iSixSigma

MSA – Stopwatch

Six Sigma – iSixSigma Forums Old Forums General MSA – Stopwatch

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #29421

    Mark McEntire
    Participant

    Can someone please make a suggestion for me?  How can I MSA an inspector using a stopwatch to measure cycle time.  I know I need to choose an event and measure it several times for Reproducibility, but I can’t think of an event.  Please help!!!!

    0
    #75365

    Joy Cowling
    Participant

    Mark,
    Why wouldn’t you use the actual event he will be measuring in production?  If you need to test repeatability, you can video tape the event so that he can judge the same event more than one time. 
    If you could provide more information about the process, perhaps I can come up with something more specific.
    Joy

    0
    #75369

    Erik L
    Participant

    Mark,
    My recommendation would be to map out the inspection process.  With that you can look at partitioning the total time to inspect to various activities.  What granulaity are you looking for in time?  Hours, minutes, seconds, nanoseconds?  Is your measuring device appropriate for the application?  If all you’re looking at is minutes, perhaps you could have the operator hit a bell at the initiation and conclusion of activities.  You could then track the time.  If the result is a Conforming/Non-Conforming label, you could look at performing an attribute gage R&R.  If you can seed units in, where you know what the result should be, you can look at the degree to which the operator agrees with the standard.  If you have multiple operators you could then look to see the degree of agreement amongst all operators.  With these two pieces of information you could possible look at a correlation between time to perform and the degree of agreement between the operator and the master value.
    Regards,
    Erik

    0
    #75399

    James C. Bailey,Jr.
    Participant

    In Process Owner training we do the MSA with a stopwatch for measuring the time it takes for a paper helicopter to hit the ground after release.
    The person holding the various types of paper helicopters is standing on a 6′ stepladder. We use
    three people measuring when it hits the ground.
    It is a simple exercise that proves the point!

    0
    #75401

    RR Kunes
    Member

    To perform an MSA on a piece of measuring equipment you must heck it against a knon standard. If you do not you will not know if the variability is in the stopwatch ot whatever else you are measuring.
     
    Also consider the accuracy you are attempting to measure. If you are looking for milliseconds you have the wrong approach. If you are looking for minutes you are probably okay.
     
    What is the expected variance in the stopwatch?
     
    If nothing else clock the passage of the electronic time generated by most computers.
     

    0
    #75407

    TCJ
    Member

    Is the system visual or audible?  Use the time elasp between a ringing phone or drop a pen from a specific height.  You need to choose different operators for reproducibility, and multiple times for repeatability.

    0
    #75410

    Terry
    Member

    Randy is correct again!
    If you’re trying to measure in .001 minutes then you will see an error in both repeatability and reproducibility.  I already did this as a Six Sigma R&R training exercise because my first job 30 years ago was a timestudy IE.  Back then the Company I worked for wanted to “make sure those cost accounting and incentive standards were 100% accurate.”  Naive precision at its best, but I wanted to prove it with data and also thought it would make a good BB training exercise.
    There’s a lot of human error reading a stopwatch and capturing what you thought you saw at the exact moment an elemental task ended (It’s like the catapult without the aluminum foil).  Especially if it’s a repeatable assembly with many quick cycle elements and many operators doing (not) exactly the same procedure.    You’re watching an operator’s motions, then you’re scanning your eyesight back to the stopwatch for a read and record while you’re trying to watch the next element.  The 3 bank stopwatch boards are even better because they multiply the error and sometimes the levers don’t work properly (These are the boards where you hit the lever, the first stopwatch starts, you hit the lever a 2nd time, it stops the first watch for a read and starts the second watch.  You hit the lever a 3rd time it stops the 2nd watch for a read, zeros the first watch, and starts the 3rd watch.  Then the 4th time it stops the third watch, starts the first watch, and zeros the second watch.  You keep hitting the lever and observing, reading the 3 watches, and recording).  Sometimes you go blank, get bored stiff, accidently hit the lever, or forget which watch to read which adds even more error.  I hope nobody does this stuff for real anymore, except in Six Sigma training or industrial antique collections!
    Like Randy said, your level of desired precision is important.  If you’re studying someone “shoveling coal into a boxcar” like the old Frederick Taylor studies, you’re fine.  If you’re looking for time to the nearest minute or .1 minute, you’re OK unless there is significant variation in the procedure from operator to operator.  If it’s miliseconds, you’re dead.
    Good luck
    Terry

    0
    #75443

    Leo Matikainen
    Participant

    Mark,  If the question is simply how to get a repeating event, and the measurement that you seek is the start/stop function of the operator/watch try this:  Mask off the face of a clock with a second hand.  For example, run a masking tape piece up to the twelve-o-clock position, and a second to the one-o-clock position, thus forming a wedge (five second window).  Do a similar “wedge-job” on another part of the clock face; say at twenty-two past the hour and at twenty five past the hour (three-second window).  Mask off the rest of the clock face. Now you have a repeating signal where you can ask the operator to time between the first and second appearances of the second hand from behind the masked-off areas of the clock (either 22 or 38 s).  You can vary the speed of the clock hands easily enough if it is a battery powered clock by changing the battery type (that is the battery voltage).

    0
    #75502

    Andrew M. Brody
    Participant

    Each process will provide different opportunities for measurement.  I work in the paint industry and gauge R&Rs are reqired by our QS 9000 registration.  For our stopwatches,  we use a viscosity cup that has been verified by an outside Guide 25 test lab.  The measuring media is NIST traceable viscosity test oil.  Some of our stopwatches are to the second and some to the 1/10 of a second.  100 trials are run on each of the stopwatches by one technician then repeated by another technician with the results plugged into a well known and widely accepted software program.  Our results were well within the acceptable G R&R range.
    Hope this helps.  Our registrar was pleased with our methods.
    Andy Brody

    0
Viewing 9 posts - 1 through 9 (of 9 total)

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