17/03/1431

Statistical Process Control

Dr. Ashraf S. Youssef

Assistant Professor,
Professor QMS Consultant
Industrial Engineering Dept.
Fayoum University, Egypt

Course Introduction
„ Welcome!
„ Instructor introduction.
„ Student introduction.

1
17/03/1431

Learning Objectives
„ Understand quality fundamentals.
fundamentals
„ Differentiate between QC, QA, and TQM.
„ Learn quality improvement tools.
„ Learn SPC and reliability.
„ Learn Cost of Quality.
„ Understand ISO 9000:2000 QMS.
„ Become aware of Excellence Models.
„ Apply quality to oneself & organization.

Our Objective

You Enjoy While You

Learn

2
17/03/1431

CHAPTER ONE

Introduction to
Statistical Quality
Control

Meaning Of “Statistical Quality

Control

Statistical With the help of numbers or data

We study the characteristics of the

Quality
process

In order to make it behave the

Control
way we want it to behave

3
17/03/1431

What is Quality?

What is Quality?
„ “fitness for use” (Juran)
„ “conformance to requirements”
(Crosby)
„ “the totality of features and
characteristics of a product or
service that bear on its ability to
satisfy given needs” (ASQ)

4
17/03/1431

Quality Aspects
„ Quality of design
„ All goods and services are produced in various
grades or various levels of quality.
„ Quality of Conformance
„ How well the product conforms to the
specifications and tolerances required by the
design
„ Examples
„ Car
„ Air condition

Quality Characteristics
„ Physical
„ Length, Weight, Voltage, Viscosity, etc….
„ Sensory
„ Taste, Appearance, Color, etc….
„ Time Orientation
„ Reliability, Maintainability, Serviceability,
etc….

5
17/03/1431

Quality Dimensions
Conformance Durability
Reliability Serviceability
Features
Performance Perceived
Quality
Product

Source: Total quality and organization development by Lindsay and Petrick

Quality and Productivity

„ Example:
p
„ 100 product are produced in a machine, 75% is good
and 25% is nonconforming. 60% out of 25% can be
reworked at $4 per part. The direct manufacturing cost
per part is approximately $20. Determine the
cost/good part.
„ If yyou know that,, a new process
p control procedure
p is
implemented, the 25% is reduced to 5%
nonconforming, and the 60% out of them can be
reworked. Determine the cost/good part after the
process control program is implemented.

6
17/03/1431

Quality Costs (1/4)

„ Prevention Costs
„ Costs associated with efforts in design and
manufacturing that are directed toward
the prevention of nonconformance.
„ Examples
„ Quality planning and engineering.
„ New products review.
„ Product/Process design

Quality Costs (2/4)

„ Appraisal Costs
„ Costs associated with measuring, evaluating, or
auditing products, components, and purchased
materials to ensure conformance to the standards
that have been imposed.
„ Examples
„ Inspection and test of incoming material.
„ Product inspection and test.
„ Materials and services consumed.
„ Maintaining accuracy of test equipment.

7
17/03/1431

Quality Costs (3/4)

„ Internal Failure Costs
„ Costs are incurred when products, components,
materials, and services fail to meet quality requirements
and this failure is discovered prior to delivery of the
product to the customer.
„ Examples
„ Scrap
p
„ Rework
„ Retest
„ Failure analysis
„ Downtime

Quality Costs (4/4)

„ External Failure Costs
„ Costs occur when the product does not
perform satisfactory after it is supplied to the
customer.
„ Examples
„ Complaint adjustment.
„ Returned product/material.
„ Warranty charges.
„ Liability costs.
„ Indirect costs.

8
17/03/1431

Quality & the Bottom Line

Income Statement

Right features Revenue

Competitive price
Higher Market share
- Cost Improved conformance
R d
Reducedd cycle
l time
ti
Improved warranty
= Profit Less waste

Quality Improvements
„ We may define quality
improvement as a reduction in
variability of processes and
products.
„ Variation can only be described in
statistical terms.

9
17/03/1431

QC, QA, & TQM

TQM … Organization
Difficult Time Soft Employees
Leadership
Culture
Business Results
QA … Business Processes
Procedures
Policies Responsibilities

SPC … Product
Off-line
On-line

From QC to TQM
TQM … Organization
Leadership Employees
Culture
Business Results
QA … Business Processes
Procedures
Policies Responsibilities

SPC … Product
Off-line
On-line

10
17/03/1431

CHAPTER TWO

Statistical Methods
Useful in Quality
Improvement

Statistics: The Technical

Language of Quality
„ Quality is achieved if your product,
product
service, and information are
consistently on, or close to, target.

11
17/03/1431

Key Terms
„ Specification/standard/requirement.
Specification/standard/requirement
„ Nominal/target value.
„ Defective vs. non-defective.

Examples: Standard,
Target, & Defective
„ Rod diameter spec.
spec is 2 +/-
+/ 0.05
0 05 inch
„ Satisfaction is above 3 on a scale from 1 to 5.

Example Target Standard Defective

Rod

Satisfaction

12
17/03/1431

Who Sets the Standard?

„ ________________
„ ________________
„ ________________
„ ________________
„ ________________
„ ________________

Types of Standards
„ Two sided
Two-sided.
„ The shaft diameter (inches).
„ One-sided … left is bad.
„ Time to failure (hrs) of electrical insulation
material.
„ One-sided … right is bad.
„ Cars exceeding the speed limit (mph).

13
17/03/1431

Defective & Non-Defective

Categories of Data
Category Meaning Example
Nominal Data is a code. 0=Defective,
1=Non-defective.
Ordinal Data is a rank. Data order 5=A, 4=B, 3=C,
is important. 1=D, 0=E.
Interval Data is a measure; no Temperature.
true zero.
Ratio Data is a measures; there Weight.
is a true zero.

14
17/03/1431

Types of Data
Type Meaning Example
Continuous Infinite values on an Temperature,
interval. Weight.
Discrete Finite values on an Student letter
interval. grade, Traffic-light
color.
color

Example: Categories &

Types of Data
Data Category Type
Customer satisfaction survey

Service time

Go/No-Go gauge

Percentage of purity

15
17/03/1431

Statistical Overview
P
Population
l ti
A large No. of N
Mean = µ
Standard Deviation = σ

Sample
A small No. of n X
Sample Mean =
Sample Standard
Deviation = s

Descriptive Statistics
„ Centrality Measures
¾ Mode
¾ Median
¾ Mean
„ Dispersion Measures
¾ Range
¾ Variance
¾ Standard Deviation
¾ Coefficient of Variation

16
17/03/1431

Accuracy: Measures of Centrality

„ Mean = add values ÷ number of values

„ Median: is the value in the middle of
ascendingly ordered data
„ Mode: is most frequently occurring
value(s)

Example:
Measures of Centrality
„ Purpose: learn how to calculate
measures of centrality
„ Method: Hand-calculate the mean,
median, & mode of boxes weights (kg):
„ 14,, 16,, 12,, 18,, 14,, 10,, 11,, 14,, 23,, 18.
„ Time: 25-30 minutes.

