The Taguchi method represents a remarkable improvement of the classic experiment plan method. The Japanese specialist elaborated some standard instruments which simplify the experimenting process and which can be applied in a wide range of domains, that is to say 'a collection' of orthogonally fractioned experiment matrices and a set of 'accessories' - triangular tables and linear graphics which enables standard matrices to adapt to specific situations. In Taguchi's terminology, the factors that cause quality losses are called noise factors, while the performance levels targeted are called signals. Taguchi evaluates the quality of a product or of a manufacturing process through a performance synthetic indicator called signal/noise ratio, which takes into consideration both the average value of the targeted performance and its dispersion simultaneously. The paper also contains a case study regarding the quality improvement of the electrochemical cadmium plating in the drum in order to protect the representative parts of the hydraulic mining equipment. The quality criterion is the thickness of the cadmium layer target criterion. Finding the optimum configuration of the process requires the arbitration between representative factor levels in order to reach the double objective: minimising the variance and reaching an average thickness closer to the target value.

Volume 20

Numbers 11-12