INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

Quality
Learning Objectives

• Quality:
• Definition

• Dimensions

• Determinants

• Cost

• Impact

2
What is Quality? getting what you pay for

• Conformance to specifications
• Fitness for use
• Value for price paid

• Q = P/E
• Q = Quality
• P = Performance
• E = Expectations

3
What is Quality?

“The quality of a product or service is a customer’s perception of the degree to which

the product or service meets his or her expectations.”

4
Nature of Quality

• Dimensions of Quality

• Determinants of Quality

• Costs of Quality

5
Some Dimensions of Product Quality

• Performance – primary characteristics (signal to noise ratio, power - time to process customer requests)

• Features – secondary characteristics, special characteristics (remote control)

The “extra” items added to the basic features

• Reliability – Consistency of performance over time. (MTTF,MTBF,breakdowns, malfunctions)

• Serviceability – Ease of repair. (speed/cost/convenience of servicing)

6
• Durability – useful life amount of time/use before repairs

• Appearance – aesthetics, effects on human senses

• Customer service – treatment before/during/after sale

• Safety – user protection before/during/after use

7
8
Customer’s and producer’s perspectives depend
on each other
Producer’s perspective:
• production process and COST
Customer’s perspective:
• fitness for use and PRICE
Customer’s view must dominate

9
Learning Objectives

• Quality:
• Definition

• Dimensions

• Determinants

• Cost

• Impact

10
 Determinants of Quality
• Quality of design – products/service designed based on customers’ expectations and
desires.

• Quality capability of production processes –

processes must be capable of producing the products designed for the customers (surface finish in microns : m/cs. Tools.
Measuring equipment)

• Quality of conformance – refers to the degree

to which the product or service design
specifications are met. (capable processes can produce inferior product if not
operated properly.

• Quality of customer service – a superior product does not mean success;

must have quality service also

• Organization quality culture – superior product and service requires

organization-wide focus on quality
 Costs of Quality
• Scrap and rework - rescheduling, repairing, retesting

• Defective products in the hands of the customer - recalls,

warranty claims, due to awareness - law suits, lost business, …

• Detecting defects - inspection, testing, ….

• Preventing defects - training, charting performance,

• Product/process redesign, supplier development, ….

 Cost of Quality

Cost of Achieving Good Quality

Prevention costs
costs incurred during product design
Appraisal costs
costs of measuring, testing, and analyzing
Cost of Poor Quality
Internal failure costs
include scrap, rework, process failure, downtime, and price
reductions
External failure costs
include complaints, returns, warranty claims, liability, and lost
sales
 Prevention Costs
• Quality planning costs Training costs
• costs of developing and • costs of developing and
implementing quality putting on quality training
management program programs for employees and
• Product-design costs management
• costs of designing products Information costs
with quality characteristics
costs of acquiring and
• Process costs •

maintaining data related to

• costs expended to make sure quality, and development and
productive process conforms
to quality specifications analysis of reports on quality
performance
 Appraisal Costs
• Inspection and testing
– costs of testing and inspecting materials, parts, and product
at various stages and at end of process
• Test equipment costs
– costs of maintaining equipment used in testing quality
characteristics of products
• Operator costs
– costs of time spent by operators to gather data for testing
product quality, to make equipment adjustments to
maintain quality, and to stop work to assess quality
 Internal Failure Costs
• Scrap costs Process downtime costs
• costs of poor-quality products • costs of shutting down
that must be discarded, including productive process to fix
labor, material, and indirect costs
problem
• Rework costs
Price-downgrading costs
• costs of fixing defective products
to conform to quality • costs of discounting poor-
specifications quality products—that is,
selling products as
• Process failure costs
“seconds”
• costs of determining why
production process is producing
poor-quality products
 External Failure Costs
• Customer complaint costs Product liability costs
• costs of investigating and • litigation costs resulting
satisfactorily responding to a from product liability and
customer complaint resulting from customer injury
a poor-quality product Lost sales costs
• Product return costs • costs incurred because
• costs of handling and replacing customers are dissatisfied
poor-quality products returned by with poor-quality products
customer and do not make additional
• Warranty claims costs purchases

• costs of complying with product

warranties
Cost of Quality

Cost of achieving good quality Cost of poor quality

• Appraisal • Internal failure

• Prevention • External failure
Learning Objectives

• Quality:
• Definition

• Dimensions

• Determinants

• Cost

• Impact

19
 Do O&SCM and QM practices create an
impact?
• Impact on the
– Organization
– Customers
– Society

20
Analysis and Use of Quality Costs (example)
Quality costs

• Leverage effect

• Pareto analysis
Quality and efficiency/productivity

• Efficiency = Output/Input • Illustration:

• Consider the manufacture of a
mechanical component used in an
assembly.
• Leverage effect: Quality • Parts are manufactured in a process at
improvements more than pay a rate of approximately 100 parts per
day.
for themselves
• First-pass yield = 70% (conforming).
• 70% of the fallout (the 30%
nonconforming) can be reworked into
an acceptable product, and the rest
must be scrapped.
Quality and efficiency/productivity

• The direct manufacturing cost = ₹30/part.