17
17/03/1431

Consistency: Measures of Spread

„ Range = maximum value – minimum
value.
„ Variance: is the squared average
deviations.
„ Standard Deviation: is the average
deviations.
„ Coefficient of Variations : is the ratio
between Standard deviation and Average.

Example:
Measures of Spread
„ Purpose: Learn how to calculate
measures of spread.
„ Method: Hand-calculate the range,
variance, & standard deviation of boxes
g
weights (kg):
( g)
„ 14, 16, 12, 18, 14, 10, 11, 14, 23, 18.

18
17/03/1431

Example:Why Need Both

Types of Measures?
„ If the passing grade is 75,
75 evaluate the
performance of the following 3-student
classes:
„ Class 1 exam score: 0, 50, 100
„ Class 2 exam score: 48, 50, 52
„ Class 3 exam score: 78,
78 80,
80 82
„ Rank the 3 classes from best to worst.
Why?

Example:Why Need Both

Types of Measures?
Class 1 2 3

Accuracy

Precision

„ Rank is ____________________

19
17/03/1431

Accuracy & Precision

____________ ____________ ____________ ____________

____________ ____________ ____________ ____________

Exercise: Centrality
Measures
„ Calculate the sample mode,
mode
median and the mean of
the following data set.
„ Set A: 20, 20, 18, 17, 15,
17 14,
17, 14 17,
17 12,
12 10
„ Set B: 10, 20, 12, 13, 13,
16, 17, 12, 13, 17, 22

20
17/03/1431

Exercise: Dispersion
Measures
„ Calculate the sample range
range,
variance, standard deviation
and coefficient of variation.
„ Set A: 20, 20, 18, 17, 15,
17 14,
17, 14 17,
17 12,
12 10
„ Set B: 10, 20, 12, 13, 13, 16,
17, 12, 13, 17, 22

Exercise: Descriptive
Measures
„ The following Sample Percents
Isooctane
measurements 1 3.83
were obtained for 2 3.94

the analysis of 3 3.88

4 3.97
isooctane in 5 3.90
gasoline
gasoline. 6 3.94
• Calculate X , S, and the C.V

21
17/03/1431

Estimate of Standard Deviation

from a Pair of Results (X1 and X2)
„ IIn case off a single
i l duplicate
d li t pair i off
measurements, the equation for the standard
deviation can be significantly reduced as
follows
D X − X2
S= = 1
n 2

Estimate of Standard Deviation

from Duplicate Measurements
„ For a set of measurements consisting of pairs
of duplicate results, the standard deviation
can be estimated as follows:

S=
∑D 2

2k
ƒ Where K = the number of sets of
duplicates

22
17/03/1431

Graphical Representations
„ Stem-and-leaf
Stem and leaf plot
„ Histogram

„ Box-plot

Data for boxes weights (in kg):

14, 16
14 16, 12,
12 18,
18 14
14, 10,
10 11,
11 14
14, 20,
20 18,
18
06, 10, 13, 24, 21, 18, 14, 15, 14, 16,
25, 18, 16, 20, 10, 24, 14, 12, 10, 14

Example:
Stem-and-Leaf Plot

(0 – 4) 0
(5 – 9) 0* 6
(10 – 14) 1 402340014442404
(15 – 19) 1* 68688586
(20 – 24) 2 40140
(25 – 29) 2* 5

23
17/03/1431

Example:
Ordered Stem-and-Leaf Plot

(0 – 4) 0
(5 – 9) 0* 6
(10 – 14) 1 000012234444444
(15 – 19) 1* 56668888
(20 – 24) 2 00144
(25 – 29) 2* 5

Example:
Distribution Table
„ n = 30,
30 cells = SQRT(n) = 5 or 6,
6 cell
width = (25 – 6)/5 = 4
Cell Tally Frequency
6 to 9
10 to 13
14 to 17
18 to 21
22 to 25

24
17/03/1431

Example:
Histogram
Freq.

Weight
6 9 13 17 21 25

Example:
Box Plot
„ Q = 0.25 x 30 = 7.5
Q1
„ Q2 = 0.50 x 30 = 15
0 „ Q3 = 0.75 x 30 = 22.5
0* 6
1 000012234444444
1* 56668888
2 00144
2* 5

25
17/03/1431

Example:
Box Plot

6 8 10 12 14 16 18 20 22 24 26

Q1 Q2 Q3

Fish--bone or Cause & Effect Diagram

Fish
Possible causes

Machine Man

Environment Effect

Method Material
Note
Track backwards to find possible cause of Q problem (or effect).

26
17/03/1431

Fish--bone or Cause & Effect Diagram

Fish
Solve the problem of Late deliveries of XYZ Company

Possible causes

Machine Man

Environment Effect

Method Material
Note
Track backwards to find possible cause of Q problem (or effect).

Check sheet
track defects or collect data
Wednesday
Invoicing errors

Wrong Account

Wrong Amount

Accounts payable
errors

Wrong Account

Wrong Amount

27
17/03/1431

Pareto Analysis (20

(20--80
80%)
%)

Steps of making the Pareto Analysis

„ Select the measurement parameter
„ Rank the parameter in descending order
„ Calculate the percentage
„ Calculate the cumulative percentage

„ Construct a bar graph for each category and a line graph for the
cumulative percentage .

Pareto Analysis ( Example, Step1

Step1 )

At th
the endd off 1st Quarter
Q t off 2003,
2003 XYZ Company
C surveyedd the
th
performance of 4 suppliers and put the result in the following table :
Suppliers
Problems A B C D TOTAL
Incorrect Invoice 3 4 0 2 9
L t D
Late Deliveries
li i 3 4 5 6 18
Damaged Material 2 5 1 2 10
Short Delivery 2 5 3 3 13
Totals 10 18 9 13 50

28
17/03/1431

Pareto Analysis (Example, Step 2 )

% of each problem was calculated , and the result put in descending

order in the following table
Problems Suppliers % Accum.
A B C D TOTAL %
Late Deliveries 3 4 5 6 18 36% 36%
Short Delivery 2 5 3 3 13 26% 62%
Damaged Material 2 5 1 2 10 20% 82%
Incorrect Invoice 3 4 0 2 9 18% 100%
Totals 10 18 9 13 50 100 % 100%

Pareto Analysis ( Example, Step3)

Step3
120%
100%
100%
82%
80%
62%
60%
36%
40%
26%
20% 18%
20% 36%

0%
Late Delivery Short Delivery Damaged In correct
Material Invoice

29
17/03/1431

Hypothesis Testing
„ Common questions that can be answered
using the null hypothesis approach:

¾ Is the mean of the data set significantly different

from the true value?
¾ Are the means of two different sets significantly
different?
¾ Are the precisions of two different data sets
significantly different?

