iSixSigma

Thomas Whitney

Forum Replies Created

Forum Replies Created

Viewing 26 posts - 1 through 26 (of 26 total)
  • Author
    Posts
  • #192944

    Thomas Whitney
    Member

    Ok everybody, I must say I love the discussion on all this statistical philosophy. But, I worked in the insurance field for a long time and understand that the management in these fields, most transactional processes, are data phobic. Hence, my original comment that doing an average chart is one dimensional. Most of the management in the transactional field do not understand nor do they want to understand variation or control limits. They need simple pictures. That a box plot outlier to you stat heads may not be statistically significant does not mean that the data point is not different than the rest of the data in the sample. Processing insurance claims or processing mortgage documents are the same. Where one of the processor people or any other variable do something different from the rest, an outlier, is important just as which form is used is important.

    My big success in claims processing worked as follows. I was doing a group FMEA on what factors made one processing center so much worse in closing insurance policies than another? A lady in the midst of the FMEA started crying. She said that the poor performance was all her fault. She did not know that the type of insurance policies gained by the business had to be responded to in less than 24 hours. She as the clerk in the front office let insurance requests sit for days, sometimes, because she didn’t know the importance of getting request to the adjusters fast. Trish, you may have “by” variables as strange as this!

    Transactional processes are different than engineering processes in as simple of issues I just described, so I admonish myself for not presenting the following in the first place; a tenant of problem solving I hold as the most important.

    @Trish, What question are you trying to answer or what question is your management trying to answer? Define it precisely and us not so much stat heads and us stat heads can help you determine what tool to use.

    Just a thought..

    0
    #192927

    Thomas Whitney
    Member

    @garyacone @MBBinWI @Mike-Carnell My final statement on the white paper that is almost complete on the subject of sustaining the gains of a LSSL deployment.

    What Lean Six Sigma Deployments Do Not Address

    Too often companies deploy a Lean Six Sigma effort with a mind that the Lean Six Sigma deployment will address all of the quality systems holes that exist in the organization. This is not true. There is no one silver bullet to quality. LSS targets the problems needing more sophisticated forms of analysis and corrective action. It is the only problem solving system that emphasizes process management over quality problems management. It does not address:

    1. Basic data collection and data mining to find systematic procedural holes for managing quality in a manufacturing system.
    2. Sampling plans for protection from poor quality incoming materials or sub-assembly processing.
    3. Final quality audit, customer complaint or warranty management.
    4. A good version of the eight disciplines of quality improvement to solve discrete quality issues outside of systematic process issues. Not all quality issues need a full DMAIC corrective action process but do need disciplined problem solving.
    5. The seven basic tools of quality improvement for managing less complex quality problems. While these tools are embedded in LSS, they need to be separated out and tied specifically to the eight disciplines of problem solving method of quality improvement.
    6. Quality policies. Quality policies need to be in place before any basic quality or LSS activity begins. A quality policy dictates the: who, what, when, and where quality will proceed specifically including a rhythm of review and accountability delineation for items such as action registers.

    Lean Six Sigma came out of the need to provide more sophisticated tools for the more complex problems found in processes. It also switched a 100% focus on solving the same quality problems over and over again to one that also included working on the process root causes of the perpetual problems and solving those issues. It was never meant to be a replacement for an all-encompassing quality system that was able to address all types of quality threats whether process or discrete. Perhaps it will do well that companies take a few steps backwards from just LSS and readdress just the basics of a quality system too!

    Has any of you kept an old quality policy. I keep throwing away stuff I haven’t used in years so haven’t got an old copy. My mind is too like that of Jerry Garcia in his latter years to be able to reconstruct stuff without extreme effort. It is more senility than drugs though. I think the next paper may need to be on what a good quality policy looks like. I could probably manage to reconstruct one but I am also lazy, so if you have one all the better.

