A comparative study of forecasting methods for sporadic demand in an auto service station

Spare parts are essential in the automobile sector and forecasting of spare parts has always been the vital prospects in an automobile service parts station. In this paper, a comparison and efforts have been made at the service station of a reputed organisation to reduce the errors in demand forecasting of intermittent demand items. The errors are compared for various methods using mean absolute scaled error (MASE) and Syntetos and Boylan approximation (SBA) method which exhibited the least error for intermittent demand and lumpy demand pattern. While single exponential smoothing method is used for smooth and erratic demand pattern. All calculations are done in MS Excel and solver tool to find optimal values of smoothing parameters.