Electro-polishing is a process having a merit of being able to obtain a smooth surface. In the process, it is necessary to adjust the processing conditions depending on the shape. But few reports have been published on the effects of processing conditions, which are mostly determined by experience.

In this study, an electro-polishing process was optimized for austenitic stainless steel using the Quality Engineering approach. From the results it was clarified that the factors including polishing voltage, anode surface, anode surface diameter and polishing solution flaw significantly affect SN ratio and sensitivity. Under the optimum condition, it became possible to continue working without being affected by the deterioration of the polishing solution.

Quality engineering is the knowledge/methodology known as Taguchi methods widely applied all over the world, especially in the U.S. Conventionally, sampling inspection in machining line has not been conducted optimally because sampling interval was fixed at Job #1 based on the past experiences and never reviewed unless a big problem occurred. It should have been determined consid ering practical process stability. In this study, we apply 'On-line Quality Engineering' to sampling inspection after finishing process at transmission case machining line. As a result of optimizing sampling inspection conditions (extending/shortening sampling interval, adljusting cutting tools & equipment) based on practical process stability, we have the prospect of improving robust ness of product quality and reducing inspection workload.

In this study, a trial was made to effectively detect the bugs in a software for vending machines using an orthogonal array. This debugging process was evaluated considering the environmental conditions where vending machines are used. It was found from the results that debugging could be made not only for the functions of vending machine's but for the ones that relate to the customers who buy the products. Bugs were detected from a small number of combinations. It was also found to be possible that the cost of software development will become lower.

Today, the size of semiconductor devices has been reduced to the deep-sub micron level. In such processes, it is important to control pattern size as well as pattern profile. This paper outlines the optimization of the photo resist pattern size and its profile. The optimization was performed using simulation and Taguchi Methods. Despite the fact that the dimensions of profile is difficult to measure by experimentation, it was successfully optimized using simulation. The pattern from the simulation well agreed with the one by observation. The study was completed in two days compared with two to three months using traditional approaches. Moreover, from the study of the dimensions that relate to the direction in thicness, the parameters for tuning the pattern were found by the analysis including an indicative factor. The results of this study will be useful in the future study.

The face-down bonding process is a process to directly connecting an IC to a base plate. Using this process, it is possible to save space. Due to the nature of the construction, connected parts are invisible and therefore, it is difficult to inspect the product. In this report, ACF (Anisotropic Conductive Film) was used for the face-down mounting. Experiments were conducted to optimize the bonding conditions. Current/voltage relationship was considered as the generic func tion and bonding conditions as control factors. The conditions before and after the accelerating test were used as the noise factor levels. From the results of analysis, the effects of control factors were clarified. Comparing with the traditional way where many samples were used with a long time to test, it became possible to evaluate the quality with a small number of samples in a short period of time. The feasibility of using the same approach to forecast the life of the final mounted product was also studied.

The image blurring in the Organic Photoreceptor (OPC) used in copy machines has been reported as a quality problem. To solve tile problem, Quality Engineering was applied to improve the quality of the OPC. In problem solving cases, time is essential as well as quality level. From this point of view, the evaluation using SN ratio and the Loss Function is especially useful. This example shows the effectiveness of applying Quality Engineering for solving the quality problems from the market.