iSixSigma

defining opportunity

Six Sigma – iSixSigma Forums Old Forums General defining opportunity

This topic contains 12 replies, has 12 voices, and was last updated by  Mike Carnell 16 years, 6 months ago.

Viewing 13 posts - 1 through 13 (of 13 total)
  • Author
    Posts
  • #30510

    Jon B.
    Participant

    I am trying to determine the number of opportunities in a welding process. The product contain approximatly 240″ of  weld and is made from 15-20 parts. My experience tells me we have a sound process and our defects per inch of weld is very very low.
    These welded parts must not leak and are 100% tested.
    If I count each tank as 1 opportinity and it either leaks or it doesn’t the process “looks” bad <3 sigma.
    If I count an opportunity as the smallest defect that will cause a leak it is nearly 6 sigma.
    Does anyone have any experience with a process/situation like this?
    Thanks Jon
     
     

    0
    #79529

    aush
    Participant

    This is my opinion and may not be acceptable
    I think the question of defect opportunities was developed to have a common platform to compare processes in a establishment . So the question of  how to define opportunities will depend on you considering how you want to represent the process. What ever you define becomes the baseline and from that you proceed for improvement.
    as you say your process can be 6 sigma or 3 sigma depending on the opportunity defined , it is basically what you beleive is true. If you believe that your process is 6 sigma then your defect rate should be really low ,however this does not seem to be the case. So select defect opportunities which will reflect the process.
     

    0
    #79562

    Jack Yang
    Participant

    I think this is a good question in terms of identifying opportunities which is not “discreet”.  Here are my brief thoughts…
    I believe defining opportunities should depend on the actual process vs. preference.  Depending on difference welding techinique, we look at the process differently, and we look the causes of leading differently. For instance, suppose the leak points most occur between welding points, the number of inter-welding points become defect oppoutunities. 

    0
    #79563

    Ganapathi
    Participant

    Jon,
    It’s similar to no. of opportunities for a soldering process on a circuit board. The welder  most likely does a specific length (say one inch) at a time. Therefore you can define this specific length (what the welder covers in every attempt) as an opportunity. Average length over a period of time may be considered if there is variation from time to time. THIS MAY BE VERY CLOSE TO NO.OF INTER-WELDING POINTS, as suggested by Jack.
    Ganapathi

    0
    #79566

    Hunda
    Participant

    I think the best way for such a situation is to find a way to make the length of weld discrete.  Determine which is the smallest length of weld that can have only one defect.  If it is 1/1000 “, then in 240” there will be 240,000 opportunities.  You can then work with the number of the defects and the opportunities in the normal way

    0
    #79567

    Ron
    Member

    Jon,
     
    I think you answereed your own question. The process looks bad. Opportunity count is game that can be played if your trying to fool someone. Various scenarios have been used to placate this tempation.
    In your case I would look at the process outcome. It cannot leak. Building rework into your process is just plane dumb.
    At best I would identify each joining of a part(s) as an opportunity.
    But you are missing the big picture. If you must rework many of the products opportunity count is not your biggest concern.
     

    0
    #79573

    Jon B.
    Participant

    I must not have stated the issue clearly. I do beleive this welding  process is in control and the quality of the welding is, in my opinion second to none.
    My concern is, I could make a very good process “appear” to have a  problem by using too few opportunities.
    As someone suggested on this web site, I am leaning toward using the number of parts + number of connections + the number of weld transitions to measure opportinities.
    (I do have 20 plus years experience with weld quality ranging from the  nuclear defense industry to pressure vessels used in power generation to base this opinion on.)
    Thanks to everyone for thier input.
    Jon  

    0
    #79577

    shahzad
    Member

    An opportunity can be defiened by what’s listed in your current control/quality plan.  Each characteristic that you measure per the plan could result in a potential defect causing you to scrap or rework.  Therefore the number of characteristics you currently measure in a process equals the number of opportunities.
    -S

    0
    #79583

    Yeldo
    Member

    Hi Jon,
    Its a very good question as sigma calculation is pre-dominantly dependent on the opportunities for error.
    It is very necessary to therefore take a pragmatic view of how we define the opportunities. We will be able to define the opportunities for error by obtaining answers to the following questions
    a. How many processes are present in the welding process ?
    b. For each process, what are the different types of errors that have occurred in the past ?
    c. For each process, what are the different types of errors that can occur in future with the current process ?
    d. Are there any parameters present in the existing process(es) that are not covered in pts ‘b’ and ‘c’. If so, how many ?
    Once you have the figures for questions ‘b’, ‘c’ and ‘d’ for each process, we need to sum them to arrive at the figure of the opportunities to commit error.
    Bottomline is that you need to be honest in identifying all possible chances to commit errors !!
    Hope this helps.
    Thanks,
    Yeldo

    0
    #79585

    Withheld
    Member

    I’m glad I read all the responses before chiming in because I agree with your’s, Ron. You can only sell a completed weldment – not a subset of the various welds that create it. The opportunity, therefore, is one.
    Adding opportunities may (and probably is) be a necessary byproduct of the problem solving process, but should not be used as a way to paint a prettier picture than may otherwise exist.

    0
    #79660

    KennyCat
    Participant

    Ron,
    I recently completed a Six Sigma project on welding and another on fabrication type errors.  We fabricate pipe spools.  I used 8 opportunities per fabricated piece (because each spool piece is not identical, some have three pieces of pipe some have none.  Some have fittings and flanges, some have o’let connections, etc.) 
    I counted my total spools which came through that day and I counted the defects.  (Defects were brooken out according to weld [W1, W2, W3…] or fit project [F1, F2, F3…] and by the operator [1, 2, 3…], ie. fitter/welder).  This was entered in my L1 spreadsheet.
    I was able to track how many time undercut, slag, etc. showed and also was it caught in-process or final inspection, (to show improvements).
    If you like I can e-mail you my presentation.
     
     

    0
    #79662

    VINAYAK D GADRE
    Member

    Hi,
    Please send me your presentation.
    Regards.Vinayak
     

    0
    #79667

    Mike Carnell
    Participant

    Withheld,
    I am glad to see you back.
    I am going with you on this one in the overall  but probably for a different reason. When you work a project there is no reason to run DPMO (opportunity counts). DPU is a better measure and a more sensitive measure at the project level. It is also a better reflection of what your customer sees. Try explaining how good your process is to a customer that has a 100% leak rate because they don’t care about opportunities they care about – one tank at a time and they do not want any of them to leak. I do hope the people teaching this are not bypassing the link between CTQ’s and customer expectations. The money only comes from one place – customers. They are in a dpu mentality as they should be.
    Opportunities is complexity factor – period. It is what lets you level the playing field between dissimilar products. If you are Welch or Immelt and you have 12 Sectors all producing dissimilar products it is what lets you figure the difference in products. There is a distinct difference between a jet engine, light bulb, TV show and loan that are all running at 1 dpu. You can convert it to rty using e and it still isn’t right. That is why you use DPMO. When you figure who is most screwed up they get the most pressure, resources, etc. It helps with internal strategy or benchmarking to a similar process in a fifferent discipline (benchmarking a portable launch gantry against the process the roadies use to setup and tear down the stage for Pink Floyd on tour – actual case).
    I am sure there is someone out there that can figure out why they need opportunity counts on a project level. There are always exceptions. It isn’t where you start. The project level metric should be dpu.
    Yes I have worked a similar process on a thing called a charge air cooler. We had welding and braising. We went from 13 dpu to 0.5 dpu in 6 weeks. Just like Dominos used to be – “the clock is running.”
    Good luck.

    0
Viewing 13 posts - 1 through 13 (of 13 total)

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