iSixSigma

Tips on implementing SPC

Six Sigma – iSixSigma Forums Old Forums General Tips on implementing SPC

Viewing 38 posts - 1 through 38 (of 38 total)
  • Author
    Posts
  • #44643

    Rodrigo
    Member

    Hi all
    Im looking for some tips on how to implement SPC in my company. If you remember working in a company that “made the jump” from no control charts to a company that now has control charts, would you be kind enough to share your learnings?
    Ive been with my company for 9 months now, no control charts in place. The company itself is 84 years old, family owned. I just dont seem to be able to explain why it is important to investigate special causes and the benefits it has. Im quickly approaching the “valley of dispair”..

    0
    #143479

    Rodrigo,
    The purpose of ‘control charts’ is to find sources of variation. Once you’ve identifiedΒ various sources of variation, by changing the way you define subgroups, the next step is to eliminate or reduce the variaition.
    Once you’ve reduced the variation you can stop plotting the control chars – after all you can’t sell control charts. (Ideally, processes should be set up in a ‘robust’ operating area of process space – ‘flat area.’)
    So .. to answer your question, the goal is actually to remove control charts. The problem is many people try to put the cart before the horse.
    My advice would be to change tack and put more emphasis on variation reduction that control charting per se.
    Andy

    0
    #143480

    Jym
    Participant

    Andy U,
    I’m confused by your response.Β  My understanding is that control charts should be used daily/weekly as part of the day-to-day activity of a Process Owner.Β  If you get rid of the control charts after you have reduced variation, how will you see if the process remains in control?
    Jym

    0
    #143481

    Rodrigo
    Member

    thanks Andy for ur post
    I agree with you where “… variation is the root cause of all evil…”.
    The catch is:Β Β  it is more dificult to explain variation without a control chart. With a control chart you can clearly “see” the variation.
    Β 

    0
    #143482

    Jym,
    When we talk about ‘stability’ we generally imply the temporal variation is ‘in control.’ In the sense thatΒ ‘in control’ implies the variation is random and there are no ‘special causes,’ or assignable causes,’ of variation.
    For example, a coating process has a film thickness characterised by a spin curve, where the film thickness varies non-linearly with spin speed. Years ago teams of people used to go around, plotting X-bar and R charts of film thickness, and calibrating acceleration and spin speeds with tacometers. Then, we found we could eliminate all this waste simply by spinning a thicker film and adjusting the film thickness to target using viscosity, whichΒ allowed us toΒ discard the ‘control charts.’
    Since the modified process is now insensitive to process variationΒ  the only concern is a sudden catastrophe, which can be detected very easily by the provision of a tolerance.
    Andy
    Β 

    0
    #143483

    Rodrigo,
    Please note I’m not advocating abandonimg charts .. I’m just trying to emphasise the originalΒ concept of these charts, as recorded by Deming.
    The X-bar and R charts are really ideal tests of homogeneity, which implies they can be used to detect multi-normal distributions, and multivarite distributions, as well as temporal inhomogeneity.
    Andy

    0
    #143498

    Brit
    Participant

    I had a similar experience and found that relating the concept of variation to daily activities (usually in the home or travelling to work) was an easy way to explain.Β  On the other hand….
    Why don’t you pick an operation that you know has assignable variation – there’s got to be one there or you wouldn’t be concerned.Β  Do the control chart by yourself through sampling throughout a few days.Β  Nothing fancy or too time consuming (keepin gin mind the rules for proper charting).Β  Chart the thing and show the ups, downs, outliers and trends.Β  Relate the movement on the chart to money and lost productivity – family owned businesses, like most, don’t like to lose money or pay too much for unproductive time.Β  Make it real to the powers that be.

