iSixSigma

Re-worked Rejects

Six Sigma – iSixSigma Forums Old Forums General Re-worked Rejects

Viewing 6 posts - 1 through 6 (of 6 total)
  • Author
    Posts
  • #49471

    Sinnicks
    Participant

    Am working with a company with a fairly high reject rate for a final product. They are inspecting 100% and rejecting ca 3%.
    That is the main problem we are working on but recently noticed they are reworking rejects and feeding them back into inspection where they fail at an even higher rate. Problem is they are measured twice and skew that data.  I can think of various ways to change this but anybody got ideas as to best practice?
     
     

    0
    #169126

    Swaggerty
    Participant

    Rolled throughput Yield??

    0
    #169135

    HF Chris
    Participant

    To me, a reject is a reject is a reject. How would you handle the reject if it occurred in the next work area… it is still a reject. You have lost time and money by reworking a part. You also have added a part into the process that is not part of the planned production schedule. I would be more concerned about how it affects your critical path right now. I can imagine cycles times with a wide spread… possibly even trending as a controlled process if you do this often. In answer to your question, if you run the part through twice you can decide not to count it twice thus keeping your same initial part count as long as you add the time for production and cost for production to that original part as if it happened at the same time and increase the count of defects for that part. HF Chris

    0
    #169140

    Ron
    Member

    Your issue is with the metrics. A true first pass yield will tell you what percent of defects you have on initial production – Metric #1
    In your repair line  (assuming you have a repair line) the output of that process is metric #2.
    It sounds as if you do not inspect your product after repair and this is the issue that is distorting your data.
    Rolled Throughput yield would be the defect rate of process #1 times the defect rate of process #2. This will really highlight your issues but you need to separate the two process measurement.
     

    0
    #169142

    BTDT
    Participant

    Mark:The primary metric for the project should be the rolled throughput yield. Management are likely only seeing the classical yield and do not see the impact of the rework in terms of costs of poor quality. Go through the calculations in the attached Powerpoint to show the team and stakeholders.Gather the data on the defects and chase the root causes. Process simulation software may help you to show the overall impact.Cheers, Alastair http://docs.google.com/Presentation?id=dftm5cbn_2gbsnj7d5

    0
    #169143

    BTDT
    Participant

    Mark:This link seems to work better.Cheers, Alastairhttp://docs.google.com/Presentation?docid=dftm5cbn_2gbsnj7d5&hl=en#

    0
Viewing 6 posts - 1 through 6 (of 6 total)

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