Target, Actual, Please Explain

I learned some important lessons today. One which I will share is to never compromise when it comes to placing visual status boards on the shop floor. Visual status boards promote problem solving and kaizen by exposing problems.
There are many excuses and reasons people give for why this is hard to do. None of these reasons are as hard as continuing to live with the problems and the people who accept them.
Fact based gemba kaizen is as simple as target, actual, please explain. If there is no target, you can’t improve. If you don’t know actual, you can’t improve. If you can’t explain, there’s a worn piece of chalk I’ll lend you to scratch out a circle to stand in and watch for a while.
Hour by hour charts, sometimes called hourly production status boards or some other inelegant name, are wonderful things. Wherever you don’t see these things, you are not seeing shop floor management.
Instructions for constructing an hour by hour chart:
1) Acquire flip chart
2) Acquire pen
3) Draw a grid, as below
4) Fill grid, as below
Target%20actual%20please%20explain.PNG
This is very easy. You can do it everywhere. No excuses. Just do it.
Simple is best. The 50 inch plasma screen digital display networked in real time with the ERP system and globally synchronized to all factory management dashboards should be used only if you have too much money and no access to pen and paper. If your hour by hour chart looks much prettier than the one above, it is muda.
My pledge is that from now on, if I ask for visual status boards to be placed on the gemba, I will not return to that gemba until this is done. Otherwise we’re wasting each other’s time.

6 Comments

  1. Keith Pincher

    August 29, 2007 - 1:53 am

    Is the entry for ‘please explain’ usually “it’s just common cause variation, dude”?

  2. Keith Pincher

    August 29, 2007 - 3:32 am

    I have a target of four red beads. I know that 20% of the beads in the box are red. When I drew out 50 beads I got 14 red, even though I know the system should produce an average around 10. How do I word my explanation? I can’t ask Deming. He’s dead.

  3. Jon

    August 29, 2007 - 8:48 am

    Hi Keith,
    Your first question:
    Most likely “it’s just common cause variation” would not be an acceptable explanation, since if a process is under control, or has predictable variation due to chance or common cause, the “target” should allow for this. If your target is zero variation, and if this is a serious target, you may say “it’s common cause variation” but not use the word “just,” which implies that you are excused you from further analysis.
    Your second question:
    I’m not sure this is an appropriate analogy to the hour by hour chart since the target is based on an expected output over a period of time, and the “please explain” is addressing variation during that period. If your target is 4 red beads, why did you draw out 14?
    If you do have a process with chance (common cause) variation to this degree, you don’t have a stable process and you should stop measuring hour by hour and start working controlling and stabilizing it.
    Unless we’re talking about quantum level processes, I think we can say that this is not common cause variation but special cause variation.
    If so, there would have been an error in the man, material, method, machine, measure or mother nature elements resulting in more than 20% of the beads being red in the box.
    The explanation would require investigation by going to the workplace, opening the box, interviewing the worker, observing the method, calling the supplier, etc.
    We can’t ask Deming, but he left plenty of students you could ask.

  4. john zavacki

    September 3, 2007 - 1:55 pm

    There’s something wrong with the reasoning here. The point of the Red Bead Experiment is to show that you cannot reduce process mean or it’s standard deviation without remove the cause of the variation, which is the read beads. If you got only two red beads, would you have improved the process? Of course not. Since 20% of the beads are red, you’re going to get a draw of none and a draw of 20% from time to time. Translated into real world terms, the common cause may be poor tooling, poor equipment, poor materials, etc. A special cause would be a broken tool, the wrong material, machine break-down, etc.
    A second point, the target is not 4 red beads, it’s 0 red beads, another reason the experiment is so frustrating. To get 0, you need also a special cause, which in this case, may be just dumb luck.

  5. Brandon Nagro

    February 21, 2008 - 4:08 pm

    I have an automotive assembly line and it contains six different stations. The vehicles are on a automated line that has the speed controlled according to the fixed takt time. If an andon is hit the whole line stops. How do I use an hour by hour chart to expose the reasons for the line stoppages????

  6. Venkatesh

    March 15, 2008 - 12:38 pm

    Our ‘Hourly Production Tracking Sheet” showed us significant variation in throughput time due the following reasons:
    1. Skid Prep Paper Work
    2. Material Replenishment (Waiting for a fork truck)
    Anybody with similar issues?
    Regards,
    Venkatesh