Transcript Class 1 - Kellogg School of Management

Operations Management:
Process Quality & Improvement Module
 Quality
& the Voice of the Customer
» What is Quality?
» Quality Programs in practice
» Voice of the Customer

Process Capability and Improvement
» Process Capability
» Checking for Improvement (Quality Wireless)

Control Charts & Voice of the Process
» Statistical Process Control (SPC)
» Quality Wireless (B)

Why 6-Sigma?
» Flyrock Tires
S. Chopra/Operations/Quality
1
8 Dimensions of Quality

Performance

Features

Serviceability
Q of design
 Aesthetics

Perceived Quality

Reliability
 Conformance

Q of process conformance
to design = process capability
Durability
S. Chopra/Operations/Quality
2
Elements of TQM
 Management
by fact
 Cross-functional (process) approach
 Culture and leadership
– Customer focus
– Employee focus
– High performance focus
» Continuous improvement
» Benchmarking

External alliances - the value chain
Source: Eitan Zemel
S. Chopra/Operations/Quality
3
Malcolm Baldridge National Quality Award


1 Leadership 110
2 Strategic Planning 80
– Strategy Development Process




3 Customer and Market Focus 80
4 Information and Analysis 80
5 Human Resource Development and Management 100
6 Process Management 100
– Product and Service Processes
– Support Processes
– Supplier and Partnering Processes

7 Business Results 450

S. Chopra/Operations/Quality
TOTAL POINTS 1000
4
Malcolm Baldridge Award Winners
















Ames Rubber Corporation (1993)
Armstrong World Industries Building Products
Operations (1995)
AT&T Consumer Communications Services
(1994)
AT&T Network Systems Group (1992)
AT&T Universal Card Services (1992)
Cadillac Motor Car Company (1990)
Chugach School District (2001)
Clarke American Checks (2001)
Corning Telecommunications Products Division
(1995)
Dana Corporation (2000)
Eastman Chemical Company (1993)
Federal Express Corporation (1990)
Globe Metallurgical Inc. (1988)
Granite Rock Company (1992)
GTE Directories Corporation (1994)
IBM Rochester (1990)
S. Chopra/Operations/Quality
Last Updated: May 28, 2002

















Karlee Company, Inc. (2000)
Los Alamos National Bank (2000)
Marlow Industries (1991)
Milliken & Company (1989)
Motorola Inc. (1988)
Operations Management International (2000)
Pal’s Sudden Service (2001)
Pearl River School District (2001)
The Ritz-Carlton Hotel Company (1992)
Solectron Corporation (1991)
Texas Instruments Incorporated - Defense
Systems & Electronics Group (1992)
University of Wisconsin-Stout (2001)
Wainwright Industries, Inc. (1994)
Wallace Co., Inc. (1990)
Westinghouse Electric Corporation Commerical Nuclear Fuel Division (1988)
Xerox Corporation - Business Products &
Systems (1989)
Zytec Corporation (1991)
5
ISO 9000

Series of standards agreed upon by the International Organization for
Standardization (ISO)

Adopted in 1987

More than 100 countries

A prerequisite for global competition?

ISO 9000: “document what you do and then do as you documented.”
Design
Procurement
Production
Final test
Installation
Servicing
ISO 9003
ISO 9002
Source: Adapted from Chase & Aquilano
S. Chopra/Operations/Quality
ISO 9001
6
S. Chopra/Operations/Quality
7
S. Chopra/Operations/Quality
8
Costs of Quality
 Cost
Quality Lever
Benefits of Building Q in Early
of Conformance
– Cost of Appraisal
Product
Design
Process
Design
Production
– Cost of Prevention
Improve
Product
100:1

Cost of Non-Conformance
10:1
1:1
– Cost of Internal Failure
– Cost of External Failure
S. Chopra/Operations/Quality
Low Visibility
Reward
Time
High Visibility
Reward
9
Components of Quality
 Voice
of the customer
– Customer Needs
– Quality of Design

