters method and Weibull distribution method, prior research proposed a method to forecast demand for auto parts remanufacturing. To examine the effectiveness of the proposed method, the authors used the actual time-series data of the sales of remanufactured alternators. The average error rate between actual and estimated data were 35.3% for Holt-Winters method and 41.7% for Weibull Distribution method. To improve the accuracy of demand forecasting, this paper modifies the Weibull distribution by applying limitation method and categorization method. By applying modified method, the average error rate of forecasting was improved to 24.9%.