    0
    #143588

    sathish chandran
    Member

    Hi,
    ” Better late than never”. Its good that you have looked at applying SPC in you company.
    Here are my learnings;
    1) We should start inΒ a small way – choose one are or cell or line
    2) Choose those critical parameters which needs to be monitored.
    3) The operators who need to plot should be trained in the first place.
    4) Plan a session 4 hrs, where you teach them in simple language, what is the importance of control chart and how to plot it. You start with one of these —> I-MR or X bar R or X bar S chart.
    5) Plan for 2 hrs of class room training and 2 hrs of shop floor work where they collect sample and plot the chart.
    6) Keep doing this on a regular basis till all the citical parameters are covered.
    7) Once implemented , have a frequency of checking these charts and updating the control limits.
    8) As this progresses train people on how to interpret these trends and take suitable action.
    9) Slowly start replicating this to other areas.
    The point is show / demonstrate people that because of control charts there has been benefits.
    Hope this helps
    sathish chandran

    0
    #143589

    Rodrigo
    Member

    Than you very much Sathish

    0
    #143591

    Rodrigo,
    Are you and Satish one and the same person?
    If not why not ask him to provide a practical examples of a process. Ask him to describe some details about the process. What corrective actions were taken to eliminate a special cause. What was the special cause. How did he solve a common cause of varation. How did this affect rolled yield?
    Are you both just a couple of ‘armchair’ engineers?
    Andy

    0
    #143594

    Rodrigo
    Member

    No and no.(no, im not Satish and no, im not an “armchair engineer”- im not even an engineer and last time Ive checked i didnt look much like a chair.)
    Now, onto more serious stuff, my original post asked for tips on implementing Control charts. That simple.
    Elimination of special causes is a different battle that i am currently fighting, armed with FTA (fault tree analysis), 5Whys, causal factors, CAPAs, to name just a few. DIFFERENT BATTLE from original post, SAME WAR.
    All I wanted was some tips on implementing control charts. Control charts would be beneficial to immediatly spot the special cause and take the appropriate action as opposed to spot the problem “at the end of the line” days or weeks after it actually happen.

    0
    #143604

    Rodrigo,
    The reason I ask is because there is a lot of deception on this website. Some of us have actually achieved something in this field. What I find surprising is others are making recommendations that don’t actually work. YouΒ ought to be able toΒ deduce this yourself quite easily!
    For example, who gives specific examples, and who just reguritate stuff from textbooks written by quality auditors!
    I know what works and doesn’t work – not from books – but from practical experience. Do you think Motorola provided ten years of Juran training, hired several statisticans, and quality managers, yet failed to satisfy Ford and GE prior to 1984? Think about it!!!
    The worse thing ever to happen to USA quality was a small, little man, reporting what he thought should happened in the development of Motorola’s Six Sigam efforts, rather than what actually happened.
    I know from my own personal expeience of over ten years that trying to ‘sell’ X-bar and R charts just doesn’t work. It didn’t work then and it won’t work now!
    No wonder Six Sigma is in decline if Satish’s reply is considered the correct one. Clearly, I can see I’m wasting my time on this forum.
    Goodbye,
    Andy

    0
    #143605

    Orang_Utan
    Participant

    Not a single SPC chart was found in a 20,000 people Sony plant when I was a junior engineer with them.My himble observation is almost 100% failure rate of SPC in the real life. Only those with real time and date automated SPC system shows some positive results.

    0
    #143613

    Rodrigo
    Member

    andy,
    can feel your frustration when reading the post. help me out to understand a couple of things:
    On my previous job (8 years of it) we had Control charts for nearly every process of my department. When I got there, the control charts where already in place. When a special cause was detected, root cause analysis procedures were triggered.
    1- do you agree with the use of control charts to “keep an eye” on the process?
    Β 

    0
    #143614

    Peppe
    Participant

    Rodrigo, there is something strange in your first and last post. First you ask for how to implement SPC, in the last you claim to used SPC for 8 years with good results. Please, could you clarify ?
    Rgs, Peppe

    0
    #143617

    Rodrigo
    Member

    yes, easy. previous job lasted for 8 years untill i moved country again. now, new job in a company that does not use control charts.
    previous job – control charts in place way before I got there.
    new job – no control charts. fire fighting. semi-chaos.