Voice of the process
– Quality of Conformance
– Process Capability

Process Control and Improvement
S. Chopra/Operations/Quality
10
Voice of the Customer: Linking
Customer Needs to Business Processes
Business Process
Product (30%)
Sales (30%)
Overall Quality
Installation (10%)
Repair (15%)
Billing (15%)
Source: Kordupleski et al., CMR ‘93.
S. Chopra/Operations/Quality
Customer Need
Internal Metric
Reliability (40 %)
% Repair Call
Easy to Use (20%)
% Calls for Help
Features/Functions (40%)
Function Performance Test
Knowledge (30%)
Supervisor Observations
Response (25%)
% Proposals Mad on Time
Follow-Up (10%)
% Follow-Up Made
Delivery Interval (30%)
Average Order Interval
Does Not Break (25%)
% Repair Reports
Installed When Promised
% Installed on Due Date
No Repeat Trouble (30%)
% Repeat Reports
Fixed Fast (25%)
Average Speed of Repair
Kept Informed (10%)
% Customers Informed
Accuracy, No Surprise (45%)
% Billing Inquiries
Response on First Call (35%)
% Respolved First Call
Easy to Understand (10%)
% Billing Inquiries
11
Voice of the Customer:
Quality Function Deployment







What do customers want?
Are all preferences equally important?
Will delivering perceived needs deliver a competitive
advantage?
How can we change the product?
How do engineering characteristics influence customer
perceived quality?
How does one engineering attribute affect another?
What are the appropriate targets for the engineering
characteristics?
S. Chopra/Operations/Quality
12
Correlation:
Strong positive
X
House of Quality
Positive
X
X
Easy to close
7
Stays open on a hill
5
Easy to open
3
Doesn’t leak in rain
3
No road noise
2
-
-
Water resistance
+
*
Accoust. Trans.
Window
Energy needed
to open door
+
Check force on
level ground
-
Door seal
resistance
Customer
Requirements
Energy needed
to close door
Engineering
Characteristics
Negative
Strong negative
X
X
X
Competitive evaluation
X = Ours
A = Comp. A
B = Comp. B
(5 is best)
+
1
2
3
4
AB
X
X
5
X AB
XAB
A XB
X A
Importance weighting
10
6
6
9
2
3
B
Relationships:
Technical evaluation
(5 is best)
5
4
3
2
1
B
A
X
BA
X
B
B
A
X
X
A
BXA
Maintain
current level
Maintain
current level
Reduce energy
to 7.5 ft/lb.
Reduce force
to 9 lb.
Target values
Maintain
current level
Reduce energy
level to 7.5 ft/lb
Strong = 9
Medium = 3
Small = 1
BA
X
Source: Hauser and Clausing 1988
Linked Houses From Customer To
Manufacturing
House of
Quality
S. Chopra/Operations/Quality
Parts
Deployment
III
Process
Planning
Production
Characteristics
Key Process
Characteristics
II
Key Process
Characteristics
Parts
Characteristics
I
Parts
Characteristics
Engineering
Characteristics
Customer Attributes
Engineering
Characteristics
IV
Production
Planning
14
Benefits of QFD
Startup and Preproduction costs
at Toyota Auto Body
Japanese automaker with QFD made fewer
changes than US company without QFD
Design
Changes
US
Japan
Before QFD
After QFD
90% of total Japanese
changes complete
(39% of preQFD costs)
Job # 1
t
20 - 24
months
14 - 17
months
1 - 3 Job # 1
months
1-3
months
time
Source: Hauser and Clausing 1988
S. Chopra/Operations/Quality
15
More New Product Development Tools

Value analysis / Value engineering

Design for manufacturability

Robust design
S. Chopra/Operations/Quality
16
Value Analysis/Value Engineering
 Achieve
equivalent or better performance at a lower
cost while maintaining all functional requirements
defined by the customer
– Does the item have any design features that are not
necessary?
– Can two or more parts be combined into one?
– How can we cut down the weight?
– Are there nonstandard parts that can be eliminated?
S. Chopra/Operations/Quality
17
Robust Quality: Taguchi’s View of Cost
of Variability
Non-conformance to
design cost
$$$
0
Lower
Tolerance
Design
Spec
Upper
Tolerance
Traditional View
S. Chopra/Operations/Quality
Actual
value
Lower
Tolerance
Design
Spec
Upper
Tolerance
Taguchi’s View
18
Quality & the Voice of the Customer:
Key Learning Objectives
 Elements
of TQM / Baldridge / ISO 9000