    0
    #192918

    Thomas Whitney
    Member

    In my view here is what has happened to my philosophy of quality.
    Six Sigma tools when I learned it in 1983 were the tools in a course called “Advanced Diagnostic Tools for Engineers and Scientist”. I was not able to take this training until I learned basic quality tools and quality management. Because I had a strong foundation in basic quality tools, the Six Sigma advanced tools made sense to help me solve the more complex issues I encountered.
    To the Point:
    Where is basic data collection and data mining taught in SS. Where is Sampling Plans taught in any Six Sigma Curriculum? Where is a good version of the eight disciplines of quality improvement taught? Where is it taught that Control Charts are more important in the measure phase than in the control phase? Where is it that the seven basic tools of quality are taught independently as the first set of tools to be used in assessing variation? Why is the DMAIC process taught to be 5 demarcated steps in problem solving that need a review between each step? I still don’t know how there is not only an MA-IC sequence. How do you separate Measure from Analyze? I can’t help but to analyze data as I record it. My biggest issue in SS training; Control plans. In SS problem solving control plans are the cornerstone of maintaining the gains and yet to date I have never reviewed a good control plan coming from an SS project. Control plans are only second to a good quality policy document. Quality Policy you say? What is that. My point exactly, it is a lost concept. Quality policies need to be in place before any SS activity begins. A quality policy dictates the who, what, when, and where quality will proceed. It is the cornerstone of accountability. Yes, Mike C, left unto themselves mid level management will be corrupt. An enforced quality policy leaves them not unto themselves. Mike, be a little kinder to mid management. Most are basically good but maybe ignorant (ignorant not meaning stupid but ill-informed) of what to do. Once told what and how to do, say quality, they will do it.
    We forget that most of our BBs are ignorant of anything related to quality and then we teach them only advanced diagnostic tools are the way.

    Where do we go from here. Let us go backwards to just good data mining training, basic quality tools training, good quality policies training and good control plans training

    Not a rant or an opinion but a serious issue!

    0
    #192916

    Thomas Whitney
    Member

    @Mike-Carnell I accept the apology, but be assured I was not offended. It is our gift and our profession to detect when something is said that does not fit the data we have at hand. Never thought of the pink X thing before but maybe that why this is my favorite Shainin X?!

    0
    #192909

    Thomas Whitney
    Member

    @Mike-Carnell I don’t know. I watched the Juran 12 set tapes we had in Seguin 3 times and all the Harry tapes (don’t remember how many) on stats several times. Juran was tough, I think he was 90 when he made them. Harry was young with brown hair and was OK. Just the thought of watching a John tape makes me want to laser myself. I love John but he is the most monotone slow speaking person I know based on how he spoke during the must have been 50 quality surveys on which I went with him. I think I would try if I drank a dozen of John’s favorite drink, the Manhattan. No, make that 15…

    0
    #192908

    Thomas Whitney
    Member

    @Mike-Carnell Mike, it may seem BS to you but I do treat my wife this well. $132.58 for 3 lobsters (there is 3 people) and I cook, them and crack them open for her so she just gets the lobster and I make all the trimmings.
    I’ve never been home enough not to treat her well when I am. In 31 yrs of marriage, my home time has been about 10 yrs.

    0
    #192901

    Thomas Whitney
    Member

    It’s Jung, I spelled it wrong in the original. I know who would believe.

    0
    #192900

    Thomas Whitney
    Member

    I left out John Lupienski and Steve Zinkraff above. To me John was the definitive master of process assessment and it was great learning from him. I hired Steve into Seguin because all I knew were the 7 basic tools and knew we needed allot more sophisticated tools for the problems we had. Steve was a Navy Reserves friend of mine who was also an applied statistician.
    He brought the discipline of MA-IC to the plant as an analysis methodology. We had brown bag lunches every Friday where he reviewed our projects. His MA-IC process became MAIC under Harry and then DMAIC at GE.

    0
    #192898

    Thomas Whitney
    Member

    @Mike-Carnell Actually she is making me make her a full lobster dinner tonight to prove the point of making her happy.
    My laser doesn’t work, so you are safe. Canadians and Cubans are OK anyway if you don’t talk on your cell phone while driving.

    0
    #192897

    Thomas Whitney
    Member

    @Mike-Carnell Actually Texas is quit Jung in the counseling. The whole gun and laser thing is just a face persona thing. I’m really quite passive but not Young.

    0
    #192893

    Thomas Whitney
    Member

    @PraveenGupta @KatieBarry Katie, I appreciate your comments on etiquette. I like it when Gary or Mike or that Darth Vader guy on this site kicks my but with strong and slanderous words when I say something not to their liking. It is the only way they can get through to a strong willed person like me. I have a Masters Degree in Psychology and went to a school that taught all forms of counseling. I learned that if you are working with a New York client use Adler techniques (a get in your face technique) because that works best with them; If you are working with a mid-westerner use a Young technique (very nice and attentive) because that works best with them. However, no Psychologist is good at all techniques and should and do only use one. If censor is the game who do you censor, Adler or Young. If you talk about respect, then respect the psychologist technique to whom you dare to reveal your thoughts. Do not read our inputs if they offend you communication style.

    I work at the tactical level of LSS deployment for a reason. All this talk of high level methods and which school of process change is right, or who is the father of six sigma or no bores me.There is only one thing that matters. Is the tactical level six sigma change agent able to create change?