Test 1: Comparing the Mean With

the True Value

„ The
h largest
l difference
d ff that
h could
ld be
b
expected as a result of indeterminate
error is given by: s
± t( )
n
− s
H o : ( X − μ ) p± t ( )
n
− s
H1 : ( X − μ ) f± t ( )
n

30
17/03/1431

Test 2: Comparing Two Means

„ There are two occasions when we wish
to determine if two independently
obtained results are essentially the same.
− − n1 + n 2
H o : ( X 1 − X 2 ) p ± ts p
n1 n 2

− − n1 + n 2
H 1 : ( X 1 − X 2 ) f ± ts
t p
n1 n 2
Where Sp is the pooled standard deviation of the two samples.

( n1 − 1) S 1 + ( n2 − 1) S 2
2 2

Sp =
n1 + n2 − 2

Exercise: Comparing Two Means

„ Compare the two
means, of the
following data set.
„ Set A: 20, 20, 18,
17,, 15,, 17,, 14,, 17,,
12, 10
„ Set B: 10, 20, 12, 13,
13, 16, 17, 12, 13,
17, 22

31
17/03/1431

Test 3: Comparing Two Precisions

„ To determine if the standard deviation
from one data set is significantly different
from the standard deviation of another
data set H o : Fc p Ft
H 1 : Fc f Ft
2

Fc =
Var 1
= S 1
2
Var 2 S 2

•The larger variance is always numerator.

•Ft is the expected tolerance base on Normal distribution.

Exercise: Comparing Two Precisions

„ A ship of copper ore was purchased by a

local metal refiner. The analysis
certificate made out while the ship was
being loaded, show that %Cu =14.66
with a standard deviation of 0.07% for 5
measurements. When the ore arrived at
the refinery, it was analyzed and the
following results were obtained: %Cu =
14.58, 14.61, 14.69, 14.69, and 14.64.
Should the refiner accept the ore?

32
17/03/1431

Normal Distribution

Standard Deviation (σ )

Mean ( μ )

Normal Distribution Principle

99.7 Percent
95 Percent
68 Percent

Y - 3s Y -2 s Y-s Y Y+s Y +2s Y +3s

33
17/03/1431

Process Capability
Lower Upper
Specification Specification

Process variability
matches specifications
Lower Upper
Specification Specification

Process variability well Lower Upper

within specifications Specification Specification

Process variability
exceeds specifications

CHAPTER THREE

Methods and Philosophy

of Statistical Process
Control

34
17/03/1431

Statistical Methods for

Quality Improvement
„ Acceptance Sampling
„ Statistical Process Control Charts (SPC)
„ Design of Experiments
„ Process Capability studies

Statistical Quality Control

Can be applied to:
„ Engineering Problems

„ Inspection Problems

„ Operating Problems

„ Management
M t Problems
P bl
„ Accounting and Clerical Problems

35
17/03/1431

Process
„ Is any set of conditions
conditions, or set of
causes, which work together to produce
a given results.
Decision Variables

Input Process Output

Uncontrollable inputs

Control Chart
Purpose: To monitor and analyze variation of a process to
determine whether the variation is normal or abnormal.
510 UCL
480

450
LCL
420

390

360
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30

US System 3.09 Sigma Limits

ES System 3.00 Sigma Limits

36
17/03/1431

Control Charts Type

„ Variables Control Chart
„ X-bar and R Charts
„ X-bar and S Charts
„ MA Control Charts
„ Attributes Control Charts
„ P-Chart
P Chart
„ np-Chart
„ C-Chart
„ U-Chart

Type of Variations in Production

„ Within-Piece
Within Piece Variation
„ Piece-To-Piece Variation
„ Time-To-Time Variation

37
17/03/1431

Sources of Variation
„ Equipment
„ Material
„ Environment
„ Operator
„ I
Inspection
ti Activity
A ti it

Causes of Variation
Cause of Variation

Assignable Causes Chance Causes

They are These causes

Large and are difficult
readily to detect
identified

38
17/03/1431

The Control Chart Method

„ A graphical record of the quality of a
particular characteristics
„ It shows if the process is in a stable
state with chance causes of variation or
not.

CHAPTER FOUR

Variable Control Charts

39
17/03/1431

Variables Control Charts

„ X-bar
X bar and R Charts
„ X-bar and S Charts
„ MR Control Charts

Objectives of Variable Control Charts

„ Quality Improvement
„ Determine the Process Capability
„ Setting Effective Specification Limits
„ Deciding to release the product to the
next phase or not

40
17/03/1431

Control Chart Building Technique

„ Select the Quality Characteristics
„ Choose the Rational Subgroup
„ Collect the Data
„ Determine the Trial Central Line and the
Control Limits
„ Establish the revised Central Line and
the Control Limits
„ Achieve the Objectives

Selection of Quality Characteristics

„ Must be measurable and expressed in
numbers
„ Seven basic units were used
a) Lengths e) Substance
b) Mass f) Time
c) Current g) Length intensity
d) Temperature
„ Variables can be treated as Attribute
„ One Characteristic only is selected

41
17/03/1431

Choosing the Rational Subgroup

„ The Inspection Cost
„ Test (Inspection) Type
„ Pre-Control Rule
„ Keep The results in tighter and sensitive
Control Limits
Usually sample of 5 is taken and
sampling frequency is every half
hour.

Establish X-bar and R Charts

σx =σ / n
UCL= X + zσx CL= x LCL = X − zσ x
-OR-
UCL= X + A2 R CL = x LCL = X − A2 R

(R) ‫ﺧﺮﻳﻄﺔ اﻟﻤﺪى‬

UCL = D4 R CL = R LCL= D3R

42
17/03/1431

Establish X- bar and S Charts

s
X Chart S = ∑ n

UCLX = X + A3σ UCL = B6σ

S
CL S =C4 σ
LCLX = X − A3σ LCLS = B5 σ

Use S, B4, and B3 if sigma is unknown, and CLs = S

Establish MR (Individual) Chart

In so
some
e cases, especially
espec a y in dest
destructive
uct e test, The
e
QE can’t take more than one sample each time.

