Systems Improvement Part 2

In my last post, I explained that in this series of posts, I wanted to write about what I refer to as system's improvement.  I presented a basic definition of what a system is as defined by [1] Arnold and Wade as presented in their paper entitled, A Definition of Systems Thinking: A Systems Approach.  In their white paper they presented what they referred to as, “The System Test.”  While Arnold’s and Wade’s intention was to use this test to verify the requirements for a system’s thinking definition, I explained that my use will be to outline the basic structure of a manufacturing system and the thinking that goes along with it.  I explained in my last post that I had changed the wording originally presented by Arnold and Wade to describe the three characteristics of a manufacturing system, namely its purpose, the elements within a system, and the interconnectedness of the system’s elements.  I finished my last post by stating that we will consider at a simple example to better understand system's thinking in manufacturing. 

Systems always exist to realize a specific purpose and in reality, the purpose should be viewed as the goal of the system or the objective toward which all effort should be directed.  If we are attempting to improve our current system, then we must do so with our goal in mind.  Improvement implies that change will be required from the system’s current status, but because changes to our system can be either good or bad, we must do so with the ultimate system goal clearly in the forefront. Let’s now consider at a simple example. 

In Figure 1 below, we see the cross section of a simple piping system used to transport water (i.e. its purpose) starting from Section A.  Each of the different pipe diameters represents the basic elements of this interconnected piping system.  This system is gravity fed with water entering Section A, then flows into Section B and continues until water reaches a receptacle directly beneath Section I.  Suppose there was an increasing demand for more water and you have been assigned the responsibility to satisfy this increased demand.  What would you do and why would you do it?

Figure 1

It should be apparent that if more water is required, then you must first identify that part of the system that is limiting or constraining the output of water through this piping system.  In Figure 1 we see that the constraining factor is Section E’s diameter.  It should be evident that in order to increase the output of water through this system, Section E’s diameter must be enlarged.  What the new diameter must be is completely dependent upon the demand requirement being placed on this system.  In other words, how much more water is required?

Figure 2

Figure 2 is this same piping system with Section E’s diameter enlarged to allow more water to flow through the system.  The new output of water has clearly increased, but is now limited by a new constraining factor, Section B.  If there was another surge in demand for water, then our focal point will now be Section B.  So how does this simple piping system relate to a typical manufacturing system?

Figure 3 is a simple, linear manufacturing system with individual cycle times listed for each interconnected step.  Parts, or raw materials, enter Step 1, are processed for 30 minutes and are then passed on to Step 2.  In Step 2, the semi-finished product is processed for 45 minutes and passed on to Step 3.  Step 3 requires 90 minutes to process the semi-finished product and then passes it on to Step 4 which requires 30 minutes to process.  When Step 4 is completed, the finished product is sent directly to either shipping or to direct sales or is stored in racks to satisfy future orders.

Figure 3

If we wanted to increase the output rate of product through this manufacturing system, the first thing we must do is to locate that part of the system that is the limiting or constraining factor.  Just like we identified Section E in our piping system as the constraint, we must do the same thing for this manufacturing system.  Whereas in our piping system, in order to identify the constraint, we simply looked at the volume of water passing through each pipe which was proportional to its diameter, as well as looking for a “back-up” of water waiting to pass to the next section. In our manufacturing system we must identify which step has the longest cycle time.  Here we see that Step 3, at 90 minutes, is clearly the longest cycle time, so it is labeled as the system constraint.  If we wanted to increase the output of this manufacturing system, we would undoubtedly need to reduce the time required at Step 3.

This system, in its current state, can produce one part every 90 minutes because that is the rate of the system constraint.  Even though Steps 1, 2 and 4 can produce parts at much higher rates than Step 3, the total system is limited or constrained by Step 3’s output rate. Table 1 is a step-by-step summary of cycle times and output rates for this system for a typical 8-hour day.

Step #	Cycle Time	Output Rate for 8 Hours
1	30 minutes	16.0
2	45 minutes	10.7
3	90 minutes	5.3
4	30 minutes	5.3

Table 1

Clearly, Step 1’s output capacity is dominant at 16 parts every 8 hours, while Step 2’s rate is approximately 11 parts every 8 hours.  The system constraint (Step 3) can only produce product at about 5 parts in 8 hours which then limits what Step 4 can produce.  That is, Step 4 can only produce what Step 3 delivers to it.

The question now becomes, in its current state, how fast should Steps 1 and 2 be running?  In my next post, we will answer this question and continue our discussion on systems improvement.

References:

[1] Ross D. Arnold and Jon P. Wade, A Definition of Systems Thinking: A Systems Approach, 2015 Conference on Systems Engineering Research

System Improvement Part 1

In this series of posts, I want to write about what I refer to as system's improvement. In this series, it is my intention to first, lay out the basics of what systems are and then share some of the wonderful experiences I have had over the years in a variety of different companies in a variety of different industry segments.  It is my hope that you will achieve meaningful takeaways from this series and that those takeaways will be helpful in the future.  In this series I will write about industry segments such as manufacturing, healthcare, maintenance, etc. for both large and small companies.  Having said this, I will focus on manufacturing systems to drive home key points.