• Reworked parts incur an additional processing charge = ₹6

Cost / good part = ?

Total Yield after rework = ?
Quality and efficiency/productivity

• The direct manufacturing cost = ₹30/part.

• Reworked parts incur an additional processing charge = ₹6

Cost / good part = ((₹30*100) + (₹6*21))/91 = ₹ 34.35

Total Yield after rework = 91 good parts / day
Quality and efficiency/productivity
• Quality management procedures reduce the variability and thereby process
fallout from 30% to 5%.

• Of the 5% fallout produced, about 70% can be reworked & 30% are
scrapped
Cost / good part = ?
Total Yield after rework = ?

• Quality management led to

• ?% reduction in manufacturing costs
• ?% increase in productivity
Quality and efficiency/productivity
• Quality management procedures reduce the variability and thereby process
fallout from 30% to 5%.

• Of the 5% fallout produced, about 70% can be reworked & 30% are
scrapped
Cost / good part = ((₹30*100) + (₹6*3.5))/98.5 = ₹30.67
Total Yield after rework = 98.5 good parts / day

• Quality management led to

• 10.7% reduction in manufacturing costs
• 8.2% increase in productivity
Quality and efficiency/productivity

• Typical cost of deploying quality management procedures (Cost

of doing things right)
• 3-4 % of revenues

• Cost of doing things wrong

• 20-35% of revenues
 Quality–Cost Relationship
• Cost of quality
– difference between price of nonconformance
and conformance
– cost of doing things wrong
• 20 to 35% of revenues
– cost of doing things right
• 3 to 4% of revenues
 Measuring Product Yield
and Productivity
Yield=(total input)(% good units) + (total input)(1-%good units)(% reworked)

or
Y=(I)(%G)+(I)(1-%G)(%R)

where
I = initial quantity started in production
%G = percentage of good units produced
%R = percentage of defective units that are successfully reworked
 Computing Product Yield
• Motor manufacturer
• Starts a batch of 100 motors.
• 80 % are good when produced
• 50 % of the defective motors can be reworked
Y =(I)(%G)+(I)(1-%G)(%R)
=

Increase quality to 90% good

Y =
 Computing Product Yield
• Motor manufacturer
• Starts a batch of 100 motors.
• 80 % are good when produced
• 50 % of the defective motors can be reworked
Y =(I)(%G)+(I)(1-%G)(%R)
= 100(.80) + 100(1-.80)(.50) = 90 motors

Increase quality to 90% good

Y =100(.90) + 100(1-.90)(.50) = 95 motors
Computing Product Cost per Unit

( K d )( I ) + ( K r )( R)
Product Cost
=
Y
where:
Kd = direct manufacturing cost per unit
I = input
Kr = rework cost per unit
R = reworked units
Y = yield
 Cost per Unit
Direct cost = $30 Rework cost = $12
80% good 50% can be reworked

( Kd )( I ) + ( Kr )( R)
=
Y
Increase quality to 90% good
 Cost per Unit
Direct cost = $30 Rework cost = $12
80% good 50% can be reworked

( Kd )( I ) + ( Kr )( R)
= $30*100 + $12*10 = $34.67/motor
Y 90 motors

Increase quality to 90% good

$30*100 + $12*5 = $32.21/motor
=
95 motors
 Computing Product Yield
for Multistage Processes
Y = (I)(%g1)(%g2) … (%gn)

where:
I = input of items to the production process that will
result in finished products
gi = good-quality, work-in-process products at stage i
 Multistage Yield
Average Percentage
Stage Good Quality
1 0.93
2 0.95
3 0.97
4 0.92

Y = (I)(%g1)(%g2) … (%gn)
 Multistage Yield
Average Percentage
Stage Good Quality
1 0.93
2 0.95
3 0.97
4 0.92

Y = (I)(%g1)(%g2) … (%gn)

= 100 * .93 * .95 * .97 * .92 = 78.8 motors

Initial Batch Size For 100 Motors
Y
I= (%g1)(%g2) … (%gn)
Initial Batch Size For 100 Motors
Y
I= (%g1)(%g2) … (%gn)

100
= = 126.88 → 127
100 * .93 * .95 * .97 * .92
 Quality–Productivity Ratio
QPR
– productivity index that includes productivity and quality
costs

(good-quality units)
QPR = (100)
(input) (processing cost) + (reworked units) (rework cost)
 Quality Productivity Ratio
Direct cost = $30 Rework cost = $12
80% good 50% can be reworked
Initial batch size = 100

Base Case

QPR =

Case 1: Increase I to 200

QPR =
 Quality Productivity Ratio
Case 2: Reduce direct cost to $26 and rework cost to $10

QPR =

Case 3: Increase %G to 95%

QPR =

Case 4: Decrease costs and increase %G

QPR =
 Quality Productivity Ratio
Direct cost = $30 Rework cost = $12
80% good 50% can be reworked
Initial batch size = 100