Costs of Quality

Components of Quality

Voice of the Customer
– Linking business processes to customer needs
– Product Design Methodologies:
» Convert customer needs to product and process specifications: QFD
» Value Engineering
S. Chopra/Operations/Quality
19
Operations Management:
Process Quality & Improvement Module
 Quality
& the Voice of the Customer
» What is Quality?
» Quality Programs in practice
» Voice of the Customer

Process Capability and Improvement
» Process Capability
» Checking for Improvement (Quality Wireless)

Control Charts & Voice of the Process
» Statistical Process Control (SPC)
» Quality Wireless (B)

Why 6-Sigma?
» Flyrock Tires
S. Chopra/Operations/Quality
20
Process Capability
 Percent
defective
– Proportion of output that does not meet customer
specifications

Sigma-capability
– Number of standard deviations from the mean of the
process output to the closest specification limit.
S. Chopra/Operations/Quality
21
Quality Wireless (A): Capability
Distribution of Average Daily Hold Time for 2003-04
20
Within Specs
Out of Specs
18
16
Number of Days
14
12
10
8
6
4
2
99
10
3
10
7
11
1
11
5
11
9
12
3
12
7
13
1
13
5
13
9
14
3
14
7
15
1
15
5
15
9
16
3
95
91
87
83
79
75
71
67
63
59
55
51
47
43
39
0
Average Daily Hold Time
S. Chopra/Operations/Quality
22
Quality Wireless (A): Capability
 Proportion
of days within specification in 2003-04 =
491/731 = 0.672
 The call center had a mean hold time of 99.67 with a
standard deviation of 24.24. With a specification of 110
seconds or less,
σ-capability of call center = (110 – 99.67)/24.24
= 0.426
The call center is a 0.426-sigma process. Expected fraction
of days within specifications from a 0.426-sigma process
= NORMSDIST(0.426) = 0.665
S. Chopra/Operations/Quality
23
What is Process Improvement?
S. Chopra/Operations/Quality
24
Continuous Improvement:
PDCA Cycle (Deming Wheel)
Institutionalize the change or
abandon or do it again.
Plan a change aimed
at improvement.
4. Act
3. Check
Study the results;
did it work?
S. Chopra/Operations/Quality
1. Plan
2. Do
Execute the change.
25
Quality Wireless (A): Checking for Improvement
 Performance
in April 2005: Mean = 79.50, Standard
deviation = 16.86
 What is the probability of observing such a sample if
performance has not improved relative to 2003-04?
– Mean hold in 2003-04 = 99.67
– Standard deviation = 24.24
– Given that April 2005 had 30 days, we need to consider
distribution of samples of size 30. The standard deviation of
sample means = 24.24/√30 = 4.43
– Probability of observing a sample of size 30 with mean 79.50 or
less = NORMDIST(79.50, 99.67, 4.43, 1) = 2.64E-06
S. Chopra/Operations/Quality
26
Operations Management:
Process Quality & Improvement Module
 Quality
& the Voice of the Customer
» What is Quality?
» Quality Programs in practice
» Voice of the Customer

Process Capability and Improvement
» Process Capability
» Checking for Improvement (Quality Wireless)

Control Charts & Voice of the Process
» Statistical Process Control (SPC)
» Quality Wireless (B)

Why 6-Sigma?
» Flyrock Tires
S. Chopra/Operations/Quality
27
Has Process Performance Changed? Quality
Wireless (B)
 Average
hold time from September 1-10 =86.6
seconds
– Ray yells at supervisors
Performance improves from September 11-20 to an
average hold of 74.4 seconds
 What do you think of Ray’s management style?

S. Chopra/Operations/Quality
28
Performance of Inventory Manager
WIP
Award Given
WIP
month
JF MAMJ JASON
Manager repents and kicks...
month
JF MAMJ JASON DJF
WIP
.. and concludes that kick ... mgt works !?
JF MAMJ JASON DJF MAMJ
S. Chopra/Operations/Quality
month
29
Statistical Process Control:
Source of Variability

Inherent (common cause)

External (assignable cause)
Objective: Identify inherent variability and eliminate external
variability. A process is in control if it has only inherent
variability.

