Unit-5

Statistical Process Control – Data collection

Control charts in quality control
Process capability
 STATISTICAL PROCESS CONTROL
• A method of inspection by which it can be determined whether a process is in control
• Differs from Acceptance Sampling in that SPC does not make judgments about the
quality of the item processed.
• Key tool is the Control Chart of which several types exist.

• All processes are affected by multiple factors and, therefore, SPC can be applied to any
process.
• There is inherent variation in any process which can be measured and “controlled.”
• SPC does not eliminate variation, but it does allow the user to track special cause
variation.
• “SPC is a statistical method of separating variation resulting from special causes from
natural variation and to establish and maintain consistency in the process, enabling
process improvement.” (Goetsch & Davis, 2003. p. 631)
 STATISTICAL PROCESS CONTROL
• Process: Set of interrelated resources and activities that transform inputs into outputs. [Note:
Resources may include personnel, finance, facilities, equipment, techniques, and methods
• State of statistical control: State in which the variations among the observed sampling results
can be attributed to a system of chance causes that does not appear to change with time. [Note:
Such a system of chance causes generally will behave as though the results are simple random
samples from the same population) (ANSI/ISO/ASQC A3534-2-1993, clause 3.1.5).]
• Process in control, stable process: Process in which each of the quality measures (e.g., the
average and variability or fraction nonconforming or average number of nonconformities of the
product or service) is in a state of statistical control (ANSI/ISO/ASQC A3534-2-1993, clause 3.1.6).
• Chance causes: Factors, generally many in number but each of relatively small importance,
contributing to variation that have not necessarily been identified. [Note: Chance causes are
sometimes referred to as common causes of variation (ANSI/ISO/ASQC A3534-2-1993, clause
3.1.9).]
 Variation in Processes

• Common Cause variation - the variation which in inherent in the process itself; when sampled,
a normal distribution is found; a process is said to be in statistical control when only common
cause variation exists.
• Special (or Assignable) Cause variation – the variation in process output that might be
traced to a specific cause; the process is said to be out of control when a special cause
variation exists.
 Rationale for SPC
• Control of Variation

• Continuous Improvement

• Predictability of Processes

• Elimination of Waste

• Product Inspection
 Creating Control Charts
• All control charts rely on the periodic sampling and measurement of items.
• The data collected will allow the calculation of a centerline, and upper and lower control limits.
• The centerline is the mean of all samples, whereas the control limits are, conceptually, the mean
+/- three standard deviations.

Interpreting Control Charts

SPC is based upon the Central Limit Theorem which tells us,
in effect, that the samples will follow a normal distribution
regardless of the shape of the parent distribution.
Interpreting control charts is, then, all about probabilities – if
the observations aren’t probable, then there must be a
special cause variation.
 Interpreting Control Charts

Special Cause Variation is assumed to exist if:

1. Any point falls outside the control limits.
2. Nine consecutive observations fall on one side of the mean.
3. Six consecutive observations are increasing (or decreasing.)
4. 14 observations alternate above and below the mean.
5. Two of three consecutive points fall in zone C in one-half of the chart.
6. Four of five consecutive points fall in zone B in one-half of the chart.
7. 15 consecutive observations in the A zones.
8. Eight consecutive points outside of the A zones.
 Risks of SPC

• SPC has the same Type I and Type II risks as

acceptance sampling
• If the process if in fact in control but we conclude that it
is out of control, we have committed a Type I error.
• If the process if in fact out of control but we conclude
that it is in control, we have committed a Type II error.
 Methods of Data Collection
(Population and sample)

• Census Survey

• Sample Survey

• Raw statistical data (recorded in the order)

✓ Univariate data

✓ Bivariate data

✓ Multivariate data
 Types of Data
Variables Attributes
 Characteristics that  Defect-related
can take any real characteristics
value  Classify products as
 May be in whole or either good or bad or
in fractional count defects
numbers  Categorical or
 Continuous random discrete random
variables variables
Common control charts for variables & attributes
 Control Charts for Variables

 For variables that have

continuous dimensions
 Weight, speed, length,
strength, etc.
 x-charts are to control
the central tendency of the process
 R-charts are to control the dispersion of the
process
 These two charts must be used together
 Mean and Range Charts
(a)
These (Sampling mean is
sampling shifting upward but
distributions range is consistent)
result in the
charts below

UCL
(x-chart detects
x-chart shift in central
tendency)
LCL
UCL
(R-chart does not
R-chart detect change in
mean)
LCL
Figure S6.5
 Mean and Range Charts
(b)
These
sampling (Sampling mean is
distributions constant but
result in the dispersion is
charts below increasing)