Base Case
80 + 10
QPR = (100) = 2.89
100 * $30 + 10 * $12

Case 1: Increase I to 200

160 + 20
QPR = (100) = 2.89 – NO CHANGE
200 * $30 + 20 * $12
 Quality Productivity Ratio
Case 2: Reduce direct cost to $26 and rework cost to $10
80 + 10
QPR = (100) = 3.33
100 * $26 + 10 * $10

Case 3: Increase %G to 95%

95 + 2.5
QPR = (100) = 3.22
100 * $30 + 2.5 * $12

Case 4: Decrease costs and increase %G

95 + 2.5
QPR = (100) = 3.71
100 * $26 + 2.5 * $10
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

Quality Management
Quality improvement in modern business environment

Prof. Tarun Sharma

Contents/ Learning activities

• Revisit: Quality dimensions,

definitions, terminology
• Statistical methods for quality
control and Improvement
– SPC
– DOX
– AS

• Management aspects of
quality improvement
– Six sigma

2
Definitions – Meaning of Quality and Quality
Improvement
1.1.1 The Eight Dimensions of Quality

1. Performance
2. Reliability
3. Durability
4. Serviceability
5. Aesthetics
6. Features
7. Perceived Quality
8. Conformance to Standards

3
• This is a traditional definition
• Quality of design
• Quality of conformance

This is a modern definition of quality

4
The Transmission Example

• Quality is inversely proportional to variability

• Customer doesn’t see the mean of the process, s/he only sees
the variability around the target that you have not removed - Jack
Welch (retired CEO - GE)
5
 What is Quality Improvement?
• Quality Improvement is the reduction of
variability in processes and products.

• Excessive variability in process performance

often results in waste.

• Quality Improvement is the reduction of

waste.
 Quality Terminology
• Quality / Critical-To-Quality Characteristics (CTQ)
– Every product possesses a number of elements that jointly
describe what the user or consumer thinks of as quality
• Physical: length, weight, voltage, viscosity
• Sensory: taste, appearance, color
• Time Orientation: reliability, durability, serviceability

• Quality Engineering
– Set of operational, managerial, and technical activities that
a company uses to ensure that the quality characteristics
of a product are at the nominal or required levels and that
the variability around these desired levels is minimum.
 Quality Terminology
• Data Types
– Variables data are usually continuous measurements, such as length, voltage,
or viscosity.

– Attributes data, on the other hand, are usually discrete data, often taking the
form of counts.

• Specifications
– For a manufactured product, the specifications are the desired measurements
for the quality characteristics of the components and subassemblies that make
up the product, as well as the desired values for the quality characteristics in
the final product.

– In the service industries, specifications are typically in terms of the maximum

amount of time to process an order or to provide a particular service. (Cycle
Time)
 Quality Terminology
• Nominal / Target Value
– A value of a measurement that corresponds to the desired value for that
quality characteristic

• Upper Specification Limit (USL)

– The largest allowable value for a quality characteristic

• Lower Specification Limit (LSL)

– The smallest allowable value for a quality characteristic

• Nonconforming Product
– Products that fail to meet one or more of its specifications.

• Nonconformity
– Specific type of failure
 Quality Terminology
• Defective, Defects
– A nonconforming product is considered defective if it
has one or more defects, which are nonconformities
that are serious enough to significantly affect the safe
or effective use of the product.

• Concurrent Engineering
– Stressed a team approach to design, with specialists in
manufacturing, quality engineering, and other
disciplines working together with the product
designer at the earliest stages of the product design
process.
Statistical methods for quality control and
improvement

11
Statistical methods

• Statistical process control (SPC)

– Control charts, plus other problem-solving tools
– Useful in monitoring processes, reducing variability
through elimination of assignable causes
– On-line technique
• Designed experiments (DOX)
– Discovering the key factors that influence process
performance
– Process optimization
– Off-line technique
• Acceptance Sampling

12
Status of a Process

• What is the true

status of a
process?

• How is status of
the process
established?

13
Statistical Process Control (SPC)

• Walter A. Shewart • A control chart is one of the

(1891-1967) primary techniques of
statistical process control
• Trained in engineering and
(SPC)
physics
• Long career at Bell Labs • Classically, control charts
• Developed the first control are applied to the output
chart about 1924 variable(s) in a system

• However, in some cases

they can be usefully
applied to the inputs as
well.
14
 Design of Experiments
• Approach to systematically vary the controllable input factors in the process and
determining the effect these factors have on the output product parameters.

• Off-line improvement technique

• Leads to a model of the process

• One major type of designed experiment is the factorial design, in which factors are varied
together in such a way that all possible combinations of factor levels are tested.

15
Acceptance Sampling
• Inspection and classification of a sample of units selected at
random from a larger batch or lot and the ultimate decision
about disposition of the lot, usually occurs at two points:
incoming raw materials or components and final production.

16
Quality Engineering

• The primary objective of

quality engineering efforts is
the systematic reduction of
variability in the key quality
characteristics of the
product.

17
Modern Quality Assurance Systems

• Modern quality assurance

systems usually place less
emphasis on acceptance-
sampling and attempt to
make statistical process
control and designed
experiments the focus of
their efforts.