UCLX = X + 3
MR UCLMR = D4 MR
d2

MR LCL MR = D3 MR
LCLX = X − 3
d2

43
17/03/1431

State of Control
1. Process in Control
1
2. Process out of Control
A point is outside the 3σ
limits Plus AT & T rules.
A nonrandom behavior is
observed in the points

Interpretation of the Control Charts

Western Electric Rules (AT&T Rules) are

used to decide about out of control
condition.
There are 10 rules were used.

44
17/03/1431

AT&T Rules
„ One or more points outside Control Limits
„ Two of three consecutive points outside of 2σ
limits
„ A run of 8 consecutive points on one side of
the center line
„ Six p
points in a row steadilyy increasing
g or
decreasing

AT&T Rules Continued

„ Eight points in a row in zone C or beyond
„ Eight points in a row in both sides of the
center line with none in zone C
„ Four points in a row alternating up and down
„ Fifteen points in a row in zone C
„ An unusual and nonrandom pattern

45
17/03/1431

Two of three consecutive points outside of

2σ limits

A run of 8 consecutive points on one side

of the center line

150 150
140 140
130 130
120 120
Thickness

Thickness

110 110
100 100
90 90
80 80
70 70
60 60
50 50

0 5 10 15 20 25 0 5 10 15 20 25
Sequence Sequence
A B
Which process has better quality?

46
17/03/1431

An unusual and nonrandom pattern

An unusual and nonrandom pattern

200 200
180 180
160 160
Thickness

140 140
120 120
100 100
80 80
60 60
T

40 40
20 20
0 0

Time

47
17/03/1431

An unusual and nonrandom pattern

150 150
140 140
130 130
120 120
Thickness

Thickness
110 110
100 100
90 90
80 80
70 70
60 60
50 50

0 5 10 15 20 25 0 5 10 15 20 25

A Sequence B Sequence

Which process has better quality?

An unusual and nonrandom pattern

200 200
180 180
160 160
Thickness

140 140
120 120
100 100
80 80
60 60
40 40
20 20
0 0

Time

48
17/03/1431

CHAPTER FIVE

Attribute Control Charts

SPC “Attributes Charts”

Types of Attributes
„ Non-measurable Characteristics

„ Measurements that could not be made

due-to:
a) Time b) Cost
c) Easier to use “go-no-go” gage

49
17/03/1431

Non Conformity
Definition

Departure of quality characteristics from

required

The Objectives of Attribute

Charts
„ Determine the average quality level
„ Bring to the attention of any changes in
the average level
„ Improve the product quality
„ Evaluate the quality performance
„ Determine the acceptance criteria of a
product

50
17/03/1431

Attribute Charts Type

„ P-Chart
P Chart
„ np-Chart
„ C-Chart
„ U-Chart

P- Chart
„ P Chart can be produced to measure the
P-
quality produced by:
a) A Work Center
b) A Department
c) A Shift
d) The entire Plant
„ Also used to report the quality performance
of an operator, a group of operators, or
management.

51
17/03/1431

P-Chart Construction
„ Constant Sample Size
At least 25 subgroups should be collected

P =
∑ nP
n
Where nP is the number of Nonconforming units in each subgroup
n is the sample size for each subgroup
P(1−P)
UCL= P +3
P(1− P) LCL= P −3
n n

If LCL is negative, replace it with zero value

P-Chart Construction
„ Variable Sample Size
P =
∑ P
N

where P = Fraction of nonconforming in a single subgroup

N = Number of subgroups

P(1−P)
UCL= P +3
P(1− P) LCL= P −3
n n

where n is the average sample size of the subgroups

52
17/03/1431

nP - Chart
„ nP-Chart
Chart is similar to P-Chart
Chart
„ It is easier for personnel to understand
„ Sample size must be constant

Center Line = nP
C t l Limits
Control Li it = nPP±3 nPP(1− P)

C-Chart
Example for C-Chart
„ No. of nonconforming rivets on an airplane

„ Imperfections in a large roll of paper

„ Rust sports on a steel sheet

„ Billing errors

„ Errors in forms

„ Seeds or air pocket in a glassware

53
17/03/1431

C-Chart Construction
Center Line = C
UCL = C + 3 C

LCL = C − 3 C

Where C = Average
Wh A number
b on
nonconformities for a number of
subgroups

U-Chart
If subgroup size varies
varies, U-Chart
U Chart would be
appropriate to use.
U=
C
U =
∑C
n ∑n
U U
UCL = U + 3
n LCL = U − 3
n
Where C = Count of nonconformities in a
subgroup
n = Number of unit inspected in a
subgroup

54
17/03/1431

A Quality Rating System of

Nonconformities (MIL-STD-105D)

1.
1 Critical Nonconformities
2. Major Nonconformities
3. Minor Nonconformities

A weight
g of 9,, 3 and 1 p points are
assigned to critical, major and minor
nonconformities.

Guidelines for Implementing Control

Charts
„ Determine which process characteristics to
control
„ Determine where the charts should be
implemented
„ Apply control charts to important
characteristics
„ Add new charts if necessary
„ Implement and maintain control charts as
close to the work center as possible

55
17/03/1431

CHAPTER SIX

ACCEPTANCE SAMPLING

When are Using Acceptance Sampling

„ When testing
Wh t ti isi destructive.
d t ti
„ When the cost of 100% inspection is extremely
high.
„ When 100% inspection is not technologically
feasible or need much time.
„ When the vendor has an excellent quality history.
„ When there are potentially serious product liability
risks.
„ When there are many items to be inspected and
the inspection error rate is sufficiently high.

56
17/03/1431

Types of Sampling Plans

„ A single
i l sampling
li Plan
Pl
„ N = sample size, and c = acceptance number
„ Double sampling plans
„ N1 = sample size on the first sample, c1 = acceptance
number of the first sample.
„ N2 = sample size on the second sample, c2 = acceptance
number for both samples.
„ A multiple sampling plan

Operations With the Sampling Plan

Inspect a random sample of n = 50 from a lot.
d1=number of observed defects

Accept D1<=C1=1 D1>C2=3 Reject

the lot the lot

Inspect a random sample of n2 = 100 from a lot.

d2=number
b off observed
b dddefects
f t

D1+D2<=C2=3 D1+D2>C2=3
Accept Reject
the lot the lot

57
17/03/1431

MILITARY Standard 105 D (ANSI/ASQC Z1.4)

Procedure

1. Choose the
Ch th acceptance
t Quality
Q lit Level
L l (AQL).
(AQL)
2. Choose the inspection level.
3. Determine the lot-size.
4. Find the appropriate sample size code letter
5. Determine the appropriate type of sampling plan to
use.
6. Enter the appropriate table to find the type of plan
to be used.
7. Determine the corresponding normal and reduced
inspections plans to be used.

