SPC: Statistical Process Control

What is the Statistical Process Control

Joseph M. Juran briefly defines the statistical process control as “the application of statistical techniques to understand and analyze the variability of a process“.

It is, in conclusion, a methodology, applicable in particular to repetitive processes, which relies on statistical techniques aimed to define the process variability concept.

The purpose of this methodology is to provide information that can become valuable tools for improving the quality of a product, by limiting the most frequent errors (rework, slowdowns, and superfluous processes).

In the end, the statistical process control provides a deep knowledge of the processes with the aim of predicting their evolution so that they can intervene in real time.

Techniques and applications

Let’s start from the concept that the statistical techniques used by SPC, allow determining and explaining performances and causes of the unwanted changes compared to the normal functioning of the process under analysis.

These statistical techniques are:

• Pareto charts
• Control charts
• Cause and effect diagrams (by Ishikawa or fishbone)
• Action Plan

The possible applications are:

• Process controls verification
• Correlation analysis between processes
• Corrective actions
• Plants assessment

How statistical process control works (SPC)

The process manager can define the variables of the control charts, i.e. define the concept of out of control.

Every possible oscillation is described by the colors, which indicate the different compliance with the parameters established by the process manager.

The red areas indicate negative values that invalidate the quality of the product or other aspects of the production process.

Variables Definition

Given this configuration, we move on to the next step: data acquisition.

This happens in two ways: some data are acquired directly by the system; others are acquired manually by operator.

Data acquisition

For each variable it is possible to draw up the control chart with its values over time. It will therefore be possible to visually verify any out of control.

After creating the control charts, we will define the causes of out of control and the flow of corrective actions.

As we said before, one of these tools is the OCAP (Out of Control Action Plan), a flowchart that contains the series of operations to be followed in order to manage an out of control.

By using the Pareto chart we will be able to identify, among a series of causes, those that most affect the phenomenon under examination, and therefore, allow us to objectively evaluate the intervention priorities in solving problems.

Finally, the cause/effect diagram (or Ishikawa diagram) is used as a tool to identify the causes, i.e. the relations between a characteristic and its factors, and therefore the solution of the problem.

Process capacity analysis

Process capacity is also displayed through:

• the calculation of frequency histograms and Gaussian analysis,
• CP and CPK calculation
• average and standard deviations calculation
• the use of cause-effect diagrams and action plans