18
Basis for SPC

• Probability Distributions

• Sampling
– Distribution of Sample Statistics

• Chance and assignable causes

20
Basis for SPC

• Probability Distributions

• Sampling
– Distribution of Sample Statistics

• Chance and assignable causes

21
Sometimes called a Sometimes called a
probability mass function probability density function

22
22
23
24
25
Standard Normal Distribution

26
• Suppose a measurement is distributed normally and has a
mean value of 0.5 m. with a standard deviation of 0.002 m.
1) How many standard deviations away is a measurement
value of 0.5039 and 0.4961 from the mean?
2) Given the distribution how likely (probability) that an
operator will get a measurement higher than 0.5039 and
lower than 0.4961?
z = (x - u) / s

where x is the measurement value, u is the mean, and s is the standard deviation.

For x = 0.5039:
z = (0.5039 - 0.5) / 0.002 = 1.95 standard deviations away from the mean.

For x = 0.4961:
z = (0.4961 - 0.5) / 0.002 = -1.95 standard deviations away from the mean.

To find the probability of getting a measurement higher than 0.5039 and lower than 0.4961, we need to calculate the area under the normal
distribution curve between these two values.
Using a standard normal distribution table or a calculator, we can find that the area to the left of z = 1.95 is 0.9744, and the area to the left of z
= -1.95 is 0.0256. Therefore, the area between z = -1.95 and z = 1.95 is:

0.9744 - 0.0256 = 0.9488

This means that there is a 94.88% probability that an operator will get a measurement between 0.4961 and 0.5039.
28
• The tensile strength of paper used to make grocery bags is
an important quality characteristic. It is known that the
strength – say x – is normally distributed with mean 40 lb/sq
inch and standard deviation 2lb/sq inch. The purchaser of
the bags requires them to have a strength of at least 35 lb/sq
inch. Calculate the probability that the bags produced from
this paper will meet or exceed the specification.
We have a normal distribution with mean u = 40 lb/sq inch and standard deviation s= 2 lb/sq inch. The purchaser requires a minimum strength
of x = 35 lb/sq inch.

To find the probability that the bags produced will meet or exceed this specification, we need to calculate the area under the normal distribution
curve to the right of x = 35.

We can use the z-score formula to standardize the distribution:

z = (x - u) / s

z = (35 - 40) / 2 = -2.5

Now we need to find the area to the right of z = -2.5 under the standard normal distribution. Using a standard normal distribution table or a
calculator, we can find that this area is approximately 0.9938.

Therefore, the probability that a bag produced from this paper will meet or exceed the specification of 35 lb/sq inch is approximately 0.9938 or
99.38%.

29
Suppose that x is normally distributed with mean 10 and
variance 9. Find the value of x – say, a – such that P(x>a) =
0.05.
We have a normal distribution with mean u= 10 and variance s^2 = 9. The standard deviation is s = (s^2) = 3.

We want to find the value of x – say, a – such that P(x > a) = 0.05.

To find this value, we can use the inverse normal distribution, also known as the z-score or standard score.

We start by finding the z-score corresponding to the probability of 0.05:

P(x > a) = 0.05

P(z > (a - u) / s) = 0.05

Using a standard normal distribution table or a calculator, we can find that the z-score corresponding to a probability of 0.05 is approximately 1.645.

Now we can use the formula for the z-score:

z = (x - u) / s

Rearranging the formula to solve for x, we get:

x=u+s*

Substituting the values we have:

x = 10 + 1.645 * 3

x = 14.935

Therefore, the value of x – a – such that P(x > a) = 0.05 is approximately 14.935.
31
Basis for SPC

• Probability Distributions

• Sampling
– Distribution of Sample Statistics

• Chance and assignable causes

33
 Central Limit Theorem

▪ Practical interpretation – the sum of independent random variables

is approximately normally distributed regardless of the distribution
of each individual random variable in the sum

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 4 – 34
 Central Limit Theorem
▪ One of the most important results in statistics. Allows us to apply
many statistical techniques to data that are not normally
distributed!

E(x ) =  x =
n
▪ Consider a variable normally distributed with mean 3.473 and
standard deviation of 1.7


© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 4 – 35
Large sample

36
Underlying distribution: Normal

37
Underlying distribution: Exponential

38
Underlying distribution: Uniform

39
• Suppose a measurement is distributed normally and has a
mean value of 0.5 m. with a standard deviation of 0.002 m.
1) How many standard deviations away is a measurement
value of 0.5039 and 0.4961 from the mean?
2) Given the distribution how likely (probability) that an
operator will get a measurement higher than 0.5039 and
lower than 0.4961?
3) If instead of single observations, sample sizes of 5 are
taken and the mean value of the sample is used (as is in a
control chart) what will be the 95% confidence interval for
the sample mean?

40
Basis for SPC

• Probability Distributions

• Sampling
– Distribution of Sample Statistics

• Chance and assignable causes

41
Chance and assignable causes of
variation
• A process is operating with only chance causes of variation present is said to be in
statistical control.
• A process that is operating in the presence of assignable causes is said to be out of
control.

42
 Six Sigma Quality
▪ Process performance is not predictable unless the process
behavior is stable.

▪ However, no process or system is ever truly stable, and even in

the best of situations, disturbances occur.

