January 31, 2008

Six Sigma—Taguchi DOE

Today we do another game-- Taguchi DOE, I just look the website and find some article talk about this experiment. Just intersting this experiment and find something in the website.

Taguchi or DOE?
Most engineers have heard of Design of Experiments (DOE) and of Taguchi Methods, but how many of us can really say we understand the difference between them?  Or that we can correctly decide when to use which technique?  In this article, we will examine the relative strengths and weaknesses of each approach, and develop some guidelines for selecting the best approach for solving our specific problems.

First of all, what are these techniques?  In a general sense, they can both be thought of as techniques for optimizing some process which has controllable inputs and measurable outputs.  In a manufacturing situation, the inputs might be settings in some production process, such as temperature of a heat treatment furnace or speeds and feeds on a milling machine.  The outputs are generally quality or productivity oriented, such as process yield or units produced per hour by our production line.  We might be trying to maximize some output (as in throughput or process yield) or minimize some other output (as in failure rate or scrap).

In a design situation, the inputs might be design decisions, and the outputs would then be performance oriented metrics.  For example, inputs might be the number of supports in a structural design, the type of material to be used, or a qualitative decision such as a drum clutch vs. a face clutch.  In these cases, outputs to be measured might be load carried, torque transferred, etc.  In either type of analysis, production scenario or design situation, we are making decisions on how to do something that will affect what we get as an output.

What do DOE and Taguchi have in common?  Besides the inputs and outputs described above, they both deal with multiple inputs.  That is, we might have two, three, five, or a dozen or more input decisions to make, all affecting some measurable output.  It would be nice if we could experiment with these inputs one at a time, optimizing our output for each input in turn, until we've selected ideal values for all input parameters.  Unfortunately, this doesn't usually work, because the inputs generally interact with each other to some extent.  For example, imagine that you are running a carburization process.  You cannot set furnace temperature to optimize yield while ignoring carbon potential, and then experiment with carbon potentials after fixing the temperature; they interact with each other and affect the levels of output together.  What DOE and Taguchi primarily have in common, then, is that they deal with multiple inputs and how they interact with each other.

How do DOE and Taguchi differ?  We will get into this soon, but the primary difference lies in how they handle the interactions between inputs.  When you remember that DOE was invented by scientists for scientists, and Taguchi methods were invented by engineers for engineers, the differences begin to make sense.  Let's look at them each in turn.


The main thing to know about DOE is that it was developed primarily within the scary world of statistics.  Okay, come back out from under your desk; we won't dwell on that part.  Just remember that the theory behind the technique comes from the classical world of pure math.  Using it, however, requires only that small amount of math that you probably remember from your college days.

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.  Of course this power comes at a price, and that price is lots of experimental runs and lots of calculations.

One of the first applications of DOE was in ancient agricultural sciences.  Early farming experimenters were starting to understand things like irrigation, fertilization, crop rotation, etc.  These are multiple inputs.  You can also see how they interact with each other:  What is the best irrigation method?  Well, that might depend on your fertilization approach.  What is the best fertilization process?  Well, that could easily depend on your irrigation techniques, and what crop you grew in that field last year.  Since any and all inputs could interact with all other inputs, a technique was needed which would model all of these inputs, and how they all relate to each other.  Thus DOE was born.

Another peculiarity of these early agricultural experiments was that they wanted to get it right the first time.  One experimental run took an entire growing season; a whole year!  You absolutely did not want to collect a year's worth of data, stroke your beard a few times, and then do Phase 2 the next year before you had an answer; your village could starve to death before you were finished.  This leads to another characteristic of the DOE approach: not only are all interactions studied, but they are all studied at the same time in one big round of tests.

These considerations can lead us to an assessment of the strengths and weaknesses of the DOE approach.  The strength is that we can investigate all possible interactions between inputs at the same time; we don't need any innate knowledge of how the process works.  The weakness is that we have no way to make use of any a priori process knowledge that we might happen to have; there is no way to make the experiment more efficient by thinking about how the inputs really do interact.  If you think about it, the strength and the weakness are really the same thing!

Other areas, besides agriculture, where DOE makes perfect sense are any complex sciences with many highly coupled inputs where practitioners have little innate understanding of the fundamental processes involved.  Biology, virology, and meteorology come to mind.  In all of these fields, DOE is a powerful and logical method to optimize process and predict outcomes.

To be fair, not all DOE-based investigations look at all possible interactions.  Those that do are called "full factorial" DOEs.  "Fractional factorial" DOEs can eliminate some interactions, and therefore slim down the amount of work that needs to be done.  But they are still based on the idea of full modeling, and then whittled down to improve efficiency.  The savings are generally fairly meager, such as a factor of two or four, and there is still no way to inject understanding of the fundamental process into the mix.

The details of performing a DOE can be found in many textbooks, and we won't duplicate them here.  But in a nutshell, in a full factorial, all-interaction DOE, tests are performed for all possible combinations of all inputs.  If you have three inputs, each with two possible settings (known as levels), you would need to perform eight tests (that is, two raised to the third power).  You can see how the number of tests can get really large really fast.  Then, the outputs are averaged for all tests where a particular input was set to a particular level, and compared to the average output for all tests where that particular input was set to its other levels.  This comparison gives an insight into the overall affect of that input on the output.  Similar calculations can show the affects of the interactions between the inputs.  After a bit of number crunching, many useful chunks of knowledge can be derived on how the inputs interact and how they affect the process output.

January 30, 2008

Six Sigma

What is six sigma?

