Tuesday, March 29, 2011

Quick Changeover Workshop - Part 2

In the last posting I listed the following techniques to reduce the time it takes to  changeover a machine from one product to another.
1.     A verbal written record of every action of the people doing the changeover
2.     A spaghetti diagram of the changeover operators’ movements.
3.     A measurement of the distant walked, either using a pedometer or counting the operators’ steps.
4.     A written capture of ideas for improving the changeover, as these come to mind.

This is the form that I use make a written record of what people are doing during the changeover that is being observed:

We record the activity of the operator being followed in detail. We especially want to be able to distinguish among the activities that could be done before the machine stops producing or after it is producing again, those that must be done while the machine is stopped, and those which are simply waste or muda.
In the example, the operator could have had the lock and tag ready before shutting the machine down.  This could be “external” time.  The actual act of locking out must be done after the machine is shut down.
Walking to the opposite side of the press adds no value to the changeover process.  If the lever were next to the electrical panel where the lockout is placed, the walking could be eliminated.
The operator must hold the lever down for 56 seconds to release the pneumatic pressure.  We again have waste.  A different release mechanism would allow him to pull the lever or push the button once and go on to the next element of activity.
Now the operator goes to fetch a hand truck. Accumulating product in a container is waste, leading to the additional waste requiring a hand truck to move it.  If the next operation were performed immediately, on an adjacent machine, there would be no need to move these parts with a hand truck.  For immediate purposes, we can think of this activity as potentially external.  The hand truck could be put close at hand before the machine was shut down.
In the workshop we typically leave the three right-hand columns blank until after the changeover is complete, focusing instead on keeping track of the running time at the start of each new activity, in the left-hand column.
Normally, we will fill many pages of this form in one changeover.  After the observation is complete, we categorize each item as Internal, External, or Muda, putting the time for the element in the corresponding column on the worksheet.  We then add up the times in each column and calculate their percentage of total time. The time that must be internal will be a small fraction of the total time, revealing that we spend most of the changeover doing things that are wasteful or could be done before after the period during which the machine is not producing.
The spaghetti diagram of the operators movement gives us a visual representation of the amount of activity that is wasted in movement without adding value.  Here is a typical first spaghetti diagram:

As we make improvements and repeat our observations we can see the impact dramatically by comparing spaghetti diagrams.
Data from a pedometer or by having an observer count the number of steps the operator takes is provides a measure of the potential for improvement and actual improvements obtained in the workshop.  The number of steps and distance traveled should drop dramatically.
Throughout the changeover the members of the team  will have ideas about ways the changeover can be improved.  If all nuts were the same size, we could work with a single wrench instead of the half-dozen sizes we use now. We would waste less time hunting for the right wrench.  If we chance the design of a clamp, we can eliminate a tool altogether.  These ideas should be written down as they come to mind.  Some teams have each observer keep his own notes.  Others have a single recorder to whom all observers give their ideas to be written down. It is useful to note the running time on the note, to place it in context for later discussions.
After the first changeover has been observed, the team gathers to discuss their observations and select improvements that can be done quickly to reduce the time it takes to changeover the equipment, writing the ideas on a flip chart. The Detailed Activity Log is completed by identifying  those activities that could be external and those that are muda.  The team considers what simple changes might eliminate the muda or make internal activity external.  These ease of doing this will vary widely, but there is always much that can be done quickly and inexpensively. 
Tool boards can be installed, on which all of the tools needed are in a designated location and within reach.  The first versions can be done with a magic marker on a piece of cardboard.  In fact they should not be made too permanent and pretty, as we often find ways to eliminate tools, requiring the tool-board to be changed.
Is hunting for fasteners taking up time?  We can pre-stage the exact fasteners that are needed, where they are needed.  Are we spending a lot of time spinning down nuts because bolts are longer than they need to be?  Let’s select bolts of optimum length.  Are we wasting time because small parts fall and must be retrieved?  What can we do with small parts trays to prevent this from happening.
In some types of changeovers we discover opportunities to standardize that will he contribute to reducing the times they take.  Standardizing die heights  can make changing stamping presses quicker.  When there are problems to standardize to a single height, we can we at least minimize the variation by standardizing to two or three heights instead of a half dozen or more.  Often we can we standardize clamps and bolts. Generally we ask, where can we eliminate variation through standardization?
A changeover improvement workshop can be exciting and inspiring.  It is not unusually to reduce the time it takes to conduct a particular changeover in half with one or two iterations, implementing low-cost / no-cost improvements after each observation.  Getting to SMED (Single-Minute-Exchange-of-Dies) will take more work, but the workshop will provide good indications of where that work needs to be done.
There is a danger that the necessary changes will not be standardized and maintained.  While the team was working, everything was meticulously pre-staged, but we did not then put the procedures into place so that the pre-staging is always done.  While the machine is running, who is going to get the tooling and material ready, so that the internal time can be kept to a minimum?  The operator may or many not be able to do it.  If not, who will?  This will require a change in the operating processes in support of changeover.
Quick changeover is not independent of equipment maintenance.  An unreliable machine can be difficult to changeover in a reliable time. 
As with a lot of lean processes, if the quick changeover is to be sustained, it requires a vision of how it fits in the overall system.   This vision must be translated into leadership that helps others see the big picture and understand all of the changes that are required.