The Cost Of Quality

Q y
(COQ):

Concepts

58
17/03/1431

Definition: Cost of Quality.

Cost to Improve Quality +

Costs due to Lack of Quality
(not right at the first time).

Quality Product Costs More ??

„ Buy a 500 $ car „ Buy a 100 $ car
breaks (original)! b k (clone)!
breaks ( l )!
„ No replacement. „ 100 SR replacement.
„ 80,000 $ car loss in
„ Good performance.
an accident.
„ Full control. „ 15 days hospitalized.
„ A peace of mind. „ Death of a child.
Total cost =. Total cost =.
I’ve been doing it for years nothing
happened

59
17/03/1431

The Way We Work (W. Conway) :

15%
Necessary
30%
Unnecessary
20% (Rework)
Value-Added
Value Added

10%
25%
25%
Unnecessary
Not Working
(Reports & Others)

Tolerance and Acceptability

„ The Japanese target.

„ Taguchi loss function.

„ The American target.

„ Any where is fine (Deming).

60
17/03/1431

Change in Variation:
Normal(50.263,0.15)

Normal (50.26, 0.32)

Normal (50.26, 0.56)

LSL USL

Cost of Poor Quality:

„ All holes are within

tolerance.
„ They are not aligned.
„ Drill -- adds cost.
„ Collapse on the road
oad -- adds
cost.
„ Fix it -- adds cost.
„ Loss of next sale -- etc.

61
17/03/1431

Poor Quality System Cost$

$

„ 11 days to fill an order.

„ The whole batch has to be
ordered from the mother
company, before handling the
customer’s order.
„ Branch is rated on speed of
filling orders, mother
company is rated on
inventory cost.

Poor Quality System Cost$

$ (Cont.1):

„ It costs $100 to buy $3 battery.

„ Paperwork.
„ Who is involved?
„ Manager/supervisor.
„ Accounting.
g
„ Building a wall for the other guys.

62
17/03/1431

Poor Quality System Cost$

$ (Cont.2):

„ Tracing discrepancies between purchasing

& receiving.
„ Most tasks are simply to satisfy the internal
demands (management, accounting ..Etc..) ,
In many cases not needed.

Poor Quality System Cost$

$ (Cont.3):

„ High backorders: conflicting

measures.
™ Headquarters inventory cost.
:inventory cost.
™ Branch: speed of filling orders.

63
17/03/1431

Performance Measure(poor Q
System):
Bank: Airline Reservation:
800 tellers. 1200 agents.
„ Measure: Number of customers served / month.
„ Policy: Fire if < average for last 3 months.
„ Results:
„ Pr.(<
P (< average iin ththe last
l t 3 month)=
th)
(0.5)(0.5)(0.5)=0.125
„ No of tellers fired (each month) = 0.125 (800)

=100
„ No of agents fired (each month) = 0.125 (1200)

=150

Performance or Loss
Measure?

„ Hiring costs.
„ Training cost.
„ Loss of morale cost.
„ Loss of customer:
„ Rushed,,
„ Transferred to customer service,
„ Call again, system is down (it is not),
„ Hang up on customer.

64
17/03/1431

Inspect the Inspector (Poor Q

System):
„ Incoming inspection of
spare parts (example):
„ Warehouse worker.
„ Add another workers.
„ Add an accountant.
„ Add a programmer.
„ … What’s next?

Cost of Quality Estimates:

„ 35% of operating expenses in
service companies (Crosby).
„ 44% of work time (Paul
revere insurance) .
„

„
30% of sales (Juran).
20% of sales (Baldrige award
applicants) .
%?
„ 20-40% of hospital costs.
„ 25% non interest banking .

65
17/03/1431

What Is a Good Level For The

COQ?

2% of sales is a
good level for Costs
of Quality (Crosby)

Quality Costs
1 Quality Costs identify opportunities for quality
1.
improvement
2. Quality Costs are the basic data for Total
Quality Management (TQM)

Quality
Q y Costs Exceeds:
„ 20% of sales dollar in manufacturing

„ 30% of sales dollar in service

66
17/03/1431

Quality Costs Categories

Preventive Costs
1. Marketing/customer/user
2. Product/Service/Design development
3. Purchasing
4. Operations
5 Quality Administration
5.
6. Others (Rent, Telephone, Travel, etc.)

Quality Costs Categories

Appraisal Costs
1. Purchasing Appraisal Costs
2. Operation Appraisal Costs
3. External Appraisal Costs
4. Review of test and Inspection data
5 Miscellaneous Quality Evaluations
5.

67
17/03/1431

Quality Costs Categories

Internal Failure Costs

1. Product Design Failure Costs
2. Purchasing Failure Costs
3. Operations Failure Costs

Quality Costs Categories

External Failure Costs
1. Complaint Investigation of Customer
2. Returned Goods-Evaluate & Repair
3. Recall Costs. Recall of products due to
quality problems
4 Warranty Claims
4.
5. Liabilities
6. Penalties
7. Lost Sales

68
17/03/1431

Quality Costs Categories

Base line is needed to relate the quality
costs to other costs such as:
1. Labor (Quality Cost/Hour of Labor
2. Production (Quality Cost/$ of Production
Cost)
3. Sales (Q
(Qualityy Cost/$
/$ of Net Sales))
4. Unit (Quality Cost/Unit)
Optimum costs with respect to
conformance percentage can be determined.

Iceberg Phenomenon :
VISIBLE Training
QUALITY Return
COSTS Rework
Quality dept.

HIDDEN Non-value-added Opportunity

COSTS activities costs
Loss of goodwill Rework
Negative word-of-mouth
word of mouth
Low morale Lost orders
Inability to get jump on
competition instead of Redundancy
running in place Turnover: customer
and employee
Management time
Waste

69
17/03/1431

Costs of Quality Categories:

100%
Prevention (5%)
90%
80%
Appraisal (25%)

70%
60%
50%
40%
Failure (70%)
30%
Cost Of
20%
Quality
10%
0%
Operating Cost

Prevention + Appraisal+ Failure = Total Quality Costs

Exercise (Classify Activities As):

Prevention, Appraisal, Failure and Red Flag.
Prevention Flag
1 Complaints handling
1-Complaints 10-Reports
10 Reports
2-Customers surveys 11-Strategic planning
3-Internal audits 12-Trouble-shooting
4-Quality education accounting
5-Review of work 13-Logging rework time
6-Meeting 14-Marketing research
7-Customer report 15-EEO Lawsuits
8-Worker’s compensation 16-Inter functional
9-Management training design team

70
17/03/1431

Behavior of Costs Of Quality:

(Ricoh Example)

$17 608
$17,608

$215
$17
$5

Manufacturing

Design Procurement Customer

P+I F Total

OLD MODEL NEW MODEL

0% Conformance 100% 0% Conformance 100%

71
17/03/1431

Th Cost
The C t Of Quality
Q lit (COQ):
(COQ)

Estimates

Contents:

„ Four Techniques.
„ Estimating Service, Non-Service & Hidden
Costs.
„ Sources.
„ Flowcharting
Flowcharting.
„ Things to Avoid / Problems with COQ.