▪ These disturbances can result in the process mean shifting off-

target, an increase in the process standard deviation, or both.

▪ The concept of a six-sigma process is one way to model this

behavior.

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 43
 Six Sigma
▪ Six-Sigma is a business strategy that seeks to improve business
performance by identifying and removing the causes of defects
and errors

▪ Motorola developed the Six-Sigma program in the late 1980s as a

response to the demand for their products.

▪ The focus of six-sigma is reducing variability in key product quality

characteristics to the level at which failure or defects are
extremely unlikely.

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 44
 Six Sigma
▪ Probability of producing a product
within these specifications is 0.9973,
which corresponds to 2700 parts per
million (ppm) defective.

▪ This is referred to as three-sigma

quality performance which sounds
pretty good

▪ Now suppose we have a product that

consists of an assembly of 100
independent components and all 100
parts must be non defective for the
product to function satisfactorily.

▪ The probability that any specific unit of

product is non defective is

0.9973 * 0.9973 *. . .* 0.9973 = 0.7631

▪ 23.7% of products are defective.

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 45
 ▪ The Motorola six-sigma concept
is to reduce the variability in the
process so that the specification
limits are at least six standard
deviations from the mean.
▪ Generally, we can only make
predictions about process
performance when the process
is stable.
▪ If the mean is drifting around,
and ends up as much as 1.5
standard deviations off target, a
prediction of 3.4 ppm defective
may not be very reliable,
because the mean might shift
by more than the “allowed” 1.5
standard deviations.
© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 46
Review

• Why are we bothered about probability distributions?

• Why do we care about sampling?

• Why do we care about standard normal distribution (z

scores)?

47
Key points

• Distinction between specification/control limits and the

Control limits

• A quality attribute whose 6 sigma variation (on each side of

the mean) coincides with the specification limits is of better
quality than the same quality attribute having its 3 sigma
variation (on each side of the mean) coincide with the same
specification limit.

• Are the control limits of the control chart same as the 3

sigma variation in the quality attribute as measured in the
entire population of that product?

48
• The control limits on the control chart are the in terms of the
standard deviation of the statistic plotted on the chart. It is
not the standard deviation of the quality characteristic.

49
 Six Sigma
▪ Typical six-sigma projects are
four to six months in duration
and are selected for their
potential impact in the
business.

▪ Six-sigma uses a specific five-

step problem solving
approach: Define, Measure,
Analyze, Improve, and
Control (DMAIC).

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 50
51
52
53
54
55
 Six Sigma (Example)
▪ Consider the visit to a fast-food restaurant. The customer orders a
typical meal: a hamburger bun, meat, special sauce, cheese, pickle,
onion, lettuce, and tomato, fries, and a soft drink.
▪ This product has ten components (independent). Is 99% good quality
satisfactory?
P(Single meal good) = (0.99)10 = 0.9044
▪ Now suppose that the customer is a family of four.
P{All meal good} = (0.9044)4 = 0.6690
▪ Now suppose that this hypothetical family of four visits this restaurant
once a month
P{All visits during the year good} = (0.6690)12 = 0.0080

© John Wiley & Sons, Inc. Managing, Controlling & Improving Quality by Montgomery, Jennings & Pfund 2 – 56
Statistical Process Control
 Lecture Outline

• Basics of Statistical Process Control

• Control Charts
• Control Charts for Attributes
• Control Charts for Variables
• Control Chart Patterns
• SPC with Excel and R
• Process Capability

3-2
 Learning Objectives
• Explain when and how to use statistical process
control to ensure the quality of products and
services
• Discuss the rationale and procedure for
constructing attribute and variable control charts
• Utilize appropriate control charts to determine if a
process is in-control
• Identify control chart patterns and describe
appropriate data collection
• Assess the process capability of a process

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-3
 Statistical Process Control (SPC)

• Statistical Process Control

• monitoring production process
to detect and prevent poor UCL
quality
• Sample
• subset of items produced to
use for inspection LCL
• Control Charts
• process is within statistical
control limits

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-4
 Process Variability

• Random • Non-Random
• inherent in a process • special causes
• depends on equipment • identifiable and
and machinery, correctable
engineering, operator, • include equipment out of
and system of adjustment, defective
measurement materials, changes in
• natural occurrences parts or materials, broken
machinery or equipment,
operator fatigue or poor
work methods, or errors
due to lack of training

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-5
Chance and assignable causes of
variation
• A process is operating with only chance causes of variation
present is said to be in statistical control.
• A process that is operating in the presence of assignable causes
is said to be out of control.

