This chapter deals with various kind of prediction problems. Traditional ways of prediction are either physical models or the use of statistical regression. There has been no good mathematical model for earthquake prediction. The statistical method using multivariate analysis is a least square method. It is good only fbr the data used in the least square. Dr. C. R. Rao, a famous multivariate specialist stated in his book that any conclusion used is his book must be checked with different data. This chapter explains the method called Taguchi-Schmidt methodology for various kind of prediction problems. It is the methods to use the Schmidt expansion among all variables whithin unit group. For an example in case of prediction of total sales of a company, choose all companies whose total sales of this year was almost same as those of last year. It is called unit group. Various data of all companies should collected. After Schmidt expansion among all variable within unit group calculate the difference of variable of all companies. Then calculate error of predition of each variable using linear slope explained in this article.

At present, in theory, manufacturers' design criteria are used to evaluate the image quality of copiers when they are shipped. In fact, however, designers of these familiar devices are not completely aware of the images that users want,and outgoing inspections tend to be conducted on the basis of experience and judgement. In the first report in this series an image was prepared as a basis for evaluation, over 1000 copies of it were collected from different manufacturers'monochrome copiers in the field, and a unit space was constructed from the copies that users judged to be good. In the second report, the unit space constructed in the first report was reconsidered, and a unit space with greater discriminatory power was successfully constructed by improving the precision of the signal levels. In the present study, the feasibility of evaluating color copies by using this unit space was studied. A sample color copy judged to be good was established as a reference point, and the Mahalanobis distance from the reference point was calculated. It was confirmed that a unit space of monochrome images can be used to evaluate color images.

As machining speed is increased and processing time is decreased, transient or instantaneous power measurement has become necessary. Instantaneous power must be measured by measuring the voltage and current of a balanced three-phase alternating-current power supply directly. Taking a moving average before power calculation can provide power waveforms less affected by measurement noise even in machines incorporating inverter control. This paper introduces an appropriate instantaneous power measuring method and a data processing method by showing examples in which these methods were successfully applied to drilling.

In many injection molding optimization experiments, molded product shapes are evaluated using molding conditions as control factors. This study concerns the stability of plasticizing units, using the screw shapes that are essential for injection molding machines as control factors. In this study the function of a plasticizing unit was defined by the proportional relationship of melted resin weight to the product of the screw revolution time and speed, and the plasticizing interval was used as a noise factor. Analysis of an L18 orthogonal allay expeitment disclosed interactions among some of the control factors. Additional experiments and analysis were therefore performed on the control factors whose effects showed valley-like S/N curves. As a final result, an approximately 8-db S/N gain was obtained.

Simulation has many advantages over the use of prototypes: not only is it lower in cost, but results are also obtained in a shorter time, and quantities that would be difficult to measure on an actual product can be evaluated. This paper concerns parameter design carried out by simulation to stabilize the functions of an automobile suspension system. CAE (computer aided engineering) tools were used for efficient design; tuning was carried out to improve functionality by the standard S/N ratio, and to obtain a suspension with a shock mechanism capable of absorbing a large amount of energy. The analysis was automated from the system side, enabling extensive calculations to be carried out. This experience may be of reference in the future use of simulation in quality engineering.

Missing data preclude calculation of the Mahalanobis distance in the Mahalanobis-Taguchi system. The feasibility of evaluation using a digital standard S/N ratio in place of missing data was therefore studied, taking health examination data as an example. The significance of this study is that it showed when there are missing values in the data, the loss of discriminatory power of a health examination using the Mahalanobis distance can be minimized by use of mean values of healthy subjects as replacement data. The results of this study offer suggestions for dealing with missing data when the MT system is applied in a variety of other fields as well as in health examination.

Quality engineering in corporate management must help the corporation resolve issues and solve problems. Quality engineering Provides general metasolutions,but persons responsible for practical management do not realize how quality engineering can help them selve the problems they face. This study addressed the question of how quality engineering should be utilized in management to enable continuous growth of the corporation. In particular, this study identified managerial issues and considered their relationships with quality engineering.

Experiments in the evaluation of AA alkaline dry cells were performed to study functional evaluation, Experiments to evaluate functionality in five categories for different purposes were studied, the functionality being stability with respect to ambient temperature. The S/N ratio was used to compare the functionality of alkaline dry cells made by eleven manufacturers. In the evaluation methodology proposed in this study, product function was evaluated as well as functionality, and total cost was also considered. The results showed different S/N trends for each evaluation. The study demonstrated the importance of clarifying the definition of functionality and employing noise factors based on usage conditions, and the necessity of evaluating variability under those conditions. The above showed that it is possible to select suitable products by considering evaluations of functionality and function, total cost, selection problems, etc. in combination with the purpose of the products.