Robust Design and Taguchi Method
The Taguchi Method was named after Dr. Genichi Taguchi and is also labeled as the Robust Design technique. Dr. Genichi Taguchi was the man who pioneered the design after World War II ended and that has developed over the years.
Unlike the Six Sigma method which aims to reduce waste in the manufacturing and during the operations phase mainly focuses on:
1) Increasing engineering productivity to quickly develop new products at a low rate, and
2) Management which is based on value; the Robust Design centers on improving engineering productivity.
This Robust Design is focused on the improvement of the significant function or role of a process or a product itself, therefore smoothing the progress of strategies and coexisting engineering. The Taguchi (Robust Design) approach rooted on a so called Energy Transformation method for engineering systems like electrical, chemical, mechanical and the like. It is a unique method which makes use of the ideal function of a process or product in contrast to the conventional approaches which mainly concentrate on “symptom analysis” as a source for development or improvement towards the achievement of Robustness and Quality Assurance.
To ensure or guarantee customer satisfaction, the Robust Design approach takes into account both
1) The noise considered as the variation from environmental to manufacturing and component failure, and
2) The cost considered as the rate of deterioration in the area. It is a technique for performing experiments to look into processes or investigate on processes where the end result depends on several factors such as inputs and variables without having a mind-numbing and inefficient or too costly operation with the use of possible and feasible mixture of values of the said variables. With a systematic choice of variable combination, dividing their individual effects is possible.
The Robust Design method is an exclusive alternative for DOE or Design of Experiments which differentiates itself from the traditional Design of Experiments focusing on the most favorable design parameters to reduce discrepancy prior to attaining the average values on output parameters. This innovative design to engineering signifies the most important leap in process and product method ever since the beginning of the Quality revolution. The Robust Design method or the Taguchi approach makes it possible for engineers to:
- Improve processes and products which are intended under a broad variety of consumer’s circumstances in their life cycle and making processes reliable and products durable
- Capitalize and get the most out of robustness by developing the planned function of a product by improving and expanding insensitivity to factors of noise which somehow discredit performance
- Alter and develop formulas and processes of a product to arrive at the performance desired at a reduced cost or the lowest rate possible but, at the shortest turnaround or time frame
- Make designs easier and processes at a reduced cost
Over the years, Six Sigma has made it possible to reduce cost by uncovering problems which occur during manufacturing and resolving instant causes in the life cycle of a product. Robust Design on the other hand has made it feasible to prevent issues or problems by rigorously developing designs for both manufacturing process and product. The Robust Design follows a crucial methodology to ensure a systematic process to attain a good output. Below are the 5 primary tools used in the Robust Design approach:
- The P-Diagram. This is used to categorize variables into noise, signal or the input, response or the output, and control factors related with a product.
- The Ideal function. This Ideal Function is utilized to statistically or mathematically identify the ideal or ultimate outline of the signal-response association as represented by the design idea for developing the higher-level system work fault free.
- Quadratic Loss Function. This is also termed the Quality Loss Function and is used to measure the loss earned or acquired by the consumer or user from the intended performance due to a deviation from it.
- Signal-to-Noise Ratio. This is used to predict the quality of the field by going through systematic laboratory tests or experiments.
- Orthogonal Arrays. These are used to collect and gather reliable information about control factors which are considered the design parameters with minimal number of tests and experiments.
The following are the 4 main steps in Robust Parameter method:
- Problem Formulation. This step would incorporate the identification of the main function, development of the P-diagram, classifying the best function and signal to noise or S/N ratio, and planning or strategizing the experiments. The tests or experiments would involve altering the noise, control as well as the signal factor logically and efficiently utilizing orthogonal arrays.
- Gathering of Data. This is the stage where experiments or tests are performed in either simulation or hardware. Having a full-scale example of the product for experimentation purposes is not considered necessary or compulsory in this step. What’s important or significant in this stage is to have a vital model or example of the product which satisfactorily encapsulates the design idea or concept. As a result, experiments or tests can be performed at a low cost or economically.
- Factor Effects Analysis. This is the stage where results or outcome of the control factors are estimated and such results are evaluated to identify and classify the most favorable arrangement of the control variables or factors.
- Prediction/Confirmation. This is the stage wherein predicting the performance or operation of the product model under the most favorable arrangement of the control variables or factors to confirm best conditions is done. After which, experiments are done under such conditions as well as comparing the results observed with the underlying predictions. If the outcome or results of the experiments done corresponds with the predicted results or predictions, final results are then implemented. However, if predictions do not match with the final results, the steps need to be repeated.
A lot of companies worldwide have saved millions of dollars or even hundreds of millions just by using the Taguchi approach. Telecommunications, software, electronics, xerography, automobiles and other engineering fields are just some of the few businesses which have already practiced the Robust Design method. With the Robust approach, rapid achievement to the full technological capability of designs and higher profit can be considered consistent.