    0
    #143622

    Orang_Utan
    Participant

    You shall give us some tips on how to implement SPC instead of asking how to do it. Seemed you have learned nothing from your so-called past 8 years experience on SPC implementation.

    0
    #143623

    Orang_Utan
    Participant

    Not a single SPC chart was found in a 20,000 people Sony plant when I was a junior engineer with them.My humble observation is almost 100% failure rate of SPC in the real life. Only those with real time and data automated SPC system shows some positive results.

    0
    #143625

    Rodrigo
    Member

    i am sure i have missed something here..

    0
    #143627

    Peppe
    Participant

    Clear picture, but I’m still confused … If you are confident in SPC, as I understand from your posts (wrong ?), which is yourΒ exact question ?Β Your new company seems be the best for every improvement project you wish put in place, they have nothing… SPC efficacy isΒ a deeper matter …
    Rgs, PeppeΒ 

    0
    #143629

    Mayes
    Participant

    It seems that in his previous company Rodrigo used control charts that were already implemented by someone else.Β  He had good results using what someone else started.Β  He is now in a company that doesn’t use control charts and he is asking howΒ he gets started implementing control charts when they have never been used in the new company.

    0
    #143630

    Sayyed
    Participant

    1. Identify CTQs – essential parameters in your product/service.
    2.Β Conduct fishbone analysis to identify causal X’s that you & your team think could have an impact on your CTQs. (Different fishbones for different CTQs)
    3. Collect data on all the Xs & your CTQs.
    4. Conduct statistical tests to understand if significant relationship exists between your X’s & CTQs.
    5. Wherever a relationship does exist, those X’s become you CTPs. Now put a measurement system around those Xs (if you do not have it already) & plot performance on an appropriate control chart.
    ——————
    This is what I think can be done. However wait for more responses,esp from the more knowledgeable folks in this forum who often have brilliant insights. All the Best – Neal.

    0
    #143635

    Peppe
    Participant

    My understanding was a little bit different, he doesn’t looking for chapters 1,2,3 of “how implement SPC”… Β there are a lot of books and articles available and I think he know how toΒ start it ….Β  he seems be more interested in how to explane (convince ?) his new company’s owners about SPC benefits, so he can startΒ  to implement it.Β  Unfurtunately, he haven’t found SPC supporters yet, hereΒ …
    Rodrigo, could you clarify, so that who interested can give appropiate answers ?
    Rgs, Peppe

    0
    #143636

    Rodrigo
    Member

    ok, now we are getting somewhere.
    this thread turned out to be complex when it didnt have to.
    Ray summed up everything quite nicely.

    0
    #143637

    Peppe
    Participant

    WhenΒ it is beginningΒ to be clear what you want, you close it ?Β 
    Rgs, Peppe
    Β 

    0
    #143638

    Jonathon Andell
    Participant

    If you can get the top brass to read one of two books it might help:”Understanding Variation” by WheelerThe chapters on variation in Joiner’s “Fourth Generation Management”I prefer the joiner book because it puts SPC in a more global conext. However, the Wheeler book cuts right to the chase of common cause and special cause. Both books are easy to read.I have found that many folks claim to understand the concept of common vs. special cause, but it never is reflected in their decision making – every data point is treated like a special cause. Joiner is good at pointing out how and why that equals money down the drain, which is the best reason to change.Hope this helps.

    0
    #143708

    Philip
    Participant

    Suggest you get the owners a copy of “Understanding Variation, The Key to Managing Chaos” by Don Wheeler, and ask them to read it.

    0
    #143727

    Arthur
    Participant

    While they are at the bookshop, have them buy Wheeler’s other books:
    “Understanding SPC” and “Advanced Topics in SPC”
    by Donald J. Wheeler, Ph.D.Fellow American Statistical AssociationFellow American Society for Quality
    Wheeler is one guy in this morass of ignorance who knows what he is talking about.
    He says of Six Sigma “an example of gross incompetence and a triumph of computation over common sense”

    0
    #143729

    Orang_Utan
    Participant

    Ask Dr.Wheeler how much money saved for the society with his indepth knowledge on SPC compared to “incompetence” Six Sigma?This is main difference in the mindset for an academian and a practitioner.

