Lecture The management and control of quality - Chapter 14: Statistical process control

Lecture The management and control of quality - Chapter 14: Statistical process control. This chapter presents the following content: Statistical quality control, statistical process control, types of variations, causes of variation, assignable causes,...

Statistical  Quality Control

Statistical  Process Control

 Statistical process control  is the 

application of statistical techniques to  determine whether a process is 

delivering what the customer wants.

 Acceptance sampling  is the 

application of statistical techniques to  determine whether a quantity of 

material should be accepted or rejected  based on the inspection or test of a 

sample.

Variation from Common  Causes

Variation from Special Causes

 Two basic categories of variation in output  include common causes and assignable 

 Common causes are the purely random, 

unidentifiable sources of variation that are  unavoidable with the current process.

– If process variability results solely from 

common causes of variation, a typical 

assumption is that the distribution is symmetric,  with most observations near the center.

 Assignable causes of variation are any 

variation­causing factors that can be 

identified and eliminated, such as a machine  needing repair.

 The red distribution line below indicates that the process produced a preponderance of the tests in less than average time Such a distribution

is skewed, or no longer symmetric to the average value.

A process is said to be in statistical control when the location, spread,

or shape of its distribution does not change over time

 After the process is in statistical control, managers use SPC procedures

to detect the onset of assignable causes so that they can be eliminated.

Location Spread Shape

© 2007 Pearson Education

Capability and Control

Process capability calculations make little sense if the process is not in statistical 

control because the data are confounded 

by special causes that do not represent the inherent capability of the process

Capability Versus Control

Control Capability

Capable

Not Capable

In Control Out of Control

IDEAL

necessary

Interpreting Control Charts

When a process is in statistical control, the points on a control chart fluctuate 

randomly between the control limits with 

no recognizable pattern

Shift in Process Average

Identifying Potential Shifts

Cycles

Trend

– Continue to collect and plot data – Take corrective action when 

necessary

Process Monitoring and Control

Control charts indicate when to take action, and more importantly, when to leave a process alone

  Spreadsheet Template 

Special Variables Control Charts

Charts for Individuals

Control charts for individuals offer the advantage of being able to draw 

specifications on the chart for direct comparison with the control limits

Choosing between C­ & U­charts

Confusion often exists over which chart 

is appropriate for a specific application, because the c­ and u­charts apply to 

situations in which the quality 

characteristics inspected do not 

necessarily come from discrete units

Control Chart Formulas

x and s

x and R no

yes

constant sample size?

p-chart with variable sample size

no

p or np

sampling unit?

c u

In determining the method of sampling, samples should be chosen to be as 

homogeneous as possible so that each sample reflects the system of common 

causes or assignable causes that may be present at that point in time

In practice, samples of about five have been found to work well in detecting 

process shifts of two standard deviations 

or larger. To detect smaller shifts in the process mean, larger sample sizes of 15 

to 25 must be used

Economic Tradeoffs

Pre­Control

nominal value

Green Zone

Yellow Zones

Red Zone

Red Zone

Pre­control is not an adequate substitute for control charts and should only be