Statistical

Process
Control
Learning Objective
• Define the concept of statistical process
control (SPC).
• Explain the rationale for SPC.
• List the steps for developing a control
chart.
• Describe management’s role in SPC.
• Explain the rationale for using the quality
tools before employing a control chart.
Learning Objective
• Summarize the authority operators who
use SPC have over processes.
• List the three broad phases of the
process for implementing/deploying SPC.
• List the factors that are the worst
inhibitors of SPC.
Statistical Process Control Defined
Statistical process control (SPC) is a statistical
method of separating variation resulting from
special causes from variation resulting from
natural causes in order to eliminate the special
causes and to establish and maintain
consistency in the process, enabling process
improvement.
 Rationale for SPC
The rationale for SPC is much the same as that
for total quality. It should not be surprising that
the parallel exists because it was Walter
Shewhart’s work that inspired the Japanese to
invite W. Edwards Deming to help them get
started in their quality program in 1949 to 1950.
SPC was the seed from which the Japanese grew
total quality.
 Rationale: Control of Variation
The output of a process that is operating properly can be
graphed as a bell-shaped curve. The horizontal x-axis represents
some measurement, such as weight or dimension, and the
vertical y-axis represents the frequency count of the
measurements, that is, the number of times that particular
measurement value is repeated. The desired measurement value
is at the center of the curve, and any variation from the desired
value results in displacement to the left or right of the center of
the bell. With no special causes acting on the process, 99.73% of
the process output will be between the {3s limits. (This is not a
specification limit, which may be tighter or looser.) This degree
of variation about the center is the result of natural causes. The
process will be consistent at this performance level as long as it
is free of special causes of variation.
Frequency Distribution Curve:
Normal Curve
Frequency Distribution Curve
Bimodal Frequency Distribution Curve
Frequency Distribution Curve:
Narrowed (Less Variation)
Rationale: Continual Improvement
Continual improvement is a key element of total quality.
One talks about improvement of products, whatever they
may be. In most cases, it would be more accurate to talk
about continual improvement in terms of processes than
in terms of products and services. It is usually the
improvement of processes that yields improved products
and services. Those processes can reside in the
engineering department, where the design process may
be improved by adding concurrent engineering and
design-for-manufacture techniques, or in the public
sector, where customer satisfaction becomes a primary
consideration.
 Rationale: Predictability of Processes
The problem is unpredictable processes. If the
same customer had approached a firm that was
versed in SPC, the results would have been
different. The managers would have known with
certainty their capability, and it would have
been clear whether the customer’s
requirements could, or could not, be met. They
would know because their processes are under
control, repeatable, and predictable.
 Rationale: Elimination of Waste
Parts that do not meet specifications are waste. To
prevent defective products from going to
customers, more is spent on inspection and
reinspection. This, too, is waste. All of these
situations are the result of some process not
producing what was expected. In most cases, waste
results from processes being out of control;
processes are adversely influenced by special
causes of variation. Occasionally, even processes
that have no special causes acting on them are
simply not capable of producing the expected
result.
 Rationale: Product Inspection
It is normal practice to inspect products as they
are being manufactured (in-process inspection)
and as finished goods (final inspection).
Inspection requires the employment of highly
skilled engineers and technicians, equipment
that can be very expensive, factory space, and
time. If it were possible to reduce the amount of
inspection required, while maintaining or even
improving the quality of products, money could
be saved and competitiveness enhanced.
 Rationale: Product Inspection
This also applies internally. When a company’s
processes are determined to be capable of
producing acceptable products, and after they
are in control using SPC, the internal quality
assurance organization can reduce its inspection
and process surveillance efforts, relying to a
greater degree on a planned program of process
audits. This reduces quality assurance costs and,
with it, the cost of quality.
Capability
Index
(Cpk) Capability index (Cpk) shows how
well parts being produced fit into
design limit specifications.

Cpk >= 1 is capable ; Cpk < 1 is not capable

Capability
Also, useful to calculate
Index
probabilities and defect
(Cpk)

You can use the normal distribution table (Microsoft

Excel with NORMSDIST function) to find out the
probability.
Process Capability
 Exercise 1

let’s assume our process is centered at 1.251 and σ =

0.00083 (σ is the symbol for standard deviation).
• Capable or not?
• How many defects?
 Solution
The second method is used when the process
average is not assumed to be coincident with
the specification average and is denoted as Cpk

Cpk = min[((1.251−1.245)/ 3(0.00083)) or

((1.255−1.251)/ 3(0.00083))]
Cpk = min [2.4 or 1.6]
This is Cpk >= 1 ➔ Capable
 Solution

Using Microsoft Excel

NORMSDIST(−7.2289) =2.43461E-13 =
0.00000000000024361 and
1-NORMSDIST(4.8193) =1-0.99999928 =
0.00000072
 Solution
This fraction defective above the upper spec is
1−0.99999928 = 0.00000072 of our parts.
Adding these two fraction defective numbers
together we get 0.00000000000024361. We can
interpret this to mean that we only expect about
0.72 parts per million to be defective. Clearly,
this is a great process. You will discover as you
work the problems at the end of the chapter
that this is not always the case.
 Exercise 2
The quality assurance manager is assessing the
capability of a process that puts pressurized grease in
an aerosol can. The design specifications call for an
average of 60 pounds per square inch (psi) of pressure
in each can with an upper specification limit of 65 psi
and a lower specification limit of 55 psi. A sample is
taken from production and it is found that the cans
average 61 psi with a standard deviation of 2 psi.
• What is the capability of the process?
• What is the probability of producing a defect?
Solution
Solution
Six Sigma
Western Electric Rules
Exercise
Control or
Out of
Control?
 Solution
• Mean = 10,17
• Standard Deviation = 0,23
• Zona C1 = Mean + SD = 10,40
• Zona C2 = Mean – SD = 9,94
• Zona B1 = Mean + 2SD = 10,63
• Zona B2 = Mean – 2SD = 9,71
• Zona A1 = Mean + 3SD = 10,86
• Zona A2 = Mean – 3SD = 9,48
SPC Control or Out of Control?

A
B
C
C
B
A
SPC is Out of Control

A
B
C
C
B
A
 Control Chart Development
Just as there must be many different processes, so
must there be many types of control charts. Lists
the seven most commonly used control chart types.
You will note that the first three are associated with
measured values or variables data. The other four
are used with counted values or attributes data. It is
important, as the first step in developing your
control chart, to select the chart type that is
appropriate for your data. The specific steps in
developing control charts are different for variables
data than for attributes data.
Common Control Charts and Their
Applications
X- and R-Charts
Factors Table for x and R-Charts
X-Charts
X-Charts
R-Charts
 R
R-Charts
X- and R-Charts
Chart for an Unstable Process.
The Process Is in Statistical Control.
p-Chart
p-Chart Formulas
p-Charts
Subgroup 1 Subgroup 2
p-Chart Control Limit Calculations
p-Chart.
c-Chart Data
c-Chart Formulas
c-Charts
c-Chart: Power Supply Defects
In Control or Out of Control
Succession of Control Charts
In Control and Capable Are Not the
Same Thing
Thank You