    0
    #143731

    Pinto
    Participant

    Go back to the trees Orang_Utan.Β  Next you’ll be claiming Deming was just an academic.Β 
    Who is six sigma’s “great man” ?Β … and please don’t say that foolΒ Mikel with his 3.4 DPMO nonsense.
    … and just ask Ford and GM how much six sigma has cost them Β !!!

    0
    #143758

    Orang_Utan
    Participant

    Fact 1: Dr. Deming is an academic?? He was first rejected by American people for being a practitioner.Fact 2: 3.4 DPMO was not mooted by Mikel Harry.Fact 3: Six Sigma costs nothing to Ford and GM. Is their people own mindset and attitude costs them a lot.

    0
    #143760

    Pinto
    Participant

    US Top 25 Business Leaders includes Deming :http://www.cnn.com/2005/US/02/28/cnn25.top.business/At least we agree that Mikel is a fool.Β Β 

    0
    #143764

    Orang_Utan
    Participant

    Shame on you, you are ignorant plus arrogant.Go read histroy on Dr. Deming. He was RE-recognized as a guru in America only in 80s after a TV show called “If Japan Can, Why CanΒ’t We?” was aired by NBC. He gave his first lecture in Japan in 1950 to Japanese managers. No one in America believe his philosophy except Japanese until America was in deep trouble against Japanese product quality in 80s. Six Sigma based on Deming’s variation concept was born as a response to this challenge.The highest quaity award in Japan is Deming award, not Malcom Baldrige award.

    0
    #143780

    Philip
    Participant

    I most wholeheartedly agree. However, Wheeler does not criticise Six-Sigma per se. He just criticises Six-Sigma done wrong. Unfortunately, this is true of about 80% of practitioners and consultants, including the industrial leaders like GE and Motorola. Wheeler’s book “Guide to Data Analysis” is a “diamond mine” of understanding and contains the keys to doing “Six-Sigma” effectively. In particular Wheeler debunks garbage like “proportion defective” under the extreme tails of a (hypothetical) normal distribution, and the completely spurious “1.5 sigma shift” as justification for striving for Six-Sigma, and demonstrates an alternative, and credible, alternative justification based on minimising the total cost of production and use.

    0
    #143825

    s. k. sharma
    Member

    In addition to establishing measurement for CTPsΒ it is also important to establish measurment sysstem analysis.(MSA , Gauge R & R). so that you estabish that your measurement variation is less then process variation.
    rgds…..S.K.Sharma

    0
    #143908

    Dennis
    Participant

    Salutations Rodrigo,
    As a novice to Six Sigma methodology I can’t offer advice on your control chart problem but, it appears to me that you are having problems with getting the family to buy into the program so to speak. I suggest you identify the “power behind the throne” and approach them with a “bottom line” scenerio. If you can present the cost savings the methods will produce and link them to an increased bottom line you may have a lot more success in your endeavers. I wish you good luck.
    Dennis

    0
    #144082

    Scott
    Member

    As an MBBΒ working with the very industrial companies mentioned earlier (GE and Motorola) forΒ overΒ twenty yearsΒ I leave you all withΒ one idea. Learn the truth and purpose of SPC with theΒ methodology of finding root causes from Six Sigma and the methodology of Toyota’s Production System and you will differentiate your quality and profitability from where you are today.
    Don’t wait on your boss. IfΒ he/she doesn’t understand, competition will take care of that issue. Observe and share the success from your implementation and recognize the team members that helped. Reading Wheeler is a good idea, too, and get out and see real successes.
    Tell us how it goes. Good luck and keep balancing asking with trying. Get started tomorrow.

    0
    #144088

    GB
    Participant

    Excellent advice Rick.
    Β 

    0
Viewing 38 posts - 1 through 38 (of 38 total)

The forum ‘General’ is closed to new topics and replies.