A new quality control method based on statistics of extremes for preventing recalls for mass production products

Abstract

Recently, recalls for mass production products such as car components have been frequently reported even for the case that the rate of defects is only of the order of ppm or less. The objective of this paper is to propose the solution to avoid the recall problem of the order of ppm for mass production products. Even if the defect rate is of the order of ppm or less, most of remaining safe products have to be recalled and be replaced by new components. Such a recall causes a great cost deficit if the very rare defect is possibly related to fatal accident. However, it is very difficult by the conventional quality control methods to find the defects of the order of ppm or less at the stage of design and production. This paper proposes a new practical quality control method to avoid the defects of the order of ppm or less for mass production products based on the statistics of extremes which has been successfully applied to fatigue strength evaluation of defective materials. First, several examples of the quality control method to avoid the troubles mainly caused by failures and damages of components will be presented. Next, it will be shown that the same approach also can be applied to other problems such as the optimum control of operational parameters and the selection of optimum materials through the index based on the statistics of extremes. It will be also shown that the same method can be applied not only mass production components but also to avoid the troubles and failure accidents for large machine components of small number production. The stress-strength model approach will be reviewed from the viewpoint of the statistics of extremes.