    An LSS change agent needs to keep a couple of principles in mind every second of their interface with anyone in a process. Your skill only allows you to work “on” a process not “in” a process. We are change agents and only change agents. But, the people in the process at the first interaction with you will assume you think you are an expert at working in the process. They think that you think that even though they live 25% to 50% of their life in that process, you think you know what they need to change to make it better. Our job is only to take what they know or believe is true, convert it to data and assess whether what they know is true or think is true is true. The assess part is the catch. Just taking data is statistics. Assessing data is definitely statistics.

    Examples are good. In my last assignment there was a problem with achieving the proper thickness of a cracker. Thickness is a very important parameter for the cracker. There was not one person in the entire bakery that did not absolutely state that wet dough weight was the primary factor in determining thickness. All efforts in the past and now were placed on dough weight variation with no success. I quietly in the background took data and did a regression on dough weight to thickness. The results showed there was no relationship! I assembled all the subject matter experts and operators in a room and showed the results. A riot erupted with most stating strongly that my data was obviously wrong. Then, a very quiet and unassuming R&D person raised his hand a stated, “It is the meal content of the cracker that effects the thickness the greatest. And depending on the flour used there is an interaction”. Making a long story short we used “statistics” to get the meal and flour ratios correct. Problem solved.

    I think I will get into this “Father of Six Sigma” stuff. What I just stated above about our role as change agents and how to do it, I learned form Bill Smith, Marty Rayl, Gary Cone and Mike Carnel. Marty and Rick were best friends and collaborated allot as to what six sigma should look like and Marty even bounced stuff off of people like Gary, Mike and even me. The work being done in the automotive sector was equivalent to an R&D effort for six sigma because the automotive sector was years ahead in implementing statistical control at the demand of our customers. We started this stuff in 1983. So the date of 1987 is wrong; Six sigma started in Motorola in 1983 in the automotive sector. It expanded company wide in 1987. Marty had Rick interview me for a position in the Com division after my “training period” was over. Rick was brutal but at the end of the interview complemented me in introducing some good new techniques in the change management process of six sigma. I think Rick interviewed a whole lot more people than just me. So who is the father of six sigma? I don’t think Rick would accept the title singularly.

    As a consultant in the field you are only as good as your customer says you are. At the tactical level of the Fortune 50 company Gary spoke of the operations VPs all tell me they have heard of me and want me to go work in their divisions. Not a boast a fact. Why? Because after my first engagement in the company the OVP was impressed. He had never had a BUM and a Plant manager tell him that he needed to meet this consultant who can turn any argument into numbers and then turn those numbers into improvement actions.

    Enough said…

    0
    #192875

    Thomas Whitney
    Member

    Well, I do not consider myself an expert at anything but I think I might be running the risk that people will say I am an expert at having my wife say I am a nice guy and that we are soul mates in our relationship. Anyone who knows me in the business world would not say I am a nice guy nor very good at developing business relationships. I don’t know what it is I do so often to make my wife say such things, but she does. I do know this, I am extremely passionate about being her soul mate and making her happy. I do not plan how to do this, nor do I read “How To Make Your Relationship Better” books.

    I am passionate about the profession I have chosen. I am told I am very good at what I do and some comments lead me to believe some people think I am an “expert” in this field. I have never planned how to be an expert in my field nor how to be better at my profession or more knowledgeable about my profession, nor do I read books on the subject of my profession.

    The term “expert” is a title. If you want the title as an expert, you will never be an expert. If you are totally passionate about what you do and do what you do because of who you are and the passion you possess for what you do, you run the risk of people titling you as an expert.

    Not a rant but a philosophical musing…..

    0
    #192872

    Thomas Whitney
    Member

    @MBBinWI Fortunately, i take medication for my road rage thinking….

    0
    #192869

    Thomas Whitney
    Member

    @Mike-Carnell I didn’t say the assessment process sucked, I said that how the projects coming out of the assessment process were managed sucked.

    When I was studying laser physics and laser design in college I wrote a proposal for a laser design that could be installed on the front of a car. When an ass-hole driver was in front of you, you could evaporate the entire car or at least evaporate the brain of the dumb-ass driver. I think I want to pull out my safely hidden design (the professor took the original design and burned it, he was nice forgiving person) and build it. Since it is against the law to talk on cell phones while driving and Texas blue book law allows the shooting of any Damn-Yankee or in self defense without charges, we just need to evaporate any person on a cell phone in self defense or that is not an obvious Texan and therefore a Damn-Yankee. This should eliminate at least 60% of the cars on Houston highways and make traffic patterns free and clear. Perhaps other States should consider how to use this great device within their legal system.