6
 SPC in Quality Management
• SPC uses
• Is the process in control?
• Identify problems in order to make
improvements
• Contribute to the TQM goal of continuous
improvement

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-7
 Quality Measures:
Attributes and Variables
• Attribute
• A characteristic which is evaluated with a
discrete response
• good/bad; yes/no; correct/incorrect
• Variable measure
• A characteristic that is continuous and can be
measured
• Weight, length, voltage, volume

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-8
 SPC Applied to Services

• Nature of defects is different in services

• Service defect is a failure to meet customer
requirements
• Monitor time and customer satisfaction

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-9
 SPC Applied to Services

• Hospitals
• timeliness & quickness of care, staff responses to requests,
accuracy of lab tests, cleanliness, courtesy, accuracy of
paperwork, speed of admittance & checkouts
• Grocery stores
• waiting time to check out, frequency of out-of-stock items, quality
of food items, cleanliness, customer complaints, checkout
register errors
• Airlines
• flight delays, lost luggage & luggage handling, waiting time at
ticket counters & check-in, agent & flight attendant courtesy,
accurate flight information, cabin cleanliness & maintenance

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-10
 SPC Applied to Services

• Fast-food restaurants
• waiting time for service, customer complaints, cleanliness, food
quality, order accuracy, employee courtesy

• Catalogue-order companies
• order accuracy, operator knowledge & courtesy, packaging,
delivery time, phone order waiting time

• Insurance companies
• billing accuracy, timeliness of claims processing, agent
availability & response time

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-11
 Where to Use Control Charts

• Process
• Has a tendency to go out of control
• Is particularly harmful and costly if it goes out of control
• Examples
• At beginning of process because of waste to begin
production process with bad supplies
• Before a costly or irreversible point, after which product is
difficult to rework or correct
• Before and after assembly or painting operations that
might cover defects
• Before the outgoing final product or service is delivered

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-12
 Control Charts
• A graph that monitors process quality
• Control limits
• upper and lower bands of a control chart
• Attributes chart
• p-chart
• c-chart
• Variables chart
• mean (x bar – chart)
• range (R-chart)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-13
 Process Control Chart
Out of control
Upper
control
limit

Process
average

Lower
control
limit

1 2 3 4 5 6 7 8 9 10
Sample number

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-14
 Normal Distribution

• Probabilities for Z= 2.00 and Z = 3.00

95%
99.74%
-3 -2 -1 =0 1 2 3

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-15
 A Process Is in Control If …

1. … no sample points outside limits

2. … most points near process average
3. … about equal number of points above
and below centerline
4. … points appear randomly distributed

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-16
 Control Charts for Attributes
• p-chart
• uses portion defective in a sample

• c-chart
• uses number of defects (non-conformities) in a
sample

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-17
 p-Chart

UCL = p + zp
LCL = p - zp

z = number of standard deviations from process average

p = sample proportion defective; estimates process mean
p = standard deviation of sample proportion

p(1 - p)
p = n

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-18
 Construction of p-Chart

NUMBER OF PROPORTION
SAMPLE # DEFECTIVES DEFECTIVE
1 6 .06
2 0 .00
3 4 .04
: : :
: : :
20 18 .18
200

20 samples of 100 pairs of jeans

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-19
 Construction of p-Chart

total defectives
p= =
total sample observations

p(1 - p)
UCL = p + z =
n
UCL =

p(1 - p)
LCL = p - z =
n
LCL =

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-20
 Construction of p-Chart

total defectives
p= = 200 / 20(100) = 0.10
total sample observations

p(1 - p) 0.10(1 - 0.10)

UCL = p + z = 0.10 + 3
n 100
UCL = 0.190

p(1 - p) 0.10(1 - 0.10)

LCL = p - z = 0.10 - 3
n 100
LCL = 0.010

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-21
 Construction of p-Chart

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-22
 p-Chart in Excel
Click on “Insert” then “Charts”
to construct control chart

I4 + 3*SQRT(I4*(1-I4)/100)

I4 - 3*SQRT(I4*(1-I4)/100)

Column values copied

from I5 and I6
© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-23
 c-Chart

UCL = c + zc
c = c
LCL = c - zc

where
c = number of defects per sample

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-24
 c-Chart

Number of defects in 15 sample rooms

NUMBER
OF
SAMPLE DEFECTS

1 12 c=
2 8 UCL = c + zc
3 16
: : LCL = c - zc
: :
15 15
190

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-25
 c-Chart

Number of defects in 15 sample rooms

NUMBER
OF
SAMPLE DEFECTS
190
1 12 c= = 12.67
15
2 8
UCL = c + zc
3 16
= 12.67 + 3 12.67
: : = 23.35
: : LCL = c - zc
15 15 = 12.67 - 3 12.67
190 = 1.99

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-26
c-Chart

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-27
 Control Charts for Variables

▪ Range chart ( R-Chart )

▪ Plot sample range (variability)
▪ Mean chart ( x -Chart )
▪ Plot sample averages

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-28
 x-bar Chart:  Known

UCL = =
x + z x-
LCL = =
x - z -x
Where - + x- + ... + x-
= x
X= 1 2
k
k

 = process standard deviation

x = standard deviation of sample means =/ n
k = number of samples (subgroups)
n = sample size (number of observations)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-29
 x-bar Chart Example:  Known
Observations(Slip-Ring Diameter, cm) n

Sample k 1 2 3 4 5
-x

We know σ = .08

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e

3-30
 x-bar Chart Example:  Known

= x- 1 + x- 2 + ... + x- k
X= k

=
UCL = =
x + z -x LCL = x - z -x

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e

3-31
 x-bar Chart Example:  Known

= 50.09
X = _____ = 5.01
10
=
UCL = x + z -x LCL = =
x - z -x
= 5.01 + 3(.08 / 5 ) = 5.01 - 3(.08 / 5 )
= 5.12 = 4.90