72
17/03/1431

Four Techniques for Estimating

Overall COQ:

„ Whole Account.
„ Whole Person.
„ Activity Analysis.
„ Labor// Resource
Claiming.

Whole Account:

„ Training.
Training
„ Audit.

„ Legal.

„ Rework / Returns.

Each is classified (P , A , F).

73
17/03/1431

Whole Person:
„ Inspectors.
„ Expeditors (Purchasing / Custom
Clearing / Dept collectors … etc.).
„ Lawyers.

„ Co
Complaint
p a t Handling
a d g personnel.
pe so e
„ Warranty personnel.

„ Software Debugging Engineers.

Each is to be classified (P , A , F).

Activity Analysis:
Self reported time spent on quality.
Advantages Disadvantages
„ Increases Awareness. „ Requires Education.
„ Requires ≤ 1 hour. „ Disruptive.
„ Simple spreadsheet. „ Lots of Analyses.
„ Interdepartmental „ Workers
comparisons. comparisons.
„ Easily replicated. „ Requires dictionary.

74
17/03/1431

Activity Analysis(Cont.1):
„ 1) Requirements:
„ Dictionary of activities.
„ Data collection instrument.
„ Analyst.
„ Software.

Activity Analysis(Cont.2):
„ 2)) Results
„ Categorized as:
Prevention, Appraisal, and Failure.
„ Associated dollars.
„ Percentages.
„ Specify to where costs reside.
„ Priority ranking.

75
17/03/1431

Labor/ Resource Claiming:

„ It differs from activity analysis:
„ Time is actually tracked rather than
estimated.
„ The cost requires analyzing time &
resource allocations.
„ Overtime,
Overtime
„ Supplies,
„ Equipment,
„ …..etc.

Estimating Non-Service Costs:

„ Reworked Units
Units.
„ Material.
„ M/C time lost.
„ Utility (Electricity / Oil / Gas ….).

76
17/03/1431

Estimating Hidden Costs:

„ Find Whole person/whole account
then, x 2, 3 or 4.
„ Although well documented, it may not be
convincing to some.
„ More than 25% of non-
manufacturing (service) is reworked
(Crosby).
„ Exclude whole person to avoid duplication.

Estimating Hidden Costs (Cont.)

Lost customers (TARP):

„ U h
Unhappy customer
t tells
t ll 10-12
10 12 people.
l
„ 13% of unhappy customer tell 20 people.
„ 2% of potential customers will not do business
(based on negative word-of-mouth).
„ 96% of customers do not complain.
„ Surveys why they are dissatisfied & leaving.
„ In banks:
„ 40% leave due to poor service,
„ 60% leave for other reasons.

77
17/03/1431

Problems with Quality Costs:

„ Lack of agreement on definitions &
components.
„ Even departments within companies tend to
develop their own definitions.
„ Lack of comparability over time or between
industries.
„ Diffi lt tto b
Difficult build
ild systems
t to
t capture
t d
data.
t
„ Financial measures for short-term, while
quality is definitely a long-term strategy.

Overcoming Obstacles:

Leadership

Resistance Inaction

L k off Commitment
Lack C it t

Sabotage Delays

Limited Resources

78
17/03/1431

From QC to TQM
Day
y5 TQM … Organization
Leadership Employees
Culture
Business Results
Day 4 QA … Business Processes
Procedures
Policies Responsibilities

Days 2 & 3 SPC … Product

Off-line
On-line

What is
ISO 9000:2000 QMS ?
„ ISO = International
I t ti l Organization
O i ti forf
Standards
„ 9000 = code to denote QMS Family
„ 2000 = yyear of last revision
„ Is a family of standards for implementing a
Quality Management System (QMS)

79
17/03/1431

The ISO 9000 Family

ISO 9004 ISO 9001 ISO 19011

Performance QMS Model Auditing
Improvements

ISO ??? ISO 10013

Sector Guidelines Quality Manual
ISO 9000
Fundamentals & Vocabulary

Why ISO 9001:2000 QMS ?

80
17/03/1431

What is QMS?
„ QMS is a:
„ System: a set of logically related processes
that turn input into output.
„ For Managing: administering activities and
resources to achieve an objective.
„ Quality: customer/user satisfaction.

Input Process Output

ISO 9001 Model: Process-Based QMS

Continual Improvement of the QMS
TS

MER SATISFACTION
N

clause 4
MER REQUIREMENT

info. flow management

responsibility QMS
clause 5
clause 6 clause 8
resource meas., analysis info. flow
management improvement
CUSTOM
CUSTOM

clause 7
Input product Output
realization product
value adding
value
activities
adding activities

81
17/03/1431

ISO 9001 Model:

Key Processes
Product realization and
ER REQUIREMENTS
S

ER SATISFACTION
measurement processes.
clause 5
„ Management
clause 6 processes.
clause 8

„ Provision for resources processes.

CUSTOME
CUSTOME

clause 7
Input product Output
realization product
value adding
value
activities
adding activities

Process Versus Function

Approach
Process

Functions/
Departments

Order Product

82
17/03/1431

ISO 9001 Model:

ER REQUIREMENTS
S The PDCA Cycle

ER SATISFACTION
management
responsibility
clause 5
Plan Act
clause 6
resource meas., analysis
management improvement

CUSTOME
CUSTOME

Do Check
clause 7
Input product Output
realization product
value adding
value
activities
adding activities

ISO 9001 Model: Process-Based QMS

Continual Improvement of the QMS
TS

MER SATISFACTION
N

clause 4
MER REQUIREMENT

info. flow management

responsibility QMS
clause 5
clause 6 clause 8
resource meas., analysis info. flow
management improvement
CUSTOM
CUSTOM

clause 7
Input product Output
realization product
value adding
value
activities
adding activities

83
17/03/1431

ISO 9001 Certificate

Realization Process
Plan
Identify
Document
Gap
QM

Implement

Evidences

Register Certificate

Identify
Gap

Identifying the Gap

Organization ISO 9001
Standard

shall
shall
shall
shall
shall shall

Plan

84
17/03/1431

Identify

Documenting: Gap

The QM Structure Document

Level
Policies I
shall
shall
shall
shall
Level
shall shall Procedures II
(Department)
ISO 9001
Standard Level
Work Instructions III
(Job/Task)

Quality Records Level

External Documents IV

Identify

Example: Gap

Developing Policy Document

„ OCCD reviews its quality management system for

suitability, adequacy, and effectiveness twice a year,
during the first week of the fall and spring semesters.
„ The review topics include follow-up on actions and
recommendations from the previous meeting, audits
results, customer feedback on services, status of
preventive and corrective actions,
actions and
recommendations for improvements.
„ Records of the management review meetings are
maintained on MF-01.

