Quality Management

1
Learning
g Objectives
j
• To understand concepts of Total Quality and Six Sigma:
– Quality defined
– Qualityy control to Qualityy Management
g
– Cost of Quality
– Six Sigma Quality
– Methodology and Tools for Six Sigma Quality
• To understand the concepts of Statistical Quality Control (SQC) : -
– Statistical Process Control (SPC)
• Control Charts for Variables( x-Charts) and R-Charts)
• The Central Limit Theorem
• Natural and assignable causes of variation
• Process Capability
• Control Charts for Attributes (p-charts and C-charts)
– Acceptance
A t Sampling
S li
• Producer’s and consumer’s risk
• Operating Characteristic (OC) Curves
• Averageg Outgoing
g g Quality y

2
 Meaning of Quality

 Webster’s Dictionary
 degree of excellence of a thing
 American Society for Quality
 totality of features and characteristics
that satisfy needs
 Consumer’s and producer’s
p
perspective

3
 Meaning of Quality:
Consumer’s Perspective
 Fitness for use
 how well product or
service does what it is
supposed to
 Quality of design
 designing quality
characteristics into a
product or service
 A Mercedes and a Ford are
equally “fit for use,” but with
different design dimensions

4
 Meaning of Quality:
Producer’s Perspective
 Quality of conformance
 Making
ki sure product
d or service
i isi produced
d d
according to design
 if new
ne tires do not conform to specifications,
specifications y
wobble
 if a hotel room is not clean when a gguest checks in,,
hotel is not functioning according to specifications
of its design

5
 M i off Quality
Meaning Q li
Meaning of Quality

Producer’s Perspective Consumer’s Perspective

Quality of Conformance Quality of Design

Production • Conformance to • Quality characteristics Marketing

specifications • Price
• Cost

Fitness for
Consumer Use

6
 Quality characteristics of goods
and services
• Functionality - how well the product or service does job
for which it was intended
• Appearance - aesthetic appeal, look, feel, sound and
smell of the product or service
• Reliability - consistency of product or services
performance over time.
• Durability - the total useful life of the product or
service.
• Recovery - the ease with which problems with the
product or service can be rectified or resolved.
• Contact - the nature of the person-to-person contacts
that take place.
place
7
 Quality Specifications
• Design quality: Inherent value of the
product in the marketplace

– Dimensions include: Performance, Features,

Reliability/Durability, Serviceability,
Aesthetics, and Perceived Quality.

• Conformance q quality:
y Degree
g to which
the product or service design
specifications are met
 From “Quality
Quality Control
Control” to
“Quality Management”
1 Inspection
– Error detection of finished goods
– Rectification

2 Quality Control
– Statistical Methods applied to processes
– Quality
Q li standards
d d
– Process improvement

9
From “Quality
Quality Control
Control” to
“Quality Management” continued
3 Quality Assurance
– Quality Systems
– Quality Costing
– Quality Planning
– Problem solving
4 Quality Management
– Everyone
E involved;
i l d teamwork; k staff
ff empowerment
– Involves customers and suppliers
– Quality strategy
10
 Costs of Quality

Appraisal Costs

External Failure Costs of

Costs Prevention Costs
Quality

Internal Failure
Costs
11
 Costs of Quality
• Prevention Costs • Appraisal Costs
– New product quality
planning
– Product improvement
– Process
P iimprovementt
– Staff training and
p
development
– Incoming materials
inspection
– Monitoring production
for quality issues
– Customer surveys
– Q
Quality
y laboratories
– Finished goods
inspection

12
Costs of Quality continued

• Internal Failure • External Failure

Costs Costs
– Scrap – Warranty / guarantees
– Reworking / correcting – Returned goods
– Downgrading
di – Handling
dli complaints
l i
– Retest – Allowances / litigation
– Plant downtime – Loss of goodwill
– Investigation /
troubleshooting

