JJBlack & Associates, Inc.

Variability the Hidden Waste

There are basically only two ways an organization can take away from their bottom line. The first is lack of capital goods, which is the easier of the two to solve. This can be lack of: tools, people, methods, supplies, infrastructure, space, knowledge, etc. Generally the organization can put together a business plan to solve the first, but the second is much harder to see even though it is typically in front of you all of the time.

Variability is the second way to take away your profits, it is truly a form of waste (muda). It is the more hidden of the two methods and typically present in almost every process in our operating system, as we have observed.

We built a simple case study to illustrate our point by using a simulation model; we call it our Variability Model.

Now consider a simple manufacturing system consisting of two independent five station work cells. Within each cell parts progress through each of the five machines in a sequential manner; every part was move between all five machines. Between each machine is a part queuing stand where the downstream machine puts completed parts and the upstream machine pull in raw parts for their operation. The queue stands are referred to as a Kanban system where the operator cannot put more parts in the area unless it has fewer parts than the Kanban limit, which we have set to one for this example Each of the five machine cells could be allowed to operate in several ways, we will consider two ways.

For the top five station cell the owner allows the work teams, that make up the cell, to vary their process time at each machine within a given variance. The bottom five station cell has each worker always complete their work at a constant cycle time.

The base cycle time is 2 minutes and it takes 10 seconds for an operator to either pick up a new part or deposit a finished part on the Kanban stand. Each employee makes $10.00 hr and the finished parts can be sold for $5.00 each. We will run the model with 3 scenarios by allowing the top cell to vary their cycle time from 10% to 30%. We set the cycle time variance up with a simple triangular distribution, which means the cycle time could vary from a minimum to maximum cycle time as the table shows below:

To ensure that we have some confidence in our results we run the model for 52 weeks of 40 hours each to arrive at an average loss per week, over a year's time. In addition, we added a 5 hour warm up period to each run to remove initial bias from start up, meaning we don't count the first 5 hours to ensure that the system is running at steady state when we begin collecting data.

Now this is a very simple system with only 5 individual stations and considering 2 different work strategies. Cell 1's team is told they have some flexibility in their work pattern and that they should always try to meet their target of 2 minutes to complete each task. Therefore, they most likely get each job done in 2 minutes, however sometimes they will work a little faster and get the job done in the minimum time and sometimes it might take them the maximum time, again depending upon the amount of variance allowed by scenario.

Let's look at what happens after we average 52 weeks of 40 hours per week worth of production for each scenario. The average weekly throughput goes down as variability increases!

From the table above we can see that Variability has raised its ugly head and cost the business from $940 to over $6,000 per year depending upon the amount of variance in the cycle time. That is a very significant amount when you think that most companies have many more than 5 employees, and many more than one process.

As I mentioned before, variance sneaks into every business in many ways. Consider some of the following examples and you can surely see why every business should try to eliminate it from their business processes:

• An Engineering firm utilizing a number of Project Managers doing their own work, with each allowed to do their work in their own way - the throughput of each manager will vary based upon their individual methodologies and could even worsen if the business is departmentalized.
• Preventive Maintenance levels varying on each machine in the production system - up time of the machine will vary and most likely increase as their age increase.
• Office clerks which have different filing systems - other not familiar with their system will spend extra time finding needed materials.
• Employees allowed to keep important documents on their own company hard drives - other will not have immediate accessibility causing increased time.
• Having ISO business practices installed which utilizes a great deal of forms which must be completed and signed - often employees wait until someone forces them to fill out the forms, or the time to get them signed may have to wait for the availability of certain individuals to be present or willing causing increased process time.
• I'm sure that you can think of many more examples ...

Through our application of Lean and Six Sigma tools JJBlack & Associates, Inc. can help your business or organization find and remove the variation waste in your system. Please contact us for more information.

Support Our Sponsors