iSixSigma

Correct Control Chart/Tool

Six Sigma – iSixSigma Forums Old Forums General Correct Control Chart/Tool

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
    Posts
  • #30321

    Edwards
    Participant

    I’ve implemented a plan to reduce waiting time of units being processed. The criteria is that 95% of the units that arrive at the processing station must begin being processed within 1 hour of arrival.
    What type of control chart/tools would be best to monitor this waiting time to gauge the effectiveness of the effort for a period of two months after the implementation?
    Datas of the waiting time are available on a daily basis. Note that the units arrives in a variable batch size (1 to 10) at anytime. Thanks.

    0
    #78842

    Hemanth
    Participant

    Hi this is just a thought. I would monitor waiting for each call/request, and plot it on a I-MR chart. Once I have sufficient data points on waiting time I would look at 95% confidence interval for the distribution, and check if that is greater than 1 hour to verify my claim that 95% of requests are processed under 1 hour time.
    Hope this helped.
    Hemanth

    0
    #78843

    Edwards
    Participant

    Thanks Hemanth. I have never plot a I-MR chart. Can’t seem to find it in my statistics software (Statgraphics). Any good source you can point out? Seems like I-MR is uncommon as I cannot find much info on it online.

    0
    #78844

    Hemanth
    Participant

    Hi
    I-MR stands for Individual – Moving Range chart. As the name suggest you plot individual values for waiting time on the individual chart (corresponding to X bar chart on Xbar and R chart). For calculating UCL and LCL use typical standard deviation formula of Root mean square of deviation from the mean.
    Plot moving range on the moving range chart. Moving range can be calculated by taking difference between two consecutive points. For calculating use d2 value for subgroup size 2.
    In a way I-Mr is similar to Xbar R chart, the only thing being here subgroup size is one as in your case.
    Look in book on SPC by wheeler its a nice book on Statistical process control.

    0
    #78853

    Robert Butler
    Participant

      In addition to following Hermanth’s suggestion for tracking individuals I would also recommend that for the two month test period you track and plot the arrival times and quantity per arriving batch.  If you just track daily time you are assuming that there is no difference between a batch of one and a batch of ten and you are assuming that something arriving at 8:30 A.M. will have the same chance for initial touch time as something arriving five minutes before lunch break or fifteen minutes before quitting time.  Granted that gathering this additional data may be cumbersome and/or inconvenient but if you make it clear to everyone that this data gathering effort has a reason and has a definite “drop dead” date you will probably have little difficulty gathering it.  The combined data will permit a check for trending in waiting time as a function of arrival time and batch size. 
      If the check reveals no connection well and good.  If it does, it will help guide you in making appropriate changes to your plan.

    0
    #78877

    Ron
    Member

    An ImR chart is the totally wrongtool to use. ImR charts are single point data and have very littler information contained therein.
    I would suggest based on the frequency of your incoming product an XBar & R chart.  This has the ability to detect trends both within and between the sampling times.
     
    After you establish that the process is in control utilize an EWMA chart to detect small shifts in the mean. You will not have to change sampling schemes.
     

    0
    #78880

    Erik L
    Participant

    I’ll piggy-back off of what Robert is suggesting here.  My recommendation would be to create a multiple regression for this scenario.  With the variables selected as qnty, time, etc. you have two options really.  First, track the output as the physical time to begin processing-hence keeping the response in a continuous manner.  Second, make the response binary-Logistic Regression.  Processed or not processed within the 1 hour window would become your ouput response.  For the first option, you could create a regressional relationship and then super-impose 95% CIs and PIs for prediction and potential control of the process.  If the band is too broad for these limits, control charting could help you by partitioning the variation between and within the subgroups and provide the springboard to where the biggest source lies.   Good luck with the analysis.
    Regards,Erik

    0
    #139329

    Anupam Rohatgi
    Participant

    I think you should make a Value Stream Map.
    Calculate the Process Lead Time = WIP / Exit Rate.
    Exit Rate = Finished Product/ Hr.
    High WIP means high waiting time.

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

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