    0
    #192865

    Thomas Whitney
    Member

    @Mike-Carnell I resent your totally outrages claim that senior managers have no integrity. When I was VP at that big retailer chain all of my senior management buddies had total integrity. We did absolutely no work, set no budgets or goals, and managed no budgets. We did this with total integrity to this ascribed work ethic. The lower level guys spent what they needed to ensure they got the greatest good deals from the supplier bribe, I mean negotiation, trips.
    The CEO determined the bonus plan (which didn’t really rely on performance) and managed stock holder bitching if numbers could not be manipulated enough by law to make all things seem ok; if it happened some big storm or banking problem caused consumer confidence to fall and that was the root cause of poor stock performance on some quarter. The stock would rise again line always worked
    Now this might suggest that lack of integrity might possibly exist at the executive level which is not technically the senior management level.
    As far as store level people we didn’t pay them enough to actually need to lower pay scales or have lay-offs no matter what cash problems existed. So they were never technically screwed either.

    0
    #192840

    Thomas Whitney
    Member

    @dushyantthatte Q1. The top level for a factory metric is always the output measure. Any business is in business to make stuff. And all factories are built to a capacity to make enough of the stuff to make money. Someone sets that amount and the factory needs to build it. So, the measure is always how much stuff your expected to make per day broken down to units per hour. The success rate to the factory of achieving the units per hour is the measure or yield (Six Sigma). Counter measures are the throughput (TOC) and waste (Lean).

    Q2. I only do CI this way and I’ve done it this way for 30 yrs. and I always get incredible results results.
    Q3. NO.

    0
    #192833

    Thomas Whitney
    Member

    @Stevo I’m going to try and not go into another long rant, which I would if I didn’t need to type it out but let me change your words a little. The line people are ignored because most exempts are too arrogant to listen to them or think it is below themselves to go mingle with them.
    There is not one, not one major breakthrough I have been a part of that wasn’t a result of the Red X input from a line operator/associate. When the machine tech of a solder machine told me the red X as I was taking data in my discovery search for the number one defect issue in the plant, I asked him why he never told anyone else. He said he had been saying it for 10 years but no engineer would listen! Motorola spent hundreds of thousands of dollars if not millions on trying to solve this problem and the Red X rested in the head of a lowly machine tech for 10 years! I was called a hero and tried to explain it wasn’t me it was the machine tech but I still got the credit! Even more arrogance in the face of truth. I better stop now…..

    0
    #192832

    Thomas Whitney
    Member

    @Mike-Carnell I didn’t keep my control charts I did by hand, but I also remember being able to look at a list of numbers and tell you the average and standard deviation cause I hand calculated these values so often. I bought a little machine that I could type in the numbers and it printed out control charts about 3 inches wide like a ticker tape machine. Man what a productivity saver.

    @MBBinWI I was the first person to get that first Compaq “luggable” and was ecstatic at all the memory it had! And it didn’t seem that heavy to carry around. I remember when you had to write the command code for Minitab.

    And I have a green, red and pink shirt I wear when I do projects now. I can feel which color of X I am going to find.

    However, regardless of from whence six sigma originated (though I suspect it did have something to do with beer) there is no question in my mind that I wrote and have the original and best Six Sigma Training Material……Gary KNOWS this to be true

    0
    #192826

    Thomas Whitney
    Member

    @cseider Nice to hear from you again. Thanks for the compliment, but I couldn’t have done so much if the CI team whern’t one of the best with whomI have ever worked.

    0
    #192825

    Thomas Whitney
    Member

    For those that have read the GOAL, MBBinWI reminded me of my one line summation of the book, ” Where is Herbie now?”
    I absolutely agree with MBBinWI summation of how all the tools and methods fit together. I tend to encourage the total package of CI project deployment to follow the general sequence : Stability, Shape and Throughput (constraint elimination} to get the fastest ROI.

    Stability: Highly unstable process, those that have enormous swings in KPOs, don’t need allot of measuring or observation to find projects rich in returns. The instability makes it rather futile in the short term quest for high rates of return to put too much effort on Shape or Throughput projects especially when resources are constrained. I personally love leading efforts in highly unstable processes. It is easy to become a hero!

    Shape: Once a process is basically stabilized process variation reduction and lean activity (LSS) become the easiest target projects for high ROI. These projects need to be systematically linked to the business financial/customer requirement metrics as I talked of above. I am about ready to move on after this stage, the easy variation/lean work is done and returns become less. Too much work and not a high enough chance to be a hero!

