Production regions of the Atractylodes lancea rhizome, which is utilized as a material for Kampo medicines, have been inferred by use of MT system methods to analyze componential patterns of the plant. In a previous report, a new method referred to as the RT method, which compares the variance of the distane of the adjoint matrix, was applied to the pattern analysis, using the standardized signal-to-noise ratio. It has been found that a higher level of distinguishability is obtainable than with the RT method by applying a weighting strategy to the standardized signal-to-noise ratio, using the signal-to-noise ratio of a target-is-best characteristic. This method, which applies Taguchi weighting to the RT method, will be referred to as the RT method.

The MT method was used to study footbathing, a method by which elderly people can bathe easily and which has proven effective in reducing fatigue. In this study we increased the number of examinees to revalidate the footbath effect. When human activity has to be objectively evaluated, it is extremely difficult to set quanfitative evaluation standards. We therefore created a unit space thought to represent low human fatigue, and used questionnaire data employed in Kampo medicine and physical data, which could be measured relatively easily to make a comprehensive evaluation possible. We also compared foot-bathing with a simple rest break taken without a footbath, and performed a factor diagnosis to obtain changes in distances of factors relevant to fatigue, which showed that changes in physical condition are highly individual. Foot-bathing was also compared with full bathing, showing that full bathing produces more fatigue than footbathing.

Although some liquid crystal panel tests can be performed with testing apparatus,most are still sensory tests, done by human inspectors. Testing methods are developed to detect defects that have occurred in the past, so when a new type of defect is found, a new test item must be added, but this becomes a vicious circle. This study attempts to evaluate panels by the MT method, using good panels as a reference. The results indicate that this method may be practical for point defects, for which the Mahalanobis distance varies greatly with the severity of the defect. For area defects, problems were identified, including the necessity to have coordinate data.

Golf is one of the many sports in which players compete for scores. If a certain player's competence is assumed to be constant, the factors that determine his or her golf score are wind, temperature, etc. In this study, we tried predicting scores from these factors using the TS method. We also made a study of the optimal conditions for getting good scores, and in the final stage we actually played golf and compared the results. The data included mostly category data such as wind direction, which we numerized to perform the analysis. Athletics is just one area in which getting into condition for a big event scheduled for a certain date and time is very important. This study is applicable for such purposes,and also for maintaining the optimum state of manufacturing conditions,because although the problems to be solved and the data used are different, the way in which the numeric data are handled is similar.

Although color image forming technology has made remarkable progress, comprehensive methods of evaluating color image quality still rely on sensory tests. Image forming technology, however, has matured to the point that comprehensive evaluation by sensory tests is reaching its useful precision limit. As long as there is no highly precise scale that can be used for comprehensive evaluation, we have no choice but to depend on sensory tests, even though they leave us with very vague criteria for use in deciding what technology to adopt, which obstructs further technological progress. This study accordingly attempts to develop a comprehensive scale for evaluating physical properties to enable enginerring improvements to be made in ranges that cannot be recognized visually, by eliminating sensory factors from the signal space. On the assumption that the newly developed comprehensive evaluation scale will be used to improve engineering quality, a quantification of image evaluation is proposed in which both engineering and product quality are included, by conducting an engineering evaluation of the product qualities needed for acceptance in the market.

For domestic industries that practice powder metallurgy, a major issue is creating a system that can maintain a comparative advantage in cost competition with overseas industries by greatly increasing production speed by means of technical innovation in production methods. This optimization study aims at greatly increasing production speed while suppressing variation in weight by examining the relation between design factors in the pressure forming process and variation in weight experimentally. The ideal state of a powder filling machine is that the weight of the filling powder is proportional to filling time so that metal powder falls at a constant rate from the shoe box and the weight of the powder in the cavity does not vary. To evaluate powder filling, we optimized powder filling using a proportional relation between the dwell time of the powder box and the filling weight, and obtained a certain degree of reproducibility in the S/N ratio. The results of this study were applied to commercial production and succeeded in improving productivity 1.6-fold.