Training courses Consultancy Engineering tools The way we work
Tolerance design

Statistical Process Control

In most industries it is not realistic to allow a process to make bad parts which are then removed by inspection. Statistical Process Control is a family of techniques which are used in problem solving to improve the process and then in routine production to maintain a high level of performance. It may lack some of the glamour of more fashionable methods but has stood the test of time and can give major benefits in return for a small amount of effort.
Control charts are the core of Statistical Process Control. They are a way of recording data which shows whether a process is stable, if a trend is developing or if an unusual event has occurred. Analysis of existing charts gives valuable insights into the likely causes of problems.
The methods are built on statistical foundations but do not require deep, mathematical understanding for routine use. Simple charts for "standard" situations are easy to prepare and very easy to employ. The skill comes in designing a chart which balances statistics with knowledge of the process to provide a practical tool in unusual circumstances.

Course information

The concepts together with the basic skills for initial applications can be introduced in a two-day training course. To gain the greatest benefit, though, it will be necessary to add further specialised methods which are relevant for your industry. It is here that a customised training course is vital since the range of techniques is so broad : to cover them all would take more than a week.

A two day course is aimed at engineers and managers in manufacturing, production engineering and quality functions. The content does include some specialised topics, but if more than a few are needed then the course will extend to three days. An alternative is to run a support workshop a few months after the initial training. Training courses for operators concentrate on the practical aspects of SPC and avoid statistics. In large organisations it may be better to train in-house trainers for this task.

Course content
The course is a blend of teaching and exercises which are tackled by small teams. The content depends on the needs of the organisation and can include the following :
• Basic techniques for charts for variables
• Basic techniques for attribute charts
• Sampling plans
• Control and capability
• Capability studies and indices
• Capability for non-normal distributions
• Short run or small batch techniques
• Charts for individual measurements
• Processes where the average drifts with time
• Processes which produce several parts simultaneously
( multiple fixtures, cavities or extruders )
• Cumulative Sum ( CUSUM ) methods
• Management issues

Expected results
By the end of the course delegates will be able to apply the techniques to their processes.