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e

3-32
 x-bar Chart Example:  Unknown

_ _
UCL = =x + A2R LCL = x= - A2R

where
=
x = average of the sample means
_
R = average range value

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e

3-33
 Sample
Control Size Factor for X-chart Factors for R-chart
n A2 D3 D4
Chart 2
3
1.880
1.023
0.000
0.000
3.267
2.575

Factors 4
5
0.729
0.577
0.000
0.000
2.282
2.114
6 0.483 0.000 2.004
7 0.419 0.076 1.924
8 0.373 0.136 1.864
9 0.337 0.184 1.816
10 0.308 0.223 1.777
11 0.285 0.256 1.744
12 0.266 0.283 1.717
13 0.249 0.307 1.693
14 0.235 0.328 1.672
15 0.223 0.347 1.653
16 0.212 0.363 1.637
17 0.203 0.378 1.622
18 0.194 0.391 1.609
19 0.187 0.404 1.596
20 0.180 0.415 1.585
21 0.173 0.425 1.575
22 0.167 0.435 1.565
23 0.162 0.443 1.557
24 0.157 0.452 1.548
25 0.153 0.459 1.541

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-34
 x-bar Chart Example:  Unknown
OBSERVATIONS (SLIP- RING DIAMETER, CM)
SAMPLE k 1 2 3 4 5 x R
1 5.02 5.01 4.94 4.99 4.96 4.98 0.08
2 5.01 5.03 5.07 4.95 4.96 5.00 0.12
3 4.99 5.00 4.93 4.92 4.99 4.97 0.08
4 5.03 4.91 5.01 4.98 4.89 4.96 0.14
5 4.95 4.92 5.03 5.05 5.01 4.99 0.13
6 4.97 5.06 5.06 4.96 5.03 5.01 0.10
7 5.05 5.01 5.10 4.96 4.99 5.02 0.14
8 5.09 5.10 5.00 4.99 5.08 5.05 0.11
9 5.14 5.10 4.99 5.08 5.09 5.08 0.15
10 5.01 4.98 5.08 5.07 4.99 5.03 0.10
Totals 50.09 1.15

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-35
36
37
 x-bar Chart Example:  Unknown
_ ∑R
____
R= k

_
= x
___
x=
k
_
=
UCL = x + A2R
_
=
LCL = x - A2R

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-38
 x-bar Chart Example:  Unknown
_ ∑R 1.15
____ ____
R= k = 10 = 0.115

_
= x
___ 50.09
_____
x= = = 5.01 cm
k 10
_
=
UCL = x + A2R = 5.01 + (0.58)(0.115) = 5.08
_
=
LCL = x - A2R = 5.01 - (0.58)(0.115) = 4.94

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-39
 x- bar
Chart
Example

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-40
Piston rings
Control Limits = 3 sigma
Piston rings
Control limit .99 Confidence level
 R- Chart

UCL = D4R LCL = D3R

R
R=
k
Where
R = range of each sample
k = number of samples (sub groups)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-43
44
Chapter 6 Introduction to Statistical Quality Control, 6th Edition by Douglas C. Montgomery. 45
Copyright (c) 2009 John Wiley & Sons, Inc.
 R-Chart Example
OBSERVATIONS (SLIP- RING DIAMETER, CM)
SAMPLE k 1 2 3 4 5 x R
1 5.02 5.01 4.94 4.99 4.96 4.98 0.08
2 5.01 5.03 5.07 4.95 4.96 5.00 0.12
3 4.99 5.00 4.93 4.92 4.99 4.97 0.08
4 5.03 4.91 5.01 4.98 4.89 4.96 0.14
5 4.95 4.92 5.03 5.05 5.01 4.99 0.13
6 4.97 5.06 5.06 4.96 5.03 5.01 0.10
7 5.05 5.01 5.10 4.96 4.99 5.02 0.14
8 5.09 5.10 5.00 4.99 5.08 5.05 0.11
9 5.14 5.10 4.99 5.08 5.09 5.08 0.15
10 5.01 4.98 5.08 5.07 4.99 5.03 0.10
Totals 50.09 1.15

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-46
 R-Chart Example
_
UCL = D4R =
_
LCL = D3R =

Retrieve chart factors D3 and D4

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-47
 R-Chart Example
_
UCL = D4R = 2.11(0.115) = 0.243
_
LCL = D3R = 0(0.115) = 0

Retrieve chart factors D3 and D4

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-48
 R-Chart Example

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-49
 Using x- bar and R-Charts Together
• Process average and process variability must be
in control
• Samples can have very narrow ranges, but
sample averages might be beyond control limits
• Or, sample averages may be in control, but
ranges might be out of control
• An R-chart might show a distinct downward
trend, suggesting some nonrandom cause is
reducing variation

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-50
 Control Chart Patterns

• Run
• sequence of sample values that display same
characteristic
• Pattern test
• determines if observations within limits of a control
chart display a nonrandom pattern

