The research deals with methodology proposed for a final product data fully inspected for decision-making on improvement or continuing production by using some methods of forecasting . The proposed methodology was applied in the General Company for Electrical Industries-Motors factory for air cooled engine output (1/4 Hp) for different years and proposed two ways to arrange data due to its large size and irregular production quantities: the first method by dividing the data to chapters for the rejected production, size and the most frequent defects ,the cost and losses using the Histogram, the second method is by divided production data in batches of size (1,000) units produced and used two methods(Linear Trend Model and Single Exponential Smoothing ) to forecast and calculated the cost.
The search results showed that the single exponential smoothing is the best forecasting through the measurement of sum of relative squares error MAPE and the measurement of absolute deviation MAD and the mean average deviations MSD.