Statistical Process Control

Charts
QA-LAB
 Statistical Process Control (SPC)

Statistical Process Control or SPC (sometimes called

quality control), is a set of methods used in industry for the
management of product quality.

There are two general approaches to quality management.

•The inspection approach (i.e. inspect the final product and
send defects back for re-work or to scrap), and
•The prevention approach which focuses on the
production process(versus the final product). This allows
for adjustments (corrections) to the PROCESS before
defects are produced.
Goal - Prevention
• Monitor behaviour of process using measurements,
to determine if operation is statistically stable
• All production processes result in variation, hence:
all products will exhibit some degree of variation.
• There are two sources of variation:
– Random variation
– Non Random variation
Variability

• Random • Non-Random
– common causes – special causes
– due to identifiable factors
– inherent in a process
– can be modified through
– can be eliminated operator or management
only through action
improvements in the – Ex. Periodic calibration
system of instruments
– Ex. Blackout of
electricity
Basic Principles
The Concept of Control
Out of Control
Out of Control
Out of Control
 Control Charts
• A graph that establishes
control limits of a • Types of charts
process
• Control limits – Attributes
– upper and lower bands of » p-chart
a control chart » c-chart
– Variables
» range (R-chart)
» mean (x bar – chart)

Copyright 2006 John Wiley & Sons, Inc. 4-12

Quality Measures
• Attribute/qualitative
– Attribute data are counted and cannot have fractions or
decimals. Attribute data arise when you are determining only the
presence or absence of something: success or failure, accept or
reject, correct or not correct. For example, a report can have four
errors or five errors, but it cannot have four and a half errors.
Attribute charts require larger sample sizes - 50 to 100 parts in a sample

• Variable/quantitative
– a product characteristic that is continuous and can be measured
– For example: time, weight, distance or temperature can be
measured in fractions or decimals.
Variable charts require smaller samples - 2 to 10 parts in a sample

4-13
R-Charts
Monitor range to determine whether variability is stable.
Range provides an indication of dispersion, and is easy to calculate.
R

Always look at the Range chart first. The control limits on the X-bar
chart are derived from the average range, so if the Range chart is
out of control, then the control limits on the X-bar chart are
meaningless.
control limits - determined by looking at sampling properties of range computed
from observations
X-Bar Charts
X-bar charts are used for testing stability of the mean
operation
• calculate averages from samples of size n taken at
each time step
• centre-line - determined from either
» target or specification value
» average of sample averages for data set collected when
process was operating normally
» Control Limits - determined using the average range
during normal operation, which is a reflection of usual
process variability
Stopping Rules / Control Chart Pattern
One such set of guidelines are known as the Western Electric
Stopping Rules.

 8 consecutive points on one side of the center line

 8 consecutive points up or down across zones
 14 points alternating up or down
 2 out of 3 consecutive points in zone A but still inside the
control limits
 4 out of 5 consecutive points in zone A or B
Western Electric Stopping Rules
Western Electric Stopping Rules
Western Electric Stopping Rules
Western Electric Stopping Rules
Western Electric Stopping Rules
Western Electric Stopping Rules
Ishikawa / Fishbone Diagram
• Created by Japanese quality expert Kaoru Ishikawa.
• Helps to find the possible root cause of a problem in
a structured manner.
• However, this diagram only express hypotheses
about the causes of a problem or the reason why
something is not happening as desired.
• Therefore, it cannot replace empirical testing of the
hypotheses since it does not tell which is the real
cause of the problem.