iSixSigma

Control Charts in Job Shop Operations

Six Sigma – iSixSigma Forums Old Forums General Control Charts in Job Shop Operations

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #28686

    howe
    Participant

    Any suggestions on how to set control limits in a job shop operation where short production runs are common?  By the time you have enough data to establish new limits, the run is over.  And we can’t use the limits from the previous run since there is often significant variability between runs (i.e. new material lot).  This makes control charting very difficult.

    0
    #71751

    Shree Phadnis
    Member

    Dear Mike,
    Use Z charts.
    Shree Phadnis

    0
    #71793

    Chaz
    Participant

    There is quite a number of books on Short Run SPC, you can also search this site reference information, or check your town library.
    In short (no pun intended) the method consist of plotting multiple parts on the same chart using the deviation from nominal.
    As to your problem with lots of materials – I would call this an opportunity for improvement. Most likely, not knowing the details, you woudl have to fix the lot to lot inconsistencies. That applies to all charts.

    0
    #71838

    howe
    Participant

    We can’t do different parts in the same charts.  The parts don’t have the same dimensions to check.  They are usually a completely different part design with different measureables.  We are not dealing with a family of similar parts with different part numbers…we are dealing with completely different parts.

    0
    #71841

    Bobby
    Participant

    You may want to look at DFSS instead of DMAIC.  Basically with DFSS you will ensure your process meets all necessary requirements (specs, CTQs, etc.)

    0
    #71843

    Erik Laufer
    Participant

    I’d have to agree that the lot-to-lot variability is a concern.  Focus on eliminating this as the major source of variability in the process.  Do you know that that is the biggest source?  A hiercarchial COV approach might be a good way to analyze the process and look for positional, temporal, and sequential sources of variation within your process.
    A more fundamental question is whether your charting the right things.  It’s often a mistake to do tons of charting.  There should be charts against a few critical metrics. 
    To the reply of different dimensions, etc., the response to use the Z-chart (or ZmR) is still valid.
    If none of these are palatable, then start using IX-MR charts, one per metric, and begin charting (though hardly preferred, one should focus on the process performance vice chasing specific product characteristics).  If you can’t control the lot-to-lot variation per run, you’ll have to re-set the limits when a new lot comes on board.  The IX-MR should give you a good glimpse into the process in as little as 5 points.
    Regards,
    Erik

    0
    #71846

    howe
    Participant

    For the most part, we cannot control the lot-to-lot variation.  We are at the mercy of our material suppliers.  My bigger concern is being able to chart short runs.

    0
    #71900

    Bobby
    Participant

    Sounds like your material supplier could use some improvement if they are providing materials that vary so greatly. Are there other companies that you can obtain your materials from? If so, you might be able to determine which provides a more consistent grade of material. Even if one costs more than the other, the decrease in rework, scrap, etc may offset the increased cost. You may be able to encourage them to improve the quality/consistency of the materials they provide.

    0
Viewing 8 posts - 1 through 8 (of 8 total)

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