TAGUCHI DOE (Design of Experiments)

Return to Our Services

DOE/Taguchi - It is a technique to lay out experimental (investigation, studies, survey, tests, etc) plan in most logical, economical, and statistical way. You can potentially benefit from it when you want to determine the: most desirable design of your product, best parameters combination for your process, most robust recipe for your formulation, permanent solution for some of your production problems, most critical validation/durability test condition, most effective survey/data collection plan, etc.

DOE theory starts with the assumption that all inputs might be interacting with all other inputs.  This is a powerful statement.  The technique makes no assumptions about some inputs being independent, and therefore can handle any interactions that might be lurking somewhere in your process.  When you have no idea what interactions you need to be worrying about, DOE might be the choice for you. 


To understand Taguchi methods, its helps to realize that it came from the world of design engineering.  Taguchi methods start with the assumption that we are designing an engineering system, either a product to perform some intended function, or a production process to manufacture some product or item.  Since we are knowledgeable enough to be designing the system in the first place, we generally will have some understanding of the fundamental processes inherent in that system. 

Another distinction in Taguchi methods is the recognition that there are variables that are under our control, and variables that are not under our control.  In Taguchi terms, these are called Control Factors and Noise Factors, respectively.  An early step in a Taguchi analysis is to use our understanding of the fundamental process to try to identify all of the important Control Factors and Noise Factors.  In our heat treatment example above, Control Factors would include temperature, carbon potential, time in the furnace, part layout and orientation, etc.  Noise Factors might include the inherent variability in the furnace temperature distribution, errors in timing, outside air temperatures and humidities, etc.  We would want to ensure that all important Control Factors and Noise Factors are represented in our experiments.

In our full factorial DOE approach, our experimental plan was quite simple:  we tested all possible combinations of input values.  But what do we do in a Taguchi analysis?  We will test a very small subset of all possible combinations, but which combinations do we use?  The answer to this question is one of the most powerful parts of the Taguchi method.  Again, without going into too much detail, a variety of experimental plans have been published for various numbers of Control Factors and levels.  You don't have to design them for yourself; they are readily available in many textbooks.  Some examples include:

In this chart, the L4 array, for example, is an experimental layout designed for three control factors, each at two levels.  This might be two different materials for some new product, two manufacturing methods, and two design variations.  We might be interested in how these factors affect power consumption by our new product.  A full factorial DOE of all combinations of these factors would require eight tests, two to the third power.  The Taguchi L4 layout recommends a subset of four of these eight tests which allow the main effects to be determined.  Similarly, the L27 layout allows thirteen factors, each at three levels, to be tested with a mere 27 tests, rather than the somewhat prohibitive 1,594,323 tests required by a full factorial set of tests.  The "L", by the way, stands for Latin, since most of these designs are based on classical Latin Square layouts, appropriately pruned for efficiency.

Another major difference between DOE and Taguchi is the use of our Noise Factors.  In a traditional DOE, each combination of inputs is tested once.  In a Taguchi test, each combination which is tested (remember, we are only testing certain combinations) is tested several times.  Each of these replications is different, however, in that we use different levels of our Noise Factors.  For example, we might test a particular set of inputs at high temperature and high humidity, high temperature and low humidity, low temperature and high humidity, and low temperature and low humidity.  Even though we don't really control temperature and humidity in a production setting, we want to vary it in our tests to inject maximum variability into our experimental procedures.  This lets us determine not only which combinations of inputs give us maximum output, but which gives us the most repeatable output, which is often even more important.  Taguchi calls this quality robustness, or insensitivity to noise.

Today's highly competitive environment leaves no time for trial and error. Taguchi DOE is a proven way to identify and understand how process factors affect output. With this information, you can develop the optimum balance between factors and achieve major improvements in quality, cost and productivity. This training program emphasizes application over theory and statistical calculations. It is a practical way to conduct and analyze true knowledge-generating experiments using DOE software, hands-on teaching aids and real-world examples.



Unit 1: Foundations
Overview of the steps of designed experimentation:
- Planning
- Conducting
- Analyzing
Objectives of designed experiments
Terms and concepts
Benefits of using designed experiments
Guidelines for successful experimentation

Unit 2: Design
Characteristics of a good design
Best order and manner in which to run trials
Features, advantages, disadvantages of design types:
- Design matrix considerations: orthogonality, balance, aliasing, resolution, interactions, blocking, replication, repetition, randomization
- 2-level designs: full factorials, fractional factorials, Plackett-Burman
- 3-level designs: full factorials, fractional factorials, Box-Behnken, Box-Wilson (central composite designs)

Unit 3: Analysis
How to take a practical approach to interpreting results
Simple graphical analysis
Interpreting ANOVA
Interpreting multiple regression analysis

Unit 4: Application
Practical advice on implementing DOE
Explanations of actual experiments and their results
Recommendations for how to ensure successful experiments
Dealing with "failed" experiment

Return to Our Services


News RSSFeedSmall
Get the latest QSG News delivered to your inbox with QSG's RSS Feed

Sites that offer useful information and resources in relevant industries

Site Map
Can't find what you are looking for? Perhaps our
Site Map
can help


Copyright 2014 © Quality Support Group, Inc. All rights reserved.