iSixSigma

DPMO for a production process

Six Sigma – iSixSigma Forums Old Forums Europe DPMO for a production process

Viewing 14 posts - 1 through 14 (of 14 total)
  • Author
    Posts
  • #23624

    aio
    Participant

    I am starting a new Six sigma project and planning the measures for the project outcomes. The project has to do with a complex manual assembly process with several process steps.
    In the literature I have seen processes evaluated by DPMO values, but I have never used it. Could you advise me where could I find some practical information how to start with DPMO calculations?
    Thanks in advance!

    0
    #56404

    Adam L Bowden
    Participant

    In the electronics world DPMO was used to identify areas to focus improvements.  Use DPMO CAREFULLY as you will get into deep and hot water fast unless you are prepared.
    One way of looking at it is:
    – count all components and multiply this by the opportunity for a defect – in the case on an electronic component it could be 1) the wrong component, 2) incorrectly fitted 3) defective i.e if you have 1000 components then you have 3000 potential oppoertunities for a defect. 
    Use the actual number of defects, divide by opportunity count then multiply by 1,000,000.
    Be carefull how you apply this !!
    Good luck,
    Adam
     

    0
    #56413

    Amrish Khanolkar
    Participant

    DPMO is Defects Per Million Opportunities. To use it first of all for any process you have to jot down all the opportunities you have to make a defect. In short the different types of defects a process can give to the product. Then you have to evaluate your current process as to how many defects it produces for a million opportunities. Here, for a single item if they are say 5 defects it has to be counted as 5 defects only.
    I will not advise you to use DPMO because you are never sure about the opportunities. Every time you think over you can add more opportunities to your list. There is no end to the opportunities. That is the reason you dont often find it in literature.
    Better is to use DPU (Defects Per Unit), which gives no. of defects that are seen per unit. This can be directly used to calculate your Rolled Throughput Yield (RTY) of your continuous processes.
    RTY = e raised to -DPU
    This gives the probability that the product will be First time right, after going through a string of processes.
    But if you still insist in using DPMO. Good Luck!

    0
    #56416

    Howard Miller
    Participant

    As DPMO is a ratio then you can use it as a rolling data set. As long as your intervals are set in stone then it should be fine. For example weekly or monthly.
    Then get a little smart and dump your data into the following equation in excel:
    NORMSINV(1-((Total Defects) / (Total Opportunities))) + 1.5
    Equation lifted from this site’s calculator.
    Then track your performance this way.
    T.

    0
    #56419

    Mike Carnell
    Participant

    Aio,
    The appropriate measurement depends on what you want to use it for. DPMO introduces a complexity factor by using the number of opportunities i.e. if you were the CEO of a company that made light bulds, jet engines, and nuclear fuel rods. If you are making a decision where you need to compare things a varying complexity. There are a lot of different opinions about opportunities so it sounds difficult but it really is not – for you try something as simple as parts plus connections. Remember it is the opportunity to do something correctly. If you start to count the number of ways you can do it incorrectly you will end up at six sigma or better through denominator manipulation.
    If you are just running a project and are not comparing it to anything else dpu is the simplest to calculate and the easiest for everyone to understand (total number of defects/ units produced). It is sensitive enough that the team will be able to see movement in the metric with small improvements.
    Sigma is the worst metric you can use at the project level. If you doa sensitivity analysis on the metric you will find it doesn’t move easily. That will typically result in the team working extremely had anf the change in the metric is extremely small. This is a metric with the same characteristics as DPMO.
    Just my opinion.
    Good luck.

    0
    #56421

    Cone
    Participant

    See where you are coming from Michael however my thoughts are that DPMO measure ensures that you have a robust process that consistently performs perfectly DPU will give you a short term efficiency measure but no use patting yourself on the back on Monday when the rest of the week your rubbish. I do agree that measuring your team efficiency by sigma can cause considerable difficulties but perhaps the main problem is educating your workers in the significance of a 1% sigma shift rather than changing the measure. Six Sigma is all about reducing variation and increasing consistency so may not be good for its integration into your business. Cheers ATB Gary 

    0
    #56423

    kalju
    Participant

    You can find practical examples good ideas from standards: IPC-9261 and IPC-7912.

    0
    #56428

    aio
    Participant

    Hi,
    I do not work in electronics industry, but thanks anyway, I think the examples can be useful.

    0
    #56429

    Mike Carnell
    Participant

    Gary,
    The idea that DPMO ensures you have a robust process sounds really nice. I would appreciate an explanation of how that works.
    As far as the team measuring itself in Sigma it makes no sense. It certainly adds no value and lacks discrimination. When you have a part that is dimensioned in 0.001 of an inch and you use a carpenters ruler to measure the measurement system lacks discrimination to sense the difference in parts. Sigma as a measurement at the project level has exactly the same issue. Run a sensitivity analysis.
    As far as educating the workers. If they are basically taught the basics of measurement systems that are relative to them they will figure out the issues around shift. As far a 1% sigma shift – irrelevant to most.
    There is a basic advantage to sigma. It allows the comparison of dissimilar things – jet engines and light bulbs. At the project level it is a complexity factor that adds no value other than esoteric discussion.
    Good luck.

    0
    #56431

    I’Anson
    Participant

    hello I am a Black Belt from Germany , and if you need an example for a Project with DPMO in the Production i can send it to you !
    Hopefully it helps  

    0
    #56432

    Joseph
    Participant

    Daniel: Can you post the prestnation here? I would very much appreciate seeing it also! Thanks for sharing.

    0
    #56433

    I’Anson
    Participant

    yes i can do it but this presention is only in german language so, did you want to have an english one so i must translate it at first . please let know.

    0
    #56434

    aio
    Participant

    Deutsch ist i.O.
    German language is no problem. Thanks in advance!

    0
    #56435

    I’Anson
    Participant

    it was my first one in 2002, now the projects will be much better! Believe me, they will be much better.
    But it doesnt works with the attached files here please send me your E- Mail adress . I have make it in english for you !

    0
Viewing 14 posts - 1 through 14 (of 14 total)

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