What is a System?

One of the keys to success in any industry is to understand that your company should be viewed and thought of in the context of a system, rather than just a collection of interrelated parts. So, you may be wondering, just what is a system? In its most basic form, a system is a group of interrelated, interdependent, and interacting parts that combine to achieve a specific purpose.  A system takes inputs in some form, acts on them in some way to produce outputs.  In reality, the outputs should have a greater value than the sum total of the inputs.  In other words, the system should add value to these inputs as it works to change them into outputs.

In 2015, [1] Arnold and Wade presented a paper entitled, A Definition of Systems Thinking: A Systems Approach.  In this paper they presented what they referred to as, “The System Test.”  The System Test, as described similarly to Figure 1 below, was devised as a means by which to test a system’s thinking definition.  The test, as presented by Arnold and Wade is relatively simple to follow and understand.  While Arnold’s and Wade’s intention was to use this test to verify the requirements for a system’s thinking definition, my use will be to outline the basic structure of a system and the thinking that goes along with it. As such, I have changed the wording originally presented by Arnold and Wade to describe the three characteristics of a manufacturing system, namely its purpose, the elements within a system and the interconnectedness of the system’s elements. So let's look at each of these three characteristics in more detail.

Figure 1

Characteristics of a Manufacturing System

1-   All systems exist to achieve a specific purpose, and one of the keys to understanding a system is to fully understand its intended purpose.  As an example, ask yourself what the purpose of a manufacturing system is. The basic purpose of a manufacturing system is to produce manufactured parts to satisfy customer’s requirements.  If all steps in the process are not functioning as they should, then the purpose will not be achieved.

2-  Every manufacturing system contains multiple elements and has at least one constraining factor that controls the output of the system. If the system’s purpose is to produce a product, then it is critical to locate the constraining factor, exploit it, and then subordinate the other parts of the system to it.

3-  The distinct order in which a manufacturing system is arranged and interconnected affects the performance of the system.  In other words, if the individual steps are not arranged in the correct order, then parts cannot be produced according to specified requirements.

4-  All steps in the process must be present in order for a system to achieve its intended purpose.  If, for example, one step experiences down time, then the system will not function for its intended purpose.

5-  In order for manufacturing systems to maintain stability, there must be a feedback mechanism in place to transmit information.  Without a mechanism to provide feedback, systems will not function to achieve its intended purpose.  Selecting the right performance metrics, for example, is critical for systems to operate effectively.

Understanding a System’s Purpose

To fully understand a manufacturing system, understanding its purpose is critical.  This fact is true whether it’s a separate entity or part of an even larger system.  For most systems the intended purpose is clear, but it is important for everyone interacting within the system to fully understand its purpose.  I say this because it’s important to understand that the output of a system is not the sum total of each of the individual components of the system. 

Systems always exist to realize a specific purpose and in reality, the purpose should be viewed as the goal of the system or the objective toward which all effort should be directed.  If we are attempting to improve our current system, then we must do so with our goal in mind.  Improvement implies that change will be required from the system’s current status, but because changes to our system can be either good or bad, we must do so with the ultimate system goal clearly in the forefront. 

In my next post, we will consider at a simple example to better understand system's thinking. 

References:

[1] Ross D. Arnold and Jon P. Wade, A Definition of Systems Thinking: A Systems Approach, 2015 Conference on Systems Engineering Research

The Interference Diagram Part 3

In my last post I demonstrated how to construct an Interference Diagram as well as creating a pie chart to summarize our finished results.  In this post we will continue our analysis and complete our discussion on the Interference Diagram.

So, if we want more Machine Output what can we do with the information in this analysis.  The question to ask is this:  “Are there any of these "interferences" that the operator can stop doing to free up more time on the machine? “Any interference you stop doing will free up more time to generate more machine output.  If you look you’ll notice that lunch and break time is not something that will go away, nor should it.  However, you might consider a person to run the machine while the operator is on lunch and breaks.  Not everyone needs to take lunch and breaks at the same time.   A staggered schedule could free up a person to operate the machine.  This simple solution could gain and hour of production time.

There are other possibilities to consider.  What about moving jobs into and out of the work area?  Should the operator be the person doing that?  Perhaps, this could be off-loaded to someone else and the operator could spend more time at the machine.  In this case, it would be worth and additional 9% increase in machine time. Also, what about looking for paperwork?  Maybe the paperwork needs to show up when the job does so the operators don’t have to look for it.  If it did, that’s another 19% increase in machine time.  By removing or reducing just three of the interferences it is possible to increase machine output by 28%.

The Interference Diagram is a great tool to find and exploit the hidden capacity of a constraint operation. In this example it’s a tool to help you analyze how you can get more from the constraint.  To verbalize out what those things are that the constraint could, or should, stop doing to free up more time to do what you want more of.