85
17/03/1431

Identify

From Standards to Gap

Quality Manual Document

Organization ISO 9001 Quality

Standard Manual

shall
shall
shall
shall
shall shall

Identify

Implementing Gap

Documentation Document

Implement

Organization ISO 9001 Implementing

Standard Quality Manual

shall
shall
shall
shall
shall shall

86
17/03/1431

Identify
Document

Getting ISO 9001

Gap

Certificate Implement

Organization ISO 9001 Implementing Register

Standard QMS

shall
shall
shall
shall
shall shall

Certificate

Effective QMS:
Beyond Certificate
Certificate

87
17/03/1431

Why Audit?

Who Audits?
„ Organization.
Organization
„ Customer.
„ Registrar.

88
17/03/1431

The Audit Cycle

Plan audit Execute audit

Report findings
Not OK
Review findings Apply
corrective
OK
actions
Close-out

Corrective versus
Preventive Action
„ Preventive Action:
„ Improve.
„ Prevent occurrence.
„ Corrective Action:
„ Comply.
„ Stop recurrence.

89
17/03/1431

USA Accreditation/
Registration System

Certify Auditors Audit

(Register)

Employ

Satissfy
ANSI Standards Accredit Certify
RAB Registrars Organization
ASQ (Register)

UK Accreditation/
Registration System
DIT
Administer

NRB Standards
NACCB BSI
Train &
Register Accredit

Employ
l
Auditors Registrars

Certify
(Register)

Audit Organization Satisfy

90
17/03/1431

Registration Process

Field Corrective
Registrar selection audit action

Formal application
Registration

Corrective
C i Office
Offi
action audit Surveillance
audit

Not OK
Optional pre-audit
CA/Suspension/
Termination

Day 5:
From QC to TQM
Day
y5 TQM … Organization
Leadership Employees
Culture
Business Results
Day 4 QA … Business Processes
Procedures
Policies Responsibilities

Days 2 & 3 SPC … Product

Off-line
On-line

91
17/03/1431

Fundamentals of TQM
„ Customer-focused.

„ Process-based QMS.

„ Fact-based CI approach.

„ People.

„ Results-driven.

Internal & External

Customers/Users
Marketing

Design

Purchase Produce Sell

92
17/03/1431

Triple Roles
I’m a CUSTOMER
I’m a PROCESSOR
I’m a SUPPLIER
Inputs Outputs
Your Your
Supplier You Customer

Requirements Requirements
& feedback & feedback

Case Study:
AT&T I am
Responsible
for Quality
My Input My Output
My As a As a As a
My
Supplier Good Good Good Customer
Customer Process Owner Supplier
I will: I will: I will: Requirement
Requirement 1. Agree on and 1. Learn & apply the 1. Understand my & Feedback
& Feedback document my tools of quality & customer req. &
requirements teach others agree on and
with my supplier 2. Continuously improve document my y
2. Return defective inputs my process – reduce deliverables
to my supplier promptly defects, cycle time & 2. Reduce defects &
& tactfully know benchmarks variations in my output
3. Feed back input quality data 3. Document & display 3. Measure my output quality
to my supplier my process, defect from my customer’s
levels & CI projects perspective

LIVING OUR CI VALUES

93
17/03/1431

Fundamentals of TQM
„ Customer-focused.

„ Process-based QMS.

„ Fact-based CI approach.

„ People.

„ Results-driven.

The CI Approach
NOW TQM
Focus is on results
Plan

Act Do
Check

PAST TQM
Focus was on programs
Plan Do

94
17/03/1431

Fundamentals of TQM
„ Customer-focused.

„ Process-based QMS.

„ Fact-based CI approach.

„ People.

„ Results-driven.

People
„ Leadership.

„ Motivation & Recognition.

„ Teamworking.

„ Training.

95
17/03/1431

Leadership
„ Leaders get people to want to do things.

„ Leaders do the right things; managers do

things right.

„ Theory “X” and Theory “Y” Managers.

Exercise:
Leaders vs. Managers
Activity: Creating an agenda
______ Develop a vision and establish strategies
______ Establish detailed steps
Activity: Developing a network to achieve agenda
______ Providing policies and procedures for team
______ Influencing creation of teams
Activity: Implementing the agenda
______ Monitoring results against plan
______ Energizing people to overcome barriers

96
17/03/1431

Motivation & Recognition

„ L d are greatt motivators.
Leaders ti t
„ Motivation theories:
„ Mazlow
„ Hunt
„ Herzberg

Herzberg Motivation Model

H i
Hygiene Factors
F t M ti t
Motivators
Money Sense of achievement
Status Acknowledgement
Work conditions Authorities
Relation with
ith boss Responsibilities
Relation with peers Growth potential
Organization politics Job pleasure

97
17/03/1431

Training

Identify Needs Train Evaluate

„Current knowledge, „On-the-job vs. off- „Did you like it?
skills, & attitude the-job „Did you learn it?

„Desired knowledge, „Methods: coaching, „Did you apply it?

skills, & attitude lecturing, e-learning „Was it effective?
„Gap

„Plan

Fundamentals of TQM
„ Customer-focused.

„ Process-based QMS.

„ Fact-based CI approach.

„ People.

„ Results-driven.

98
17/03/1431

Global TQM Models

„ The USA Quality Award (BNQA), 1987

„ European Quality Award (EQA), 1988

„ Australian Quality Award (AQA), 1993

„ Canadian Quality Award (CQA), 1984

„ The Japanese Quality Award

(The Deming Prize), 1951

Why TQM Models?