To improve the system, attack common causes (methods,
people, material, machines). This is the role of management.
S. Chopra/Operations/Quality
30
Various Patterns in Control Charts
Pattern
S. Chopra/Operations/Quality
Description
Possible Causes
Normal
Random Variation
Lack of Stability
Assignable (or special) causes (e.g. tool,
material, operator, overcontrol
Cumulative trend
Tool Wear
Cyclical
Different work shifts, voltage
fluctuations, seasonal effects
31
SPC – Quality Wireless (B)
 After
the improvements, daily hold time has an
average of 79.50 and a standard deviation of 16.86.
 Since we are considering samples of size 10 (10
days), we need to consider the distribution of sample
means. Sample means have an average of 79.50 and a
standard deviation of 16.86/√10 = 5.33.
 Probability of observing 86.6 or higher even if
process is in control = 1-NORMDIST(86.6, 79.50,
5.33, 1) = 0.0915
S. Chopra/Operations/Quality
32
SPC – Quality Wireless (B)
 Probability
of observing 74.4 or lower even if process
is in control = NORMDIST(74.4, 79.50, 5.33, 1) =
0.1693
 What we need is a hypothesis test each time we
observe a sample – Does the sample belong to the incontrol population or not?
S. Chopra/Operations/Quality
33
SPC – Setting Control Limits
 Upper
Control Limit = UCL = Mean + 3σXbar
 Lower Control Limit = LCL = Mean - 3σXbar
 In the case of Quality Wireless
– UCL = 79.50 + 3×5.33 = 95.49
– LCL = 79.50 - 3×5.33 = 63.51

The process was in control when samples with means
of 86.6 and 74.4 were observed.
S. Chopra/Operations/Quality
34
Control Charts & Voice of the Process:
Key Learning Objectives
 The

role of variability in evaluating performance
A process
– in control has only inherent (from common cause) variation
– out of control has variation from an assignable cause

SPC framework for process control and improvement
S. Chopra/Operations/Quality
35
Operations Management:
Process Quality & Improvement Module
 Quality
& the Voice of the Customer
» What is Quality?
» Quality Programs in practice
» Voice of the Customer

Process Capability and Improvement
» Process Capability
» Checking for Improvement (Quality Wireless)

Control Charts & Voice of the Process
» Statistical Process Control (SPC)
» Quality Wireless (B)

Why 6-Sigma?
» Flyrock Tires
S. Chopra/Operations/Quality
36
Why 6-Sigma?






2 sigma:
69.146% of products and/or services meet customer requirements
with 308,538 defects per million opportunities.
4 sigma:
99.379% of products and/or services meet customer requirements ...
but there are still 6,210 defects per million opportunities.
6 sigma:
99.99966% – As close to flaw-free as a business can get, with just
3.4 failures per million opportunities (e.g. products, services or
transactions).
S. Chopra/Operations/Quality
37
Why 6-Sigma?
 Impact
of # of parts/stages in a process
Probability that process/product meets specs
3 -sigma
4 - sigma
5 - sigma
6 - sigma
# of steps/parts
1
10
50
100
144
369
740
1044
1590
19581
42559
100000
1000000
93.3%
50.1%
3.2%
0.1%
0.00%
99.4%
94.0%
73.2%
53.6%
40.8%
10.0%
1.0%
0.1%
0.00%
100.0%
99.8%
98.8%
97.7%
96.7%
91.8%
84.2%
78.4%
69.1%
1.0%
0.00%
100.0%
100.0%
100.0%
100.0%
100.0%
99.9%
99.7%
99.6%
99.5%
93.6%
86.5%
71.2%
3.3%
Probability that process/product
meets specs
100.0%
10.0%
3 -sigma
1.0%
4 - sigma
5 - sigma
0.1%
6 - sigma
0.0%
0.0%
1
10
100
1000
10000
100000 1000000
# steps/components
S. Chopra/Operations/Quality
38
Why 6-Sigma? Robustness to Mean Shifts
LSL
USL
LSL
USL
s=5
s = 10
99.9 %
99.9 %
100
130
160
LSL
100
USL
130
160
LSL
USL
s=5
s = 10
100
S. Chopra/Operations/Quality
143
160
100
143
160
39
Why 6-Sigma? 6-Sigma Quality at Flyrock
At the extruder, the rubber for the AX-527 tires had
thickness specifications of 400  10. Susan and her
staff had analyzed many samples of output from the
extruder and determined that if the extruder settings
were accurate, the output produced by the extruder
