Hobbies And Interests
Taguchi Method
The Taguchi Method is intended to improve the quality of any product during the design stage, reducing the need to fix a product that was already designed. It is based on the idea of designing experiments carefully to reduce the number of aspects that need to be tested. If a designer attempted to test every possible variable in a design with eight parameters and three levels, a total of 6561 experiments would have to be completed. Using the Taguchi Method, only 18 experiments would have to be done, which is an advantage.
L4 Arrays
In an experiment to determine the best way to bake a cake, the first task would be to define what constitutes a successful cake. If the options to be tested included a chocolate cake, a coconut cake and a vanilla cake, then they would be the three parameters. If the cake could be baked at either 350 F or 400 F, they would be the two levels. According to the Taguchi Method orthogonal array selector, an L4 array would be required. The first experiment would compare all three cakes at the same temperature. In the second experiment, the first cake would be baked at the lower temperature and the other two cakes at the higher temperature. In the third experiment, the second cake would be baked at the lower temperature. In the fourth experiment, the first two cakes would be baked at the higher temperature. Then the results would be compared.
Advantages
The advantages of the L4 orthogonal array are the same as those of the Taguchi Method. In general, fewer experiments need to be performed in order to determine the effect of changing variables and to determine the best way to improve quality during the design process. Variables with a large experimental effect on design quality can receive the most focus as the design process continues, while variables shown to have a minor effect can be treated as unimportant by comparison.
Limitations
Just as with the advantages of the L4 array, its limitations are those of the Taguchi Method as a whole. While the relative significance of variables can be quickly determined using this method, it does not account for the ways in which the variables affect each other. If those factors are an important part of a particular design, then the Taguchi Method cannot be used effectively. In addition, the Taguchi Method is most useful when it is applied at the beginning of a project. When a particular approach has been determined, the Taguchi Method orthogonal arrays, such as the L4 array, are no longer as useful.