(A) (B)
FIG. 31.2 (A) Normal precision and (B) Higher precision.
31.8 ROBUSTIFICATION
Robustification, as per Wiktionary, is to make designs more robust, that is,
more tolerant of unexpected events. Robustification, also called robust param-
eter design, is the process of determining the settings of the control factors that
minimize response variation from uncontrollable factors. This can be done by
understanding and controlling the interaction between these variables. While
these variables may easily be controlled in an experimental setting, outside of
the experimental world, they are very hard and are called noise variables.
This case study makes us consider if the savings of tool life was worth the
cost of poor product?
31.11 CONCLUSION
For those who have been thinking that by specifying loose tolerances at the
design stage, the rejection losses of the manufactured goods would be lower,
Taguchi philosophy and the explanation is an eye-opener and made them realize
Quality Loss Function Chapter | 31 445
that in the long run, the total losses, especially the losses due to the customer
dissatisfaction, would be higher. Hence, precision plays a significant role in
reducing the total losses, as explained in this chapter.
FURTHER READING
The following websites can be referred to for further information and addition-
ally to some text books indicated in the bibliography.
[1] en.wikipedia.org/wiki/Taguchi_methods.
[2] http://www.businessdictionary.com/definition/Taguchi-loss-function.html#ixzz37PnlNpIf.
[3] http://www.scribd.com/doc/73104345/Taguchi-Loss-Function.
[4] www.terninko.com/loss.htm.
[5] It provides an animated graph showing how the loss reduces when the kurtosis increases—
http://elsmar.com/Taguchi.html.
[6] www.annauniversity.info/NetLearn/TQM_TAGUCHI.ppt.
[7] https://www.kellogg.northwestern.edu/faculty/dranove/htm/dranove/coursepages/Mgmt%20
469/noisy-variables.pdf.