Shainin, Richard (1992, ASQC) Shainin Consultants, Inc., Vienna, VA 22182
A foundry producing engine blocks was experiencing an unacceptable scrap rate due to castings being out of tolerances. The goal was to identify and correct the source of the error.
Three product teams methodically gathered data from all possible sources of the problem then used Pareto analysis to determine most likely causes. They achieved further reduction by mapping sources of variation to a multivariate control chart that revealed the level of variation coming from each source. Several samples of three consecutive blocks were taken from the line, measured, and the measurements subjected to paired comparisons. The number of probable causes was now small enough to begin design of experiment (DOE) testing on each. The cause was identified and design changes were made to eliminate it.
Other methods that might have been applied are brainstorming, Ishikawa cause-and-effect diagrams, and histograms. The strategy described here allowed efficient reduction of a large number of variables to a number that could be easily tested.
Case study,Brainstorming,Quality tools,Pareto charts,Multivariate control charts,Design of experiments (DOE),Histograms,Ishikawa diagram (Fishbone diagram ),Cause and effect diagram