13
 Cost of Q
Quality
alit Model
Make P+A more
effective

Total
Cost
osts
Co

Prevention + Appraisal
costs

Internal + External Failure

Costs

Quality Effort
14
 Six Sigma Quality
• A philosophy and set of
methods companies use to
eliminate defects in their
products and processes
• Seeks to reduce variation in
the processes that lead to
product defects
• The name, “six sigma” refers
to the variation that exists
within plus or minus three
standard deviations of the
process outputs
t t
15
 Six Sigma Quality (Continued)

• Six Sigma allows managers to readily describe

process performance using a common metric:
Defects Per Million Opportunities (DPMO)

Number of defects
DPMO  x 1,000,000
 Number of 
 opportunities 
 for error per  x No. of units
 unit 
 

16
 Six Sigma Quality (Continued)
Example of Defects Per Million So, for every
Opportunities (DPMO) calculation. one million
letters delivered
Suppose we observe 200 letters this city
city’s
s postal
delivered incorrectly to the wrong managers can
addresses in a small city during a single expect to have
day when a total of 200,000 letters were 1,000 letters
delivered. What is the DPMO in this i
incorrectly
tl sentt
to the wrong
situation? address.

200
DPMO  x 1,000,000  1, 000
 1  x 200,000
Cost of Q
Quality:
y What might
g that DPMO mean in terms
of over-time employment to correct the errors?
17
 Six Sigma Quality: DMAIC Cycle
• Define, Measure, Analyze, Improve, and Control
(DMAIC)
• Developed by General Electric as a means of
focusing effort on quality using a methodological
approach
• Overall focus of the methodology is to understand
and achieve what the customer wants
• A 6-sigma program seeks to reduce the variation in
the processes that lead to these defects
• DMAIC consists of five steps….
18
Six Sigma Quality: DMAIC Cycle
(Continued)
1. Define (D) Customers and their priorities

2. Measure ((M)) Process and its p

performance

3. Analyze (A) Causes of defects

4. Improve (I) Remove causes of defects

5. Control (C) Maintain quality

19
E
Example
l to
t illustrate
ill t t theth process…
• W
We are the
th maker
k off this
thi cereal.
l
Consumer reports has just published an
article that shows that we frequently have
less
ess than
t a 155 ounces
ou ces of
o cereal
ce ea in a box.
bo .
• What should we do?

20
 Step 1 - Define
• Wh
Whatt is
i the
th critical-to-quality
iti l t lit
characteristic?
• The CTQ (critical-to-quality)
characteristic in this case is the weight of
the cereal in the box.

21
 2 - Measure

• How would we measure to evaluate

the extent of the problem?

• What are acceptable limits on this

measure?

22
 2 – Measure (continued)
• Let’s assume that the government says
that we must be within ± 5 percent of the
weight
i ht advertised
d ti d on the
th box.
b
• Upper Tolerance Limit = 16 + .05(16) =
16 8 oounces
16.8 nces
• Lower Tolerance Limit = 16 – .05(16) =
15 2 ounces
15.2

23
 2. Measure (continued)

• We go out and buy 1,000 boxes of cereal

and find that they weight an average of
15.875 ounces with a standard deviation
of .529
529 ounces
ounces.
• What percentage of boxes are outside the
t l
tolerance limits?
li it ?

24
 Process
Lower Tolerance Mean = 15.875
15 875 Upper
pp Tolerance
= 15.2 Std. Dev. = .529 = 16.8

What percentage of boxes are defective (i.e.

(i e less than 15.2
15 2 oz)?

Z = (x – Mean)/Std. Dev. = (15.2 – 15.875)/.529 = -1.276

NORMSDIST(Z)
O S S ( ) = NORMSDIST(-1.276)
O S S ( 1 2 6) = .100978
1009 8

Approximately, 10 percent of the boxes have less than 15.2

Ounces of cereal in them!