UCL
(x-chart does not
x-chart detect the increase in
dispersion)
LCL
UCL
(R-chart detects
R-chart increase in
dispersion)
LCL
Figure S6.5
Central Limit Theorem
Regardless of the distribution of the population, the
distribution of sample means drawn from the
population will tend to follow a normal curve

1. The mean of the sampling

distribution (x) will be the same as x=
the population mean 
2. The standard deviation of the
sampling distribution (x) will equal 
the population standard deviation x = n
() divided by the square root of the
sample size, n
Control Charts for Attributes
 For variables that are categorical
 Good/bad, yes/no,
acceptable/unacceptable
 Measurement is typically counting
defectives
 Charts may measure
 Percent defective (p-chart)
 Number of defects (c-chart)
 What SPC does not do?
• SPC only determines whether a process is in statistical
control or NOT, whether the process is producing within
specifications nor whether the process is even capable of
producing within specifications.
• We must rely on another measure AFTER we have assured
that the process is in control using SPC.
 Process Capability

• Process capability is the ability of the process, as it currently exists, to product within
specifications.
• One measure known as Cp compares the natural variation of the process to the specification
width.
• Another, more precise, measure known as Cpk compares the natural variation of the process to
the specification width and target.
• Process capability is the ability of the process to meet the design
specifications for a service or product.
• Nominal value is a target for design specifications.
• Tolerance is an allowance above or below the nominal value.
Definitions

Cp= Process Capability. A simple and straightforward indicator of process capability.

Cpk= Process Capability Index. Adjustment of Cp for the effect of non-centered distribution.

Pp= Process Performance. A simple and straightforward indicator of process performance.

Ppk= Process Performance Index. Adjustment of Pp for the effect of non-centered distribution.
 Process Capability

Nominal
value

Process distribution
Lower Upper
specification specification

20 25 30

Process is capable
 Process Capability

Nominal
value

Process distribution
Lower Upper
specification specification

20 25 30

Process is not capable

 Process Capability

Process Capability (PC) is the range in which "all" output can be produced – the inherent capability
of the process.
Process Capability and Process Specifications
Process Capability and Process Specifications

This process is CAPABLE of producing all

good output.

Control the process.

This process is NOT CAPABLE.

INSPECT - Sort out the defectives14-16
Process Capability Index
Effects of Reducing
Variability on Process Capability

Nominal value
Six sigma

Four sigma

Two sigma

Lower Upper
specification specification

Mean
Six sigma
 What is Six Sigma (6σ)?

• Sigma (σ) is a statistical concept that represents how much

variation there is in a process relative to customer
specifications.
• Sigma Value is based on “defects per million opportunities”
(DPMO).
• Six Sigma (6σ) is equivalent to 3.4 DPMO. The variation in
the process is so small that the resulting products and services
are 99.99966% defect free.

Amount of Variation Effect Sigma Value

Too much Hard to produce output within Low (0 – 2)
customer specifications
Moderate Most output meets customer Middle (3 – 5)
specifications
Very little Virtually all output meets High (6)
customer specifications
 Six Sigma Concept

Every Human Activity Has Variability...

Customer
Specification
Customer
Specification

defects
Target

Reducing Variability is the Key to Understanding Six Sigma

Six Sigma Concept

Parking Your Car in the Garage

Has Variability...

Customer Customer
Specification Target Specification

defects defects
 Six Sigma Concept

A 3  process because 3 standard deviations

fit between target and spec
Before
Target Customer
3 Specification

1

2

3
After

Target
6  Customer
Specification

By reducing the variability,

we improve the process 1 No Defects!
3

6
 What’s Wrong With 99% Quality?

3.8 Sigma Six Sigma

99% Good 99.99966% Good
20,000 articles of mail lost per hour 7 articles of mail lost per hour

Unsafe drinking water for almost 15 Unsafe drinking water for 1 minute
minutes each day every 7 months

5,000 incorrect surgical operations 1.7 incorrect surgical operations

per week per week

2 short or long landings at most major 1 short or long landing at most

airports each day major airports every 5 years

200,000 wrong drug prescriptions 68 wrong drug prescriptions

dispensed each year dispensed each year
• Lot Tolerance Proportion Defective
Problems
Problems
Reference:
J.R.Nagala, ‘Statistics for Textile Engineers’, WoodHead Publishing (2014), pp. 278-294.