After discussing various problems of the following statistical methods; (1) least square methods, (2) Bayes' theory, (3) Fisher's theory, and (4) mathematical statistic, in designing a hardware product, denoting signal factor M and its output y, the ideal function is defined as y=f(M). In the two stage design, control factors are assigned to be orthogonal array. From each experimental run, the following outputs due to the signal factor M are collected; N₀: standard condition, N₁: compounded smaller output condition of noises, and N₂: compounded larger output condition of noises. The most important point, in two stages design, is that we should not use cause and effect relationship between control factors and output. Standard signal to noise ratio and tuning method are explained here.

Thermoresistant bacteria survive in cooked food and cause the food to go bad. The detection of these bacteria takes time and the results are unreliable. Previous studies of inspection methods have proceeded by trial and error. The present study addressed the culturing of bacteria and the detection of the number of bacteria. Conditions for improving detection precision did not match the conditions for increasing the number of bacteria, but since analytical precision is critical, the conditions that improved analytical precision were selected as optimum. The dispersion of detection results was thereby reduced by a factor of 122, as compared with previous conditions, and the inspection period was shortened from one week to less than two days. Besides reducing inspection costs, this study made it possible to reduce social costs by enabling rapid response to dangerous situations.

In this study, mold design conditions were optimized by CAE parameter design to reduce variations in intake and exhaust port dimensions that influence the characteristics of a new type of rotary engine. The parameter design was performed by carrying out a direct product experiment without setting separate noise factors because, in view of the numerous types of error occurring at mold manufacturing sites and the complex interrelationships thereamong, optimization under various combinations of noise factors would be necessary, requiring lengthy CAE procedures. Efforts were made to simplify the CAE models, and a fully automated analytic system was constructed. To confirm their validity, the simplified models were compared with realistic models; the results indicated extremely high reproducibility. When factors with large effects were applied in actual core cylinders, dimensional variation of the intake ports was reduced by a factor of 1/2.5.

The function of an engine mount is to prevent engine vibration from being transmitted to the body of the vehicle. Target characteristics can be reached by iterative adjustment of control factor levels using frequency as a signal factor and vibrational acceleration as output. Conventional optimization methods can easily fall short of their targets. This report will discuss basic concepts and an iterative procedure for reaching the target frequency characteristic.

Our company has provided customers with 100% product quality assurance. We have also applied on-line quality engineering, mainly to process control in volume production, and demonstrated its efficacy in improving quality stability and consequently reducing inspection cost by optimization of inspection cycles. This study further explored the number of inspection items, types of inspection methods, in-process tolerances and other elements and evaluated their validity by loss functions, thereby pursuing the possibility of further improved quality stability. This paper reports results demonstrated in these activities, including the production engineering operation evaluation process in the production line design phase and implementation notes derived from the results.

The purpose of this study is the efficient promotion of on-line quality engineering by the introduction of the LPI methodology, in which on-line quality engineering is used as a tool, into a semiconductor fabrication process. In line with this purpose, the study is not limited to calculations involved in immediate problems; it also focuses on practical deployment. It will consider problems and solutions in each of the series of four guidance steps: recognition, proposal,cooperation, and self-support. In the recognition step, on-line quality engineering is promoted as a tool. In the proposal step, the promoters endeavor to bring the methodology to bear on a process with which production personnel are experiencing diffculties, and to create tools. In the cooperation step, production personnel take the initiative toward solving the problem. Various standards are drawn up to enable the methodology to be applied, leading toward self-support. The greatest issue in these activities is liberation from the myriad of existing constraints. This report describes the deployment of on-line quality engineering as a method of improvement and evaluation that involves a total rethinking of previous methods of improving semiconductot fabrication processes.

Garbage is flooding our everyday environment. In particular, the quantity of garbage disposed of by food-product companies and homes is increasing. A food recycling law has been enacted to reduce the quantity of garbage and promote recycling of resources, and a great number of garbage processors are now on sale, made by various manufacturers. The processing methods employed by these garbage processors can be broadly classified into two types: microbial degradation and drying. Garbage processors can also be classified as domestic or industrial depending on their size and processing capacity. Much research on the principles of the processing methods used by those processors is being carried out, but the processing conditions have not necessarily been optimized. The present study was aimed at implementing compact, low cost garbage processors to simplify garbage processing at home. The first step taken was to reduce the processing time in the drying method by optimizing the processing conditions. The optimum processing conditions were found by evaluating the results of quality engineering experiments using orthogonal arrays and S/N ratios.