Abstract: Based on the discussion about the disadvantages and limitations of traditional Taguchi's squared error loss function, the inverse normal loss function and the extensive quality loss function model with multiple quality attributes are introduced. How to get the parameters and analyze the quality level (risk function) are discussed. Some examples of applying the model in manufacturing system decision making are given.

Key words: Quality evaluation, Quality loss function, Quality management, Controllable restrictor, High frequency motor spindle, High speed spindle, ultra-high precision machining, Hydrostatic bearing