OK, so now that you’ve identified the interferences for more output, what’s next?  Many times simple fixes work very well, so let’s see how many simple fixes we can come up with and their impact on these obvious examples of waste.  To refresh your memory, here is the completed ID.

What’s a simple solution for interference #1, Looking for Paperwork?  What if we didn’t consider the job ready to work until all of the paperwork was assembled by someone else.  Couldn’t we eliminate all 1.5 hours of searching?  How about interference # 2, looking for parts?  What if we implemented a “full kit checklist” to be assembled by someone other than the machine operator?  In other words, the job isn’t moved into the work area until all of the parts, paperwork and special tools from supply were gathered up and sent  to the work area as part of the next job.  Wouldn’t that eliminate interferences 1, 2 and 7, effectively eliminating 90 plus 50 plus 20 minutes of wasted time?  Think about it, 160 more minutes of capacity without spending any money.

How about interference #4, looking for the supervisor?  What if a visual indicator system such as andon lights were installed to notify the supervisor of a need at the work station? Green would be seen as no problems, yellow might mean that the operator has a pending need and red might mean that the work area is shut down until the need is taken care of.  This kind of system wouldn’t eliminate all of the time, but let’s say 20 of the 45 minutes? Simple solution to gain an additional 20 minutes.

Finally, the movement of jobs in and out of the work area, interference #’s 5 and 6.  Just by using the andon lights (yellow) that we already installed, someone else could be used to position new jobs and move completed jobs. Again, you might not eliminate all of the wasted time, but let’s say you were able to eliminate half of it or 23 minutes.  So in this example, implementing simple solutions, we might be able to eliminate roughly 203 minutes which could all be used to produce more output.

In my next post, we will begin a new blog post series on a completely new subject.

Bob Sproull

The Interference Diagram Part 2

In Part 1 I presented the basic structure of an Interference Diagram (ID).  In this post I will present an actual example of an ID and how to construct one.

Building an Example

With these simple steps in mind let’s use this technique to dissect a situation and determine the interferences.  Let’s assume our example, for the sake of discussion, is using the ID to supplement additional systems thinking that has already occurred.  In this example we’ll assume that you have already conducted a preliminary systems thinking analysis using Goldratt’s Five Focusing steps.  In your analysis you have concluded where the constraint in your system resides. You have completed Step 1 of the Five Focusing Steps (Find the constraint).  You’ve determined that the constraint is a drilling machine that drills some very specific holes, in specific locations, and sometimes the holes are drilled at some odd angles.

There is also an assembly step that takes place at this location.  To drill some of the holes the operator must assemble the part he has, plus an additional part before he can drill the holes.  The additional parts come from another line in your plant.  The output from this machine seems too low.  There is sometimes a blockage of parts in front of it, and starvation for parts behind it.  It appears to have the classic characteristics of a system constraint.  The operator of this machine seems to be busy all the time, and yet the output is low.  Step 2 of the Five Focusing Steps states to “exploit” the constraint – to get more output.  To discover “why” the output is so low you decide to create the interference diagram.   You want to improve the output from this work station, so you ask yourself “What do I want more of?”  The answer is “More machine output.”  So, on a piece o paper or a white board write the following:

So, now you have verbalized what you want more of.  Next, you want to discover what are the interferences or those things that block the operator/machine from getting more machine output?  The first thing you notice is the amount of time the operator spends “Looking for the paperwork!”  We will assume in this scenario that a day is 8 hours in length, so we’ll base our analysis on 8 hours of available time.  You have a quick discussion with the operator to gain more information.  You add the interference to your diagram with the time estimation (some days may be more and some days may be less) of how much time the operator spends every day looking for the correct paperwork.  In this instance looking for paperwork is about 1.5 hours per day.

 Now, ask the question again - “What stops you from getting more of what you want?”  The operator offers an additional interference about how much time he has to spend looking for the additional parts he needs to do the assembly before drilling.  Add this interference to the diagram - “Looking for the parts I need”.  Now, add this as a second interference  with a quantified time estimate.

You ask the operator “are there any others times you have to leave the machine?”  He tells you, “Yes!”  He leaves the machine during lunch and the morning and afternoon breaks.  Lunch is 30 minutes and each break is 15 minutes long, so the total time is 60 minutes.  This time counts as time away from the machine because machine output stops when the operator isn’t there. So add this as interference #3 with the correct time.

Through additional conversations and observations you are able to identify additional items that take time away from machine output.  Add these additional interferences to the diagram with their estimated times.  The figure below is the completed Interference Diagram.

Chances are good that most of these interferences are caused by some type of policy constraint, or procedural constraint.  Another way to enhance the analysis is to paint a picture of how big the interference impact really is.  You can create an impactful picture by transferring this information into an excel format.  You can create a pie chart that paints the visual impact for all of the interferences. The figure below displays the pie chart for this interference analysis.

In my next post, we will discuss how to use the pie chart in this effort.

Bob Sproull