25
 Step 3 - Analyze - How can we
improve the capability of our cereal
box filling process?

– Decrease Variation
– Center Process
– Increase Specifications

26
S 4 – Improve
Step I – How
H good d iis good
d enough?
h?
Motorola’s “Six Sigma”

– 6 minimum from process center

to nearest spec
p
12

6

3 2 1 0 1 2 3

27
 Motorola’s “Six Sigma”
• Implies
I li 2 ppB
B “bad”
“b d” with
ith no process shift
hift
• With 1.5 shift in either direction from
center (process
( will
ill move),
) implies
i li 3.4 34
ppm “bad”.
12

3 2 1 0 1 2 3

28
 Step 5 – Control
• Statistical Process Control (SPC)
– Use data from the actual process
– Estimate distributions
– Look
L k att capability
bilit - is
i goodd quality
lit
possible
– Statistically
St ti ti ll monitorit the
th process
over time

29
 Analytical Tools for Six Sigma and Continuous

Improvement: Flow Chart

Material
No,
Received Continue…
Inspect
from
Material for Defects
Supplier
Defects found?

Yes

Can be used
sed to
find quality Return to
problems Supplier
for Credit
 Analytical Tools for Six Sigma and Continuous
I
Improvement: t Run
R Chart
Ch t
Can be used to identify
when equipment or
processes are not
behaving according to
Diameter

specifications
0.58
0.56
0 54
0.54
0.52
0.5
0 48
0.48
0.46
0.44
1 2 3 4 5 6 7 8 9 10 11 12
Time (Hours)
31
 Analytical Tools for Six Sigma and Continuous Improvement:
Pareto Analysis

Can b
C be used
d 80%

to find when
80% of the
problems quency
may be
Freq

attributed to
20% of the
causes

Design Assy. Purch. Training

Instruct
Instruct.
Analytical Tools for Six Sigma and Continuous
Improvement: Checksheet
Can be used to keep track of
defects or used to make sure
people collect data in a
Monday correct manner
Billing Errors

Wrong Account

Wrong Amount

A/R Errors

Wrong Account

Wrong Amount
 Analytical Tools for Six Sigma and Continuous
I
Improvement: t Histogram
Hi t

Can be used to identify

y the frequency
q y of q
quality
y
umber of Lots
s

defect occurrence and display quality

performance
Nu

0 1 2 3 4 Defects
D t R
Data Ranges in lot
 Analytical Tools for Six Sigma and Continuous
Improvement: Cause & Effect Diagram

Possible causes: The results

or effect
Machine Man

Environment Effect

Method Material

Can be used to systematically track backwards to

find a possible cause of a quality problem (or
effect)
 Analytical Tools for Six Sigma and Continuous
Improvement: Control Charts

Can be used to monitor ongoing production process

quality and quality conformance to stated standards of
quality

1020
UCL
1010
1000
990
980
LCL

970
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
36
Statistical Q
Qualityy Control (SQC)
( Q )
The application
pp of statistical techniques
q
to control quality. Often the term
statistical p
process control is use
interchangeably with statistical quality
control,, although
g statistical q
qualityy
control includes acceptance sampling as
well as statistical p
process control
37
 Routes in SQC
Statistical
Quality Control

Process Acceptance
Control Sampling

Variables Attributes
Variables Attributes
Charts Charts

38
 Quality Characteristics
Variables Attributes
• Characteristics that you • Characteristics for which
measure,
easu e, e.g., we
weight,
g t, you focus on defects
length • Classify products as either
• Mayy be in whole or in ‘good’
good or ‘bad’,
bad , or count #
fractional numbers defects
• Continuous random – e.g., radio works or not
variables • Categorical or discrete
random variables

39
 Statistical Process Control (SPC)

• S
Statistical
i i l technique
h i usedd to ensure process is
i
making product to standard
• All process are subject to variability
– Natural causes: Random variations
– Assignable causes: Correctable problems
• Machine wear, unskilled workers, poor material
• Objective: Identify assignable causes
• Uses process control charts
40
 Process Control:
Three Types of Process Outputs
(a) In statistical control and
capable of producing within
control limits. A process with
only natural causes of
variation and capable of
Frequency producing within the specified
control limits.