    Once Shape projects have gone through a few iterations, then throughput constraints become easier to see and the next highest ROI effort. TOC is a great way to manage these. Now sometimes throughput projects are pretty fun and some do have hero level returns but often they involve too much begging for capital and involve the nasty business of hard savings through personnel reduction. Only the most sophisticated managers understand the notion Gary mentioned in his interview, that people should be moved to new opportunity areas rather than be eliminated. Motorola did this well and it is probably why the company is still union free! I am out of here by the first quarter of this phase. Way too much work and social distress!

    Actually, I’m really just a manic fanatic for CI and love the craziness of the stability and early shape phases. The rest is too calm for me.

    0
    #192804

    Thomas Whitney
    Member

    I am suggesting that working projects independent of each other or with respect to their affect on the significant variation of KPO variables is not an efficient way to fix the whole business system. Any given successful project will fix a part of the business system, but it may not lead to a breakthrough improvement in the KPO variable and may have a very small financial return during the deployment stage.

    Case-in-point (in simplified terms). I recently worked in a Bakery where the KPO was product giveaway in the final package of a cracker product. The process delivered 15% more weight to each final package than was necessary. It was staunchly believed that the critical variable was wet dough weight and a project was launched. As this process was improved to a Cpm of 1 and then 1.5 the final package weight remained near 15% overpack. During the entire wet weight project no study of the variation in the four sub packages which made up the final package was made. When the variation of the sub packages was measured it was found that there was significant variation in the sub packages. This led the wet weight team to look at the variation in rows of dry (after baking) weight crackers that made up the each of the four sub packages. To their astonishment, there was virtually no significant variation across the rows. Since the dry weight of the cracker was the customer KPO variable and it had a good Cpm it did’t occur to anyone to look at the next process step for the cracker which delivered a thin layer of oil to the product. This part of the process was considered a no brainer operation. The oil process was found to have significant variation across the rows and accounted for an 8% reduction in the final package overweight.

    0
    #192796

    Thomas Whitney
    Member

    I just happen to be writing a white paper on this subject so I will give you my observations on the subject of why LSS deployments so often have issues to failures. Hopefully, what to do different part of the white paper will be out soon.

    The most common mistake made in Lean Six Sigma (LSS) deployments is after the assessment process most of the training and project focus is directed towards possible significant Xs’ (or customer requirement) identified that are expected to affect the top-level financials, customer requirements or yield. Each black belt or green belt is assigned an X (or customer requirement) as a project and are expected to form a team and optimize the X (or customer requirement) with the X (or customer requirement) being the primary measurement variable. All projects are treated as independent from the rest and the black belt/green belt must find the time to complete the LSS process! In too many cases, the key output variable (Y) is reviewed in “private” by plant staff’s especially product output measures, cost variances and labor costs. Finance people often give swags at what impact they believe the optimization of the X or customer requirement will return. In the worst cases, the rule that states each black belt project will generate a 68% improvement and return $250,000 is invoked. Plant management LSS reviews consist more of technical LSS process reviews than on whether the project is championed correctly, supported correctly or has any expectation of a breakthrough in a key output variable (Y).

    Just my ranting….

    0
    #192710

    Thomas Whitney
    Member

    Why use such a one dimensional measure such as the average. Plot the data using a box plot by day. So much more information is available and the plot is easy to explain and see things like outliers and dramatic differences in the spread and median of data.

    A note on data collection. I hope you are collecting as many “by” variables as you can such as by office location, person, type of mortgage, city etc. I would love to find out to whom the good outlier data belongs and to whom the bad outlier belongs and find out what is different, That’s Six Sigma Heaven!

    0
    #192702

    Thomas Whitney
    Member

    Oh yea, and pink X’s too. In fact, instead of lean I like JIT and all kinds of other whittling spoke wheel terms!

    0
    #192696

    Thomas Whitney
    Member

    If the question you are trying to answer is not “How capable is my process?” , then capability measures are not needed.
    If your inspection methods answer the question “Is my customer protected from defects?” then they are sufficient.

    0
    #192694

    Thomas Whitney
    Member

    All capability analysis comes down to What Question are you trying to answer. If I need a target more than anything use Cpm, if I just need to stay within spec use Cpk. But, what question are any capability values trying to answer?

    Do I need to incur the cost of inspection seems a good one. How many parts pass this go-no-go test sequentially. If it has been 10,000 stop the test!

    What is the customer tolerance for defect escapes. No it is not zero. But if you believe your process can be under the tolerance; Stop the Test!

    In my early days as a quality engineer we stopped 90% of the testing in our factory because we could demonstrate with data our capability was sufficient enough to drop the cost of testing.

    Just my ranting…

    0
Viewing 26 posts - 1 through 26 (of 26 total)