Tools and techniques are an important part of process improvement and continuous quality improvement. In this guide, we explore some of the charts, maps, diagrams and analyses used in process and quality improvement.
Control charts are used to plot each process within an operation in chronological order, allowing for the evaluation of current processes and any changes that are made.
The control chart also allows for identification of variation in a process or individual product batches. A control chart helps you quickly see where variation is happening, such as when different shifts are on duty in a manufacturing operation. This can lead to finding “special causes” – that is, the factors impacting a process. This gives a focus for where improvement measures should be taken and the means to then measure whether those measures have led to success.
Control charts typically have an upper control limit (UCL) and lower control limit (LCL), each three standard deviations from the mean. It’s worth noting that a control chart sets control limits, not specification limits. In other words, a process can be in control (within the UCL and LCL) yet still not meet customer specifications.
X-Bar R and S charts
Both of these charts plot a process over time. There are differences in how each approaches the task of plotting the “center line” between the UCL and LCL.
An x-bar R chart uses the average of ranges in the dataset, defining that range by using the difference between two data points. If one data point is 10 and the other 5, then the plotted value on a X-bar chart would be 5.
The X-bar S chart uses standard deviation. Each data point consists of the average of values within a subgroup. For example, if a subgroup consisted of 4, 7, 10 and 12, then the plotted point would be 8.25.
The ImR chart detects shifts in variation. ImR charts are best used when measurements are expensive or not necessarily related, when products require a long cycle time to produce, when each sample represents an entire batch and when output is continuous and uniform.
P charts are used specifically to monitor the percentage of nonconforming units in a process. Each part produced is inspected and determined to be “good or bad.” The percentage (or proportion) of nonconforming units in each subgroup becomes the plotted point on the chart.
A flowchart provides a visual representation of the flow of work during an operation, breaking it down into a series of events or processes. The level of detail depends on the type of chart used, but the overall goal is to analyze and then optimize workflow. Flowcharts may also identify who is responsible for doing each task in a process.
In process improvement, there are a variety of graphical charts that are used depending on the situation. They include the following popular choices:
Dot chart. Data points are plotted on a simple scale.
Histogram. Data is organized into user-specific ranges, such as age or gender groups.
Run chart. Data is displayed in a time sequence.
Pareto diagram: Data is organized in such a way as to quickly recognize the “vital few” factors of a process.
Scatter diagram: Data is plotted on horizontal and vertical axes to identify association between explanatory and response variables. An example would be level of educational attainment (explanatory) and annual salary (response).
Box plot: Numerical data groups are represented through their quartiles. This allows for quickly spotting outliers.
A spaghetti map traces an item – and sometimes a person – as they travel through each part of a process. Typically, it depicts a floor map, such as a manufacturing or office floor. The map reveals the flow of a process by tracking one item, giving teams a clear picture of where there are redundancies or other problems in a process.
SIPOC stands for supplier, input, process, output and customer. This diagram is to clarify core processes, especially in terms of who supplies inputs, the specifications placed on the inputs, identifying the customers and determining the requirements of the customers. A SIPOC diagram quickly and easily identifies the key players and what a project accomplishes.
High-Level Process Map
A high-level process map provides the major data points on how a process works and what it gets done. It does not go into the details of how a process works. This map is used to offer a quick overview of a process for those who do not need to know every detail, just the key players, the major components and what it accomplishes. This essentially takes the “process” part of a SIPOC Diagram and expands upon it.
Swimlane Process Map
A swimlane process map shows the details of a process, this time presented as lanes that offer details on the tasks accomplished by each specific entity (purchasing, supplier, manufacturing, shipping, etc.). This is a frequently used map because it clearly shows who is responsible for what over the entire course of a process.
Detailed Process Map
A detailed process map provides a breakdown of the details of any process within an overall operation. The inputs and key players for that process are established, and then a step-by-step process is mapped out until the output is achieved.
Relationship Process Map
The relationship process map focuses on the key people in a process and how information and materials flow between them. This is another high-level map that shows the major components involved in the process flow, but not the details of what is done.
Value Stream Map
Value stream maps go into great detail in mapping the flow of a project. It includes the people involved, the tasks accomplished, how the process flows from task to task, and the Takt Time (the time between the start of each product batch). Everything in a value stream map is analyzed through the lens of how it impacts the end user or customer.
This is the diagram used to brainstorm for process improvement projects. It supports teams in finding the root causes for errors and variations in a process. Typically, to find root causes, an operation is broken down into separate processes, which then in turn are divided into their component parts. These areas are typically defined as people, program, product, policy and place.
ANOVA (Analysis of Variance)
ANOVA is an analysis that helps determine whether variance in a project is due to random noise or a significant issue. The advantage of ANOVA is that it allows for measuring the impact of more than one factor as well as two or more factors in combination, rather than analyzing two samples of the same factor.
Failure Mode and Effects Analysis
This risk analysis tool can be used at any point in a process improvement project to determine the possibility of risk and allow you to make changes that prevent the problem from happening. First developed by the military, it provides a step-by-step process for finding any possible failures in the design, development, manufacturing and delivery of a product or service.
This analysis is applied at the beginning, middle and end of a project – at any point where a team either sees the potential for substantial risk or wants to measure the possibility and stay ahead of problems.
The “gap” in gap analysis is the measurement between the way things currently are and the way you want them to be. The analysis determines the exact size and nature of the gap, and then allows a project team the ability to find ways to bridge that gap. For example, a company might look at whether its performance is meeting the standard that has been set. If not, ways are found to bridge the gap between current and expected performance.
This has become one of the most popular tools in process improvement. SWOT stands for strengths, weaknesses, opportunities and threats. A SWOT analysis can be used in many different situations but is most often employed as part of strategic planning or project management. A typical SWOT analysis employs a four-block diagram where information is listed on internal issues (strengths and weaknesses) and external factors (opportunities and potential threats). SWOT only works if the team involved is honest, detailed and realistic about expectations.
The Lean methodology places high value on analyzing every process through the lens of what does and does not serve the customer. If it serves the customer, it has value and should be optimized. If it does not have value for the customer, it should be eliminated as a wasted step in the process. This is the focus of waste analysis – mapping out a process in detail and then eliminating anything that does not serve the customer. There are 8 Wastes of Lean:
Measurement System Analysis
All the charts, maps, diagrams and analytical approaches in Lean and Six Sigma are designed to measure variation within a process. But what about the measurement system itself? A measurement system analysis uses advanced mathematical formulas to determine the variance in the method being used to measure or analyze a process. This makes it easier to determine how much measurement system variance is contributing to overall system variance.