Six sigma focuses on establishing world- class business-performance benchmarks and on providing an organizational structure and road- map by which these can be realized. This is achieved mainly on a project- by project team basis, using a workforce trained in performance- enhancement methodology, within a receptive company culture and perpetuating infrastructure.

It is a statistical model, there a diagram can described this relationship.


January 29, 2008


Today we continue to play the game. This morning we divided into two group, one is product group another is logistic group and we discuss about how to improve the product line.  I think six sigma is just like a self- assessment, do it and find the defect, then discuss how to improve it and again.

Leadership PMA review

I got one day delay, finished the leadership. This PMA has a bit difficuit than before, so I was got some problems when I was wrote. But it is really less time to think more. This time still got some problem for the time manage. But I think some I was done better than the PPE PAM. Like I did not cited too much again. I try to wite every words by myself, it is better. Paul said if you can not write by yourself, you must not understand very clear. That is ture.

Anthoer study way  I really need improve is the write process. Always I like to prepare everything and all of pre-knowledge before start to write because I was used to do it when I did my first degree. But I find it is not very useful in the EEE course. Before I was studying the science technology and if I do not know the formula and law how can I do the exercise or solve the problem. But here I think just spend one or two days to look over the theory but not read word by word just know a bit what it is. Then I can analysis the question and make a glancing planning. Most of time I can write and find material, then studying, summary, and wrrite anagin.

Third I find it is usefui write the idea before write. Mind Manage is a very useful software.

I hope next time I got a good plan and step by step to do. Time manage is really important.

January 28, 2008

Running the Improvement Process ( DMAIC)

This Afternoon we do a very interesting game---Running the Improvement Process. But the product line is not very good. Because one person is not good at assembly the plane. He always make a mistake, so we can not continue to product any more. Because it is not finished.

I just think at this circumstance, is it possible to increase or add people check the product at every process. It must be more effecient, and we can enhance the quality of product. And if find the defect product, we need appoint a place to fix or repair the mistake. Maybe send back to the previous process, or appoint a special place to repair the defect only. If like this, the quality must be improved and nobody will not do nothing just waiting for the last unfinished product. Is it right?


I have not finished the PMA of Leadership yet.

January 27, 2008

Never Make an Appointment

There are a lot of people calling on business men who place a high vale on their own time--- a higher value than they place on the time of the business men on whom they call.

January 26, 2008

Six Sigma

Just walk around the website and find a simple to introduce the six sigma

Six Sigma is a set of practices originally developed by Motorola  to systematically improve processes by eliminating defects. A defect is defined as nonconformity of a product or service to its specifications.

While the particulars of the methodology were originally formulated by Bill Smith at Motorola in 1986, Six Sigma was heavily inspired by six preceding decades of quality improvement methodologies such as quality control, TQM, and Zero Defects. Like its predecessors, Six Sigma asserts the following:

  • Continuous efforts to reduce variation in process outputs is key to business success
  • Manufacturing and business processes can be measured, analyzed, improved and controlled
  • Succeeding at achieving sustained quality improvement requires commitment from the entire organization, particularly from top-level management

The term "Six Sigma" refers to the ability of highly capable processes to produce output within specification. In particular, processes that operate with six sigma quality produce at defect levels below 3.4 defects per one million opportunities (DPMO).Six Sigma's implicit goal is to improve all processes to that level of quality or better.

Six Sigma is a registered service mark and trademark of Motorola Inc. Motorola has reported over US$17 billion in savings from Six Sigma as of 2006.

In addition to Motorola, companies that adopted Six Sigma methodologies early on and continue to practice it today include Honeywell International  (previously known as Allied Signal  and General Electric (introduced by Jack Welch.


Look at some good wisdom:

I never felt there was any great risk in starting new ventures. The greader risk was missing an opprtunity.

... an innovative scientist must e recognized as necessary and vital to the future of the organization.

                           -----Robert N.Noyce, "The book of Leadership wisdom", pp 141-142

January 25, 2008

The Six Sigma organization and training

Today we finished two six sigma, one is the variantion reduce and another is the System of Profound Knowledge. I was find the Deming circle and the system of profound knowledge can use in many place. Most moudle can not leave it. I think the six sigma organization is very connect with the Deming is training, I just find a paragraph describle this.

The Six Sigma organization and training

The implementation of Six Sigma in any organization is at first disruptive because it requires not only the buy in of senior management, but an active role of management in project definition and resource allocation. It also requires extensive training of some of the best people in the organization with the understanding that their role will be 100 per cent devoted to deployment of Six Sigma activities. The heart of these activities is projects that have been defined as critical paths or breakthrough goals that affect the bottom line of the organization.
The training required to implement Six Sigma involves everyone in the
organization. The basic training is one day and covers process mapping, and an overview of designed experiments, hypothesis testing, metrics, and process modeling. Green belt training is more extensive, including a week of statistical analysis, SPC, and measurement systems analysis. The black belt training requires about one month of training, including ANOVA, game theory, and multivariate regression. Workers trained in Six Sigma have been compared to internal SWAT teams, forming to tackle a specific problem and then breaking up and reforming once they achieve the desired results. The methodology applied to each problem is taken from a standard methodology, but tailored for the specific project or problem at hand. The methodology does not, fundamentally, focus on making higher quality widgets; it focuses on making the process more robust and less subject to errors. It is then applicable to areas outside of pure manufacturing organizations.

September 2021

Mo Tu We Th Fr Sa Su
Aug |  Today  |
      1 2 3 4 5
6 7 8 9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30         

Search this blog



Blog archive

RSS2.0 Atom
Not signed in
Sign in

Powered by BlogBuilder