Q: I work for a not-for-profit children’s treatment center. Our organization is implementing a new service delivery model, and in this redevelopment process, I am involved with redesigning the client process workflow and information flow.
These two process flowcharts give a broad picture of how the client enters the organization, receives service and is discharged. Information flow gives us an idea of such activities as how the paperwork is flowing and appointments are booked.
These two overarching flowcharts have related flowcharts to explain the process in detail, what is happening in each step and the staff involved. So, we have a macro picture or overview, and then a detailed view.
As a result, I’m looking at a lot of different Visio charts and can’t figure out how to present this to the staff. If you have any suggestions or know of software I can use, it would be very helpful.
A: In the lean enterprise world, this is called material and information flows. As you pointed out, material flow—or process workflow as you described it—maps product movement through the value stream.
Information flow, as the name implies, maps the flow of information involved in the process and illustrates only what the process needs and when it needs it. In our world of information overload, information flow is intended to provide the right information in the right quantity at the right time.
Value stream mapping is a high-level approach that includes material flow and information flow in the same map to provide a big picture. A value stream map is all the actions—customer value added and business value added—required to transform raw materials into a finished product and tracks those actions from the start to delivery to the customer.
You mentioned you are looking at a lot of different Visio charts. This could be because you’re trying to include too many processes.
Although your question is about efficiently displaying the detailed process flowcharts to your staff, you should start by asking yourself if you really need that many processes to create value for your customers.
While I am a proponent of recognizing gaps and documenting processes, I also challenge myself to evaluate whether a process is necessary in the value stream. Ask your team a couple questions:
- Does this process offer value to the customer?
- Is the customer willing to pay for this process?
- Is this a process created for legacy systems to work around limitations of those systems or justify past headcounts?
Too often, we don’t challenge ourselves by asking the hard questions regarding why a certain process exists.
Your project seems to involve mapping the current material and information flow. Brainstorm with a cross-functional team and implement an efficient overall value stream that is lean and provides high-quality services to your customers in less time and with reduced costs. If you have already done this part, you should not have too many processes.
It’s also important to remember that processes should not have too many steps.1 Processes become increasingly complex due to multiple decision loops. If you can reduce the number of decision loops by simplifying the process, it will help as you try to construct your flowcharts.2
I have not come across any flowchart-mapping software that allows the user to easily construct multiple processes and the interrelationships between those processes without needing to switch screens, enlarge the views and cross-reference junction points between charts.
But there are practical ways to display high-level process maps and the underlying detailed flowcharts using simple PowerPoint presentations. The first thing I would suggest is presenting the overall process from the 50,000-foot level all the way down to the 1,000-foot level.
Imagine looking out an airplane window from the highest altitude to the point of landing. At the highest altitude, the only things you see are major landmarks—stadiums, mountains and rivers. As you get closer to the 1,000-foot level, you start to see the highways, moving cars and buildings.
The 50,000-foot level presentation involves five to seven high-level processes. As you transition to the 30,000, 10,000 and 1,000-foot levels, present the multiple process steps of the sublevel processes.
This approach will make it easier to present the information to a variety of audiences—from management to administrative support personnel—without needing to customize the material.
I would also suggest presenting the process using the suppliers, inputs, process, outputs and customers (SIPOC) approach for the major processes. This will help simplify the process for better understanding and can be presented separately in PowerPoint slides.
If you are going to present this information as a single process flow, "gray out" the process steps using PowerPoint’s coloring options so the only thing that stands out is the portion of the process you want the audience to focus on at that point in the presentation. Continue this approach until you finish presenting all the flow paths within the chart.
This approach is particularly useful when you’re presenting a complex chart with multiple decision loops. In that case, it can be difficult for the audience to comprehend the individual flow paths. This approach will reduce the distractions and ensure everyone understands the process from beginning to end.
Senior manager, quality systems
San Jose, CA
For More Information
- Boudreaux, Miriam, "On the Map," Quality Progress, November 2010, pp. 30-35.
- Manos, Tony, "Value Stream Mapping—An Introduction," Quality
Progress, June 2006, pp. 64-69.
Q: Normally we perform control charting on product quality characteristics. Is it possible to do the same thing for continuous process variables, such as temperature and pressure? If so, how do I determine the sampling frequency and subgroup size of the sample? Do we need to consider the autocorrelation of the data points in this case?
A: Yes, it’s possible to control chart continuous variables such as pressure or temperature. Typically, you would use an individual control chart with a moving range chart.
There wouldn’t be any subgrouping because the variance between elements of a subgroup for any point in time would be so tiny that the number of out-of-control points would be unusually high if subgrouping was used to calculate the process variance.
There would be some autocorrelation if the individual points are too close in time, but some experimentation would be needed to find the optimum spacing. The way to do that is through an autocorrelation plot using a statistics package and looking at lags between data points.
Also, the sampling would need to be representative, so it’s important to place the sensor where you want the most information. For example, you may wish to track a hot spot, a cool spot or an area where the temperature is homogenous. That depends on the research question.
But control charts can be incredibly useful in situations that may not occur to you immediately. For example, I use individual control charts to track gas mileage on my car. It works pretty well and makes it easy to tell how my driving habits affect performance.
The only caveat with individual control charts is that the average run length to find an out-of-control situation is typically much longer. In other words, it may take a longer time to detect an abnormality.
Packaging quality specialist