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-51
 Control Chart Patterns
• To identify a pattern look for:
• 8 consecutive points on one side of the center line
• 8 consecutive points up or down
• 14 points alternating up or down
• 2 out of 3 consecutive points in zone A (on one side of
center line)
• 4 out of 5 consecutive points in zone A or B (on one
side of center line)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-52
 Control Chart Patterns

UCL UCL

LCL LCL

Sample observations Sample observations

consistently below the consistently above the
center line center line

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-53
 Control Chart Patterns

UCL UCL

LCL LCL

Sample observations Sample observations

consistently increasing consistently decreasing

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-54
 Zones for Pattern Tests
UCL =
3 sigma = x + A2R
Zone A
= 2
2 sigma = x + 3 (A2R)

Zone B
= 1
1 sigma = x + 3 (A2R)

Process Zone C =
x
average
Zone C
=
1 sigma = x - 1 (A2R)
3
Zone B
=
2 sigma = x - 2 (A2R)
3
Zone A
=
LCL 3 sigma = x - A2R
| | | | | | | | | | | | |
1 2 3 4 5 6 7 8 9 10 11 12 13
Sample number
© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-55
 Performing a Pattern Test

SAMPLE x ABOVE/BELOW UP/DOWN ZONE

1 4.98 B — B
2 5.00 B U C
3 4.95 B D A
4 4.96 B D A
5 4.99 B U C
6 5.01 — U C
7 5.02 A U C
8 5.05 A U B
9 5.08 A U A
10 5.03 A D B

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-56
 Sample Size Determination

• Attribute charts require larger sample sizes

• 50 to 100 parts in a sample

• Variable charts require smaller samples

• 2 to 10 parts in a sample

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-57
 SPC with Excel

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-58
 INDIAN INSTITUTE OF TECHNOLOGY ROORKEE

Process Capability Analysis

Prof. Tarun Sharma
Process Capability

Process variability
• Are they aligned with each
other?
Design
specifications
Definitions

• Process variability
• Design Specifications
• Process Range = 6*sigma
• Upper Specification Limit (USL)
Where sigma is the standard
deviation of the CTQ dimension
• Lower Specification Limit (LSL)

• Tolerance Range = USL - LSL

Definitions

• Process fallout

• Fraction of the process output which does not meet the specification,
i.e., the CTQ dimension is outside of the design specifications
Numeric measure of Process Capability

• Process Capability Ratio (Cp) = Tolerance Range/ Process variation

• Process Capability Index (Cpk) = Min{(Mean – LSL)/(3*sigma),

(USL-Mean)/(3*sigma)}
 Process Capability

• Compare natural variability to design variability

• Natural variability
• What we measure with control charts
• Process mean = 8.80 oz, Std dev. = 0.12 oz
• Tolerances
• Design specifications reflecting product
requirements
• Net weight = 9.0 oz  0.5 oz
• Tolerances are  0.5 oz

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-66
 Process Capability
Design
Specifications

(a) Natural variation

exceeds design
specifications; process is
not capable of meeting
specifications all the
time.
Process
Design
Specifications

(b) Design specifications

and natural variation the
same; process is capable
of meeting specifications
most of the time.

Process
© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-67
 Process Capability
Design
Specifications

(c) Design specifications

greater than natural
variation; process is
capable of always
conforming to
specifications.
Process
Design
Specifications

(d) Specifications greater

than natural variation, but
process off center; capable
but some output will not
meet upper specification.

Process

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-68
 Process Capability Ratio

tolerance range
Cp =
process range

upper spec limit - lower spec limit

=
6

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-69
 Computing Cp

Net weight specification = 9.0 oz  0.5 oz

Process mean = 8.80 oz
Process standard deviation = 0.12 oz

upper specification limit -

lower specification limit
Cp =
6

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-70
 Computing Cp

Net weight specification = 9.0 oz  0.5 oz

Process mean = 8.80 oz
Process standard deviation = 0.12 oz

upper specification limit -

lower specification limit
Cp =
6

9.5 - 8.5
= = 1.39
6(0.12)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-71
 Process Capability Index

=
x - lower specification limit
,
3
Cpk = minimum =
upper specification limit - x
3

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-72
 Computing Cpk

Net weight specification = 9.0 oz  0.5 oz

Process mean = 8.80 oz
Process standard deviation = 0.12 oz
=
x - lower specification limit
,
Cpk = minimum 3
=
upper specification limit - x
3

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-73
 Computing Cpk

Net weight specification = 9.0 oz  0.5 oz

Process mean = 8.80 oz
Process standard deviation = 0.12 oz
=
x - lower specification limit
,
Cpk = minimum 3
=
upper specification limit - x
3

8.80 - 8.50 9.50 - 8.80

= minimum , = 0.83
3(0.12) 3(0.12)

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-74
Impact of Process Capability Studies on
management decision problems

• Make or buy decision

• Plant and process improvements to reduce process variability

• Contractual agreements with customers or vendors regarding

product quality
 Process Capability With Excel

© 2014 John Wiley & Sons, Inc. - Russell and Taylor 8e 3-76