„ They are based on TQM fundamentals

„ Their criteria define organization excellence

„ They are used to assess current status

„ They are used to motivate for results

99
17/03/1431

The USA Model

Organizational
g Profile:
Environment, Relationships, and Challenges

Strategic HR
Planning Focus
Business
Leadership
Results
Customer & Process
Market Focus Management

Information and Analysis

The European Model

Enablers Results
People
People
mance Results

Results
Leaddership

Processes

Customer
Policy
Results
Key Perform

& Strategy

Society
Partnerships
Results
& Resources

Innovation & Learning

100
17/03/1431

The Australian Model

Customer Focus

Strategy, Policy
& Planning

Quality of Organizational
Information Performance
Process,
& Analysis
Product and
Service
People

Leadership

The Canadian Criteria

„ Leadership
„ Planning
„ Supplier focus
„ Organizational performance
„ C t
Customer focus
f
„ People focus
„ Process management

101
17/03/1431

The Japanese Criteria

„ Future Plans „ Information
„ Human Resources „ Improvement
„ Policies „ Effects
„ Organization
„ St d di ti
Standardization
„ Quality Assurance
„ Maintenance

Analyzing the Models:

The USA
„ Business Results 450
„ Leadership 120
„ Information and Analysis 90
„ Strategic Planning 85
„ C t
Customer andd Market
M k t Focus
F 85
„ Human Resource Focus 85
„ Process Management 85

102
17/03/1431

Analyzing the Models:

The European
„ Customer Results 20%
„ Key Performance Results 15%
„ Processes 14%
„ Leadership 10%
„ People 9%
„ People Results 9%
„ Partnership & Resources 9%
„ Policy & Strategy 8%
„ Society Results 6%

The European Model:

Enablers & Results
Enablers 50% Results 50%

People
mance Results

People
Results
Leaddership

Processes

Policy Customer
Key Perform

& Strategy Results

Partnerships Society
& Resources Results

Innovation & Learning

103
17/03/1431

Analyzing the Models:

The Australian
„ People 20%
„ Quality of process, product, and service 20%
„ Customer Focus 18%
„ Leadership 14%
„ O
Organizational
i ti l Performance
P f 12%
„ Strategy & Policy 8%
„ Information & Analysis 8%

Analyzing the Models:

The Canadian
„ Organizational performance 24%
„ Customer focus 17%
„ People focus 17%
„ Process management 17%
„ L d hi 10%
Leadership
„ Planning 10%
„ Supplier focus 5%

104
17/03/1431

Analyzing the Models:

The Japanese

All 10 criteria are equally weighted at 10%

Closer Look:
The BNQA:2001 Model
Organizational
g Profile:
Environment, Relationships, and Challenges

Strategic HR
Planning Focus
Business
Leadership
Results
Customer & Process
Market Focus Management

Information and Analysis

105
17/03/1431

BNQA Overview
„ Is named after Malcolm Baldridge!
g
„ Is a model for Performance Excellence
„ The NIST and ASQ manage the program
„ The model is revised annually
„ 3 Performance Excellence booklets:
„ 2001 Business Criteria
„ 2001 Education Criteria
„ 2001 Health Care Criteria

BNQA Overview
„ 5 Award categories each year:
„ Manufacturing businesses
„ Service businesses
„ Small businesses (500 or fewer employees)
„ Education organizations
g
„ Health care organizations
„ 3 awards may be given in each award category
each year

106
17/03/1431

Company Benefits
„ Win awardd
Wi
„ Basis for corporate self-assessment
„ On average, publicly traded award recipients
have outperformed S&P’s 500 by four to one
„ Cheap consulting!
„ Receive analysis report from 5-6 highly

trained experts for only $5,000

Famous Winners
„ FedEx „ Xerox

„ GE „ American Express

„ Motorola „ Ben & Jerry’s

„ 3M
„ Eastman Chemical Co.

107
17/03/1431

Application Review Process

Prepare
Applicant
pp application package

Board of
Examiners Independent review
(NIST/ASQ) application package

Consensus review Judges’

application package Review
Feedback
Report
Score Best 3
Ok Yes
Conduct
Site visit

The Model Components

„ 7 Criteria

„ 18 Items (General Requirements)

„ 29 Areas to address (Specific

Requirements)

108
17/03/1431

REMEMBER

In general, Excellence and Quality

Models such as BNQA, EQA, and
ISO tell you what is excellence. They
do not tell you how to achieve
excellence.
ll

What is Six Sigma?

General
„ Is a TQM strategy
„ Developed by Motorola in 1988
„ Focuses on breakthrough process improvement
to achieve customer satisfaction, cycle time
and defect reductions
reductions, and ultimately bottom-
bottom
line improvement

109
17/03/1431

Six Sigma Success

„ Massive bottom
bottom-line
line benefits
„ Many BNQA winners (e.g. GE, Motorola)
„ Exponential increase in the number of fortune
global 500 companies adopting six sigma
„ 1985-1990: 2 (Motorola & IBM)
„ 1991 1995 6 (GE,
1991-1995: (GE Allied
Alli d Signal,
Si l Kodak,
K d k etc.))
„ 1996-1999: 31 (Dow Chemicals, DuPont, Sony,
American Express, Johnson&Johnson, Raytheon,
United Technologies, US Postal Service, etc.)

The Technical Meaning of

Six Sigma
„ Sigma is
Si i a statistical
i i l measure off process
variability
„ The number of defects per million (dpm) at
Motorola’s six sigma is 3.4
„ True six sigma processes yield 2dpb
„ Most processes in good companies operate at
three sigma, which yield 2,700 dpm (99.73%)

110
17/03/1431

Six Sigma Process

LSL USL

0.001 1350 1350 0.001

dpm dpm dpm dpm

-6σ -3σ μ +3σ +6σ

Processes Capabilities
Spec. Distance One-sided Two-sided
from μ Spec. Spec.
Less than 3σ Incapable Incapable
Process Process
3.0σ 1,350 2,700

4.5σ 1.7 3.4

6.0σ 0.001 0.002

111
17/03/1431

The Six Sigma CI Approach

& PDCA
„ Define … Determine project (Plan)
„ Measure …. Determine current (Do)
„ Analyze …. Find solutions (Do)
„ Implement …. Execute (Do)
„ Control …. (Check & Act)

Six Sigma:
Common Quality Tools
„ Pareto analysis
„ Root-cause analysis
„ Statistical process control
„ Design of experiments

112
17/03/1431

The Training Pyramid

Course Six Sigma Title
Basic White Belts
Medium Green Belts
Comprehensive Black Belts
Executive
i Champions
h i
Executive Believer

The Training Pyramid

Rank Main Role
Employees Task improvement

Supervisors Lead small IPs

Managers Lead major,

major cross-
cross
functional IPs

In-house Experts Train, provide expertise

113
17/03/1431

Why Six Sigma Succeeds?

„ ‘Real’ top management commitment
„ Goals with identical improvement rates
„ Common quality measurement techniques
„ Common language throughout the organization
„ Systematic training
„ Goal-directed incentives for people

114