This study is an application of the Mahalanobis-Taguchi system to boring diagnostic and evaluation. When applied to boring, Taguchi methods can predict the precision of finished products, diagnose problems, and predict cutting precision from electricity data. This study deals with finished product prediction and diagnosis of problems. To predict the precision of finished products, an analysis was performed with data obtained from a waveform analysis of electricity data from the first processing as a data item and the S/N ratio obtained through parameter design as a signal. As a result, a prediction system was obtained with a precision level of 5 to 6 db, depending on the way in which the unit space was determined. For diagnosis of problems, the possibility of diagnosis by the Recognition-Taguchi system was examined with damage to the cutting blade during the experiment as the problem, and 30 items of electricity data obtained by 3-second waveform analysis as the data item. It was shown that problem diagnosis is possible by setting a separate unit space and threshold value for each processing state.
