Forecasting of Service Parts Based on Fuzzy Reliability of The Product

This paper presents a forecasting model depends on the reliability of product and the failure of its parts to forecast the required quantity of spare parts. Fuzzy logic is integrated with the forecasted model to treat the uncertainty that may be exist around defining the parameters values. Fuzzification of the product reliability is constructed using alpha cut and triangular fuzzy number. The effect of fuzzy process on the forecasted required demand of spare parts will be studied in three cases: 1) fuzzification of the mean of the product reliability, 2) fuzzification of the standard deviation of the product reliability, and 3) fuzzification of both the mean and standard deviation of the product reliability. Four suggested defuzzification methods (mean-max, centroid, signed distance, and graded mean integration representation) were used to figure out the difference between the crisp and the fuzzy forecasted demand with its related costs and to save the stock with the suitable production ranges. From the results, the maximum deviation between the crisp and the fuzzy forecasted demand was resulted from the fuzzification of both the mean and the standard deviation with percentage range from 2.06 up to 5.45 that would save the non-stock out than crisp forecasting.