Parts produced by the same process and with similar conditions are never 100% identical.   There will always exist part-to-part variations.   Process Improvement is about reducing these part-to-part variations within acceptable tolerances.   It is accepted fact that reduction of process variations will result in improved part quality and less non-conforming parts being produced.

Statistical Process Control (SPC) is an effective tool to use to evaluate and monitor process variations.   There are two types of process variation which are known as Special Cause and Common Cause.   Special Cause are not natural variations seen in processes and Common Cause are the normal variations or noise that all processes will have.  Special Causes can be reduced, if not eliminated, by some adjustments and re-alignment of the process equipment settings and material.   On the other hand Common Causes are more challenging to reduce or eliminate.  It will require process modifications or major refitting to address Common Causes.

The initial step in SPC is to gather quantified data at predetermined time frequency and then plotting the gathered data on a graphical template called a control chart. The plotted data are used to calculate a set of control limits lines that are then used on the control chart.    These control limits lines then act as boundaries in determining whether the process is stable and within acceptable variation limits.   A 6 sigma limit range is typically used.  (+3sigma limit, -3sigma limit).  Through known special variations tend-markers the process is improved by eliminating/reducing these special causes that causes those special variation trends.  The process is considered within control when all the data and future plotted data are all within the control limits and the plotted data does not exhibit uncontrolled trend-markers.

The use of Process Capability is to identify and reduce common causes in the process.   This is done by calculating the mean and sigma of the process and trying to center the process production distribution within the part specification limits.   Process Capability also allows you to determine the quality or parts per million (ppm) of the process; which forms a key part of a Six Sigma study approach.    Process Capability should only be determined once the process Special Causes are found to be controlled.

Sub-processes and the machines within a process will require some adjustments/actions for a process to produce within its control limits over time.   Control Charts are used to monitor a process so that immediate remedial actions may be taken when special causes occur.   Typical special causes can be attributed to:

  1. Change in machine settings; eg, temperature, pressure, timing, etc
  2. Change in machine part; eg tool, die, fixtures,etc.
  3. Change in material or component of part being produced; raw material, new part supplier, etc.
  4. Change in operational procedure; eg. new operator, new production method, etc.

There are various types of SPC control charts.  Commonly used SPC charts are the Variable Control Charts and Attribute Control Charts.  Some of them are listed below:

Variable Control Charts:

  • Average & Range
  • Average & Std Deviation
  • Individual & Moving Range
  • Median

Attribute Control Charts

  • p
  • np
  • c
  • u