Industrial Engineering 361

Katie Maurer

Scott Lau

February 11, 2000

What is robust product design?

Robust product design is a concept from the teachings of Dr. Genichi Taguchi, a Japanese quality guru. It is defined as reducing variation in a product without eliminating the causes of the variation. In other words, making the product or process insensitive to variation. This variation (sometimes called noise) can come from a variety of factors and can be classified into three main types: internal variation, external variation, and unit to unit variation. Internal variation is due to deterioration such as the wear of a machine, and aging of materials. External variation is from factor relating to environmental conditions such as temperature, humidity and dust. Unit to Unit variation is variations between parts due to variations in material, processes and equipment. (Lochner and Matar, 18). Examples of robust design include umbrella fabric that will not deteriorate when exposed to varying environments (external variation), food products that have long shelf lives (internal variation), and replacement parts that will fit properly (unit to unit variation). The goal of robust design is to come up with a way to make the final product consistent when the process is subject to a variety of "noise".

How do you make a design robust?

Taguchi considers making a design robust in the parameter design portion of product or process design. In parameter design the goal is to find values for controllable settings that minimize the negative effects of the uncontrollable settings. Experiments are used to determine the impact of particular settings on both the controllable and uncontrollable factors. The idea here is that by observing changes in a controllable factor (such as the thickness of boards), a value can be found for that factor that reduces the effect (warping) of something that canít be controlled (the humidity outside). The ultimate goal is to find the optimal settings to minimize cost by minimizing variation.

When setting up these experiments, the factors that effect the product need to be determined. Then the factors can be separated into controllable factors and uncontrollable factors and experiments can be set up to test the effects of changing the values of each factor. There are many ways to set up these experiments. Taguchiís method involves finding correlation between variables. He uses orthogonal arrays, with the inner array consisting of control factors and the outer array consisting of "noise" factors. Each inner array is to be run with each outer array. (If six control factor experiments and three "noise" factor experiments are needed, there will have to be (six times three) eighteen experimental trials to get all the combinations). Another method for conducting these experiments is to make no attempt to control the "noise" factors, but repeatedly run the trials for combinations of control factors. (Lochner and Matar, 152) This type of experiment allows the operator to measure process variability. The trials should be taken in an environment similar to the one in which the actual use or manufacturing of the product is going to take place. A third experimental design is to identify all the control and "noise" factors (adding the control and noise factors yields k) and run an analysis using at least k +1 trials based on eight-run experiments. (You could use an eight run experiment for up to k=7, and a sixteen run experiment for up to k=15.) This will allow the interaction between variable to be seen running fewer tests than using Taguchi's method. Further instruction as to how to use this method is found in chapter four of "Designing for Quality" by Lochner and Matar.

The data found from the experimental trials is then analyzed. The analysis will depend on the method of experimentation. Plot the effect that the variables had on your variation and/or the correlation between factors. Using this data find settings for the controllable factors that are found to lower the variation caused by uncontrollable factors.

Then after the initial experiment trails are run and "optimal" settings are found confirmation experimentation is needed. By performing a series of replica experiments at the levels that were picked, we can see if the values achieved matched that of the values the model predicted. If there is disparity, there may be an interaction or noise that we didnít see and thus our experiment must be redeveloped.

What are the advantages of robust design?

Robust design has many advantages. For one, the effect of robustness on quality is great. Robustness reduces variation in parts by reducing the effects of uncontrollable variation. More consistent parts equals better quality.

Another advantage is that lower quality parts or parts with higher tolerances can be used and a quality product can still be made. This saves the company money, because the less variable the parts can be the more they cost.

A third advantage is that the product will have more appeal to the customer. Customers demand a robust product that won't be as vulnerable to deterioration and can be used in a variety of situations.

This method is also good, because you are designing the robustness into the product and process instead of trying to fix variation problem after they occur.

What are the disadvantages of robust design?

One of the disadvantages of robust design is that to effectively deal with the noise, the designer must be aware of the noise. If there is a noise factor that is affecting the product and the experiments run do not address it (intentionally or not), the only way that the product will be robust to that variation is by luck.

Another disadvantage to robust design done Taguchiís way is that the problem becomes large quickly. If you had a lot of different things to consider as control variables and/or noise variables, it would take a great deal of time to run all the experimental trials. Controlling noise variables is expense, and when lots of trials are required the dollars add up.

Another disadvantage is that by using orthogonal arrays, it assumes the noise factors are independent, which may be helpful in setting up the experiment, but is not necessarily a good assumption (Lochner and Matar, 153).

What are some examples of why robust design is important?

Consider this example adapted from "Creating Quality" by Kolarik; the designers of a radio had built and tested a breadboard. After the radio was considered a success, the specifications were passed to production and the radios began being manufactured. The first production unitís radios went into test and failed to meet marketingís product performance requirements, as did the second unit. Analysis of why the process failed produced no results. They had been following procedure and using standard acceptable parts. Next the breadboard of the original design was inspected. It was found that the designers had hand-tested and picked all the component parts. They worked much better than the manufacturerís standard acceptable parts. After review of the design it was found that there was no way to economically fix the problem without massive redesign, so personnel were assigned the task of manually sorting the components, costing the company additional time.

In this example the design of the radio needed to be robust so that it could handle the amount of variation in the set of standard acceptable parts. Because the design didnít allow for that amount of variability, it cost the company lost time. They had to stop the production process and investigate and then they had to expend further manpower in screening the parts.

Making a product robust is also a concern for companies that manufacture products for an ever-expanding market. If products are sold nation wide or even globally, the differences in the environments, conditions, and uses have to be considered for them to be a success. For example, a manufacturer of a certain type of gas grill that is sold nationally must consider the robustness of the materials used to make the grill. The people in Minnesota may use the grill in the summer only and it is stored in the garage in the winter where the temperature falls to freezing. The consumers in Arizona use the grill year round and it is stored on the deck where it is subject to sunlight, rain and higher temperatures. The manufacturer must make sure that the grill can withstand both conditions. If the freezing temperature cracks the valve connection or if the heat cause the lid to deform, they will lose the potential buyers in the respective area.

What can be said in conclusion?

Robust design is designing a way to make the final product consistent when the process is subject to a variety of "noise". This can be done through a variety of experimentation methods. The results are capable of showing how to develop a product/process that will be robust. The advantages of robust design are that the products are of good quality, cheaper, and more customer friendly than their non-robust counterparts. Although there are disadvantages, having a robust product design can give companies a large competitive edge.



Chaplan, Frank, Quality Engineering, v.11 n.2, Marcel Dekker, 1998-1999.

Chaplan, Frank, Quality Engineering, v.11 n.4, Marcel Dekker, 1999.

Kolarik, William J., Creating Quality, New York: McGraw-Hill, 1995.

Lochner, Robert H., Matar, Joseph E., Designing For Quality, London: Chapman and Hall, 1990.

Madu, Ifeanyi E., "Robust Regression Metamodel for a Maintenance Float Policy" International Journal of Quality and Reliability Management, v.16 n.4,5, 433-456, Dale Barrie, Christian Madu, 1999.

Ueno, Kenzo, Company-Wide Implementations of Robust-Technology Development, New York: Asme Press, 1997.