Lower control limit Upper control limit

(b) In statistical control, but not

capable of producing within control
limits. A process in control (only
natural causes of variation are
present) but not capable of
producing within the specified
control limits; and

(c) Out of control. A process out of

control having assignable causes of
Size variation.
W i ht length,
Weight, l th speed,
d etc.
t

41
 Theoretical Basis
of Control Charts
Central Limit Theorem
As sample size sampling distribution
gets becomes almost
large normal regardless of
enough, population distribution.

X
X
42
 Theoretical Basis
of Control Charts
Central Limit Theorem

Mean Standard deviation


X   x  x
n

X
X  
43
The Relationship Between Population
and Sampling Distributions
Three population distributions
Distribution of sample means
Beta

Mean of sample means  x

Standard deviation of x
Normal  x 
the sample means n

Uniform

 3 x  2  x  1 x x   x  2  x  3 x
(mean)
95.5% of all x fall within  2  x

99.7% of all x fall within  3  x

44
Sampling Distribution of Means
and Process Distribution
Sampling
S li
distribution of the
means
Process
distribution of
the sample

xm
( mean )

45
 P
Process C
Controll Charts
Ch
Plot off S
Sample Data O
Over Time
i
70
60
Samplle Value
e

Sample
50 Value
40 UCL
30 Average
20
10 LCL
0
1 5 9 13 17 21
Time

46
 Control Chart Purposes
• Show
Sh changes
h iin data
d pattern
– e.g. trends
• Make corrections before process is out of control
• Show causes of changes in data
– Assignable causes
• Data outside control limits or trend in data
– Natural causes
• Random variations around average

47
 Theoretical Basis
off C
Control
t l Charts
Ch t
Properties of normal distribution
95.5% of al l x fall 99.7% of al l x fall
within  2 within  3
x x

x
x  
48
 Control Chart Types
Continuous
C Categorical or Discrete
C
Numerical Data Control Numerical Data
Charts

Variables Attributes
Charts Charts

R X P C
Chart Chart Chart Chart

49
S i i l Process
Statistical P Control
C l Steps
S

Produce Good No
Start
Provide Service

Assign.
Take Sample Causes?
Yes
Inspect Sample Stop Process

Create
Find Out Why
Control Chart

50
 X Chart
• Type
T off variables
i bl controll chart
h
– Interval or ratio scaled numerical data
• Shows sample means over time
• Monitors p
process average
g
• Example: Weigh samples of coffee &
compute means of samples; Plot

51
 X Chart
Control Limits
Mean Factor that
UCL x  x  A  R increases when
sample size is
large (Refer to
LCL x  x  A  R Sample table)
Range at
Sample
Time i
n Mean at
 xi Time i n
 R i
x  i  i 1
R 
n n
# Samples
52
Factors for Computing Control
Chart Limits
Sample Mean Upper Lower
Size, n Factor, A2 Range, D4 Range, D3
2 1.880 3.268 0
3 1.023 2.574 0
4 0.729 2.282 0
5 0 577
0.577 2 115
2.115 0
6 0.483 2.004 0
7 0 419
0.419 1 924
1.924 0 076
0.076
8 0.373 1.864 0.136
9 0 337
0.337 1 816
1.816 0 184
0.184
10 0.308 1.777 0.223 53
0.184
 R Chart
• Type of variables control chart
– Interval or ratio scaled numerical data
• Shows
Sh sample
l ranges over time
ti
– Difference between smallest & largest values in
inspection sample
• Monitors variability in process
• Example:
E l Weigh
W i h samples l off coffee
ff &
compute ranges of samples; Plot

54
 R Chart
Control Limits
UCL R  D 4R
Range Factor (Refer
to table))
LCL R  D 3R

n Sample Range at
 Ri Time i
R  i 1
n # Samples
55
 X Chart and R chart Example
p
Plot a Mean chart and an R chart for the following tread wear
data. Comment the quality level and what action should be taken.
Sample Tread Wear
1 31 38 22
2 26 36 41
3 28 33 15
4 17 27 29
5 20 38 44
6 55 20 43
7 40 29 37
8 43 26 14
9 15 28 35
10 11 15 21 56
 Steps to Follow When Using
Control Charts - continued
• Investigate points or patterns that indicate the
process is out of control. Assign causes for the
variations.
• Collect additional samples and revalidate the control
limits.

57
Which Charts to use?

58
 Process Capability
The ability of a process to produce parts that
conform
f to ((engineering)
i i ) specifications.
ifi i
Process Capability relates to the inherent
variability
i bili off a process that
h isi in
i a state off
statistical control

59
 Process Capability
• Tolerances
– design specifications reflecting product
requirements
• Process capability
– range of natural variability in a process what we
measure
easu e with
w t control
co t o charts
c a ts
 P
Process C
Capability
bilit
Design
Specifications
p

(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
 P
Process C
Capability
bilit (cont.)
( t)
Design
Specifications
p

(c) Design specifications

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

((d)) Specifications
p greater
g
than natural variation,
but process off center;
capable but some output
will not meet upper
specification.
specification
Process
Process Capability Measures

Process Capability Ratio

tolerance range
Cp = process range

upper
pp specification
p limit -
lower specification limit
=
6
 Computing Cp
Net weight specification = 9.0 oz  0.5 oz
Process mean = 8.80 oz
Process standard deviation = 0 0.12
12 oz

upper
pp specification
p limit -
lower specification limit
Cp =
6

9.5 - 8.5
= = 1.39
6(0.12)
 Process Capability Measures

Process Capability Index

=
x - lower specification limit
,
3
Cpk = minimum =
upper specification limit - x
3
 Computing Cpk
Net weight specification = 9.0 oz  0.5 oz
Process mean = 8.80 oz
Process standard deviation = 0 0.12
12 oz
=
x - lower specification limit
,
Cpk = minimum 3
3
=
upper specification limit - x
3

8.80 - 8.50 9.50 - 8.80

= minimum , = 0.83
3(0 12)
3(0.12) 3(0 12)
3(0.12)
 Meanings of Cpk Measures
Cpk = negative number

Cpk = zero

Cpk = between 0 and 1

Cpk = 1

Cpk > 1

67
 p Chart
• Type of attributes control chart
– Nominally scaled categorical data
• e.g. good-bad
• Shows
S ows % oof nonconforming
o co o g items
te s
• Example: Count # defective chairs & divide
by total chairs inspected; Plot
– Chair is either defective or not defective

68
 p Chart
Control Limits
p (  p )
UCL p  p  z z = 2 for 95.5% limits;
n z = 3 for 99.7% limits

p (  p )
LCL p  p  z
n
# Defective Items
in Sample i
k k
 ni  xi
n  i 
and p  i 1
k k
Total Size of
 ni
i 1 sample i
No. of sample
Size of each sample 69
 p Chart Example
p-Chart

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

20 samples of 100 pairs of jeans

 p-Chart Example (cont.)

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(1 - 0.10)
0 10)
LCL = p - z = 0.10 - 3
n 100
LCL = 0.010
 0.20

0.18 UCL = 0.190

0 16
0.16

0.14

ective
p
p-Chart
oportion defe
0 12
0.12

Example 0.10
p = 0.10

0.08
( t)
(cont.)
Pro

0.06

0.04

0.02 LCL = 0.010

2 4 6 8 10 12 14 16 18 20
Sample number
 c Chart
• Type of attributes control chart
– Discrete quantitative data
• Shows number of nonconformities (defects) in a
unit
– Unit may be chair, steel sheet, car etc.
– Size of unit must be constant
• Example: Count # defects (scratches, chips etc.) in
eachh chair
h i off a sample
l off 100 chairs;
h i Plot
Pl t

73
 c Chart
Control Limits
UCL c  c   c Use 3 for 99.7%
limits

LCL c  c   c

# Defects in
k
 ci Unit i

c  i 1 # Units Sampled
k
74
 c-Chart
Ch t (cont.)
( t)
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
1 99
 24
UCL = 23.35
21

18

of defects
c = 12.67

15
c Chart
c-Chart Number o

12

(cont.) 9

3 LCL = 1.99
1 99

2 4 6 8 10 12 14 16
Sample number
What should we do about these?

77
 What Is
Acceptance Sampling?
• Form of quality testing used for incoming
materials or finished goods
– e.g. purchased material & components
• Procedure
– Take one or more samples at random from a lot
(shipment) of items
– Inspect each of the items in the sample
– Decide whether to reject the whole lot based on the
inspection results
78
 What Is an
Acceptance Plan?
• Set of procedures for inspecting incoming
materials or finished goods
• Identifies
– Type of sample
– Sample size (n)
– Criteria (c) used to reject or accept a lot
• Producer (supplier) & consumer (buyer)
must negotiate
79
 Operating Characteristics (OC)
Curve
• Shows how well a sampling plan
discriminates between ggood & bad lots
(shipments)
• Shows the relationship between the
probability of accepting a lot & its quality

80
 OC Curve
100% Inspection
P(Accept Whole Shipment)

100%
Keep whole Return whole
shipment shipment

0%
0 1 2 3 4 5 6 7 8 9 10
Cut-Off % Defective in Lot
81
 OC Curve with Less than 100%
S
Sampling
li
P(A
P(Accept
t Whole
Wh l Shipment)
Shi t)
Probability is not 100%: Risk of
100% keeping bad shipment or
returning good one.
Keep whole
shipment
Return whole
shipment

0%
0 1 2 3 4 5 6 7 8 9 10
Cut-Off % Defective in Lot
82
 AQL
Q & LTPD
• Acceptable Quality Level (AQL)
– Is the poorest level of quality we are willing to accept
– Producer (supplier) does not want lots with fewer defects
than AQL rejected
– For example, if an acceptable quality level is 20 defects in a
lot of 1000 parts, then AQL is 20/1000 = 2% defective
• Lot
L tTTolerance
l Percent
P t Defective
D f ti (LTPD)
– Is the quality level of a lot we consider bad
– Consumer (buyer) does not want lots with more defects than
LTPD accepted
– For example, if an unacceptable quality level is 70 defects in
a lot of 1000 parts,
parts then the LTPD is 70/1000 = 7% defective
83
 Producer’s & Consumer’s Risk
• Producer's risk ()
– Probability of rejecting a good lot
– Probability of rejecting a lot when fraction
defective is AQL
• Consumer's risk (ß)
– Probability of accepting a bad lot
– Probability of accepting a lot when fraction
defective is LTPD
84
An Operating Characteristic (OC)
Curve Showing Risks
100
 = 0.05 producer’s risk for AQL
95

75

Probability of
Acceptance
50

25

10
= 0.10 Percent
0 Defective
Consumer’s 0 1 2 3 4 5 6 7 8
risk for LTPD
AQL LTPD
Good Bad lots
Indifference zone
lots

85
 OC Curves for Different
Sampling Plans
P(Accept Whole Shipment)

100% n = 50, c = 1
n = 100,
100 c = 2

0%
0 1 2 3 4 5 6 7 8 9 10
AQL LTPD
% Defective in Lot
86
Average Outgoing Quality
(AOQ)
The expected average quality level of
g g pproduct for a ggiven value of
outgoing
incoming product quality

87
 Average
ve ge Ou
Outgoing
go g Quality
Qu y
Equation
( Pd )( Pa )( N  n )
AOQ 
N

Where: Pd = true percent defective of the lot

Pa = probability of accepting the lot
N = number of items in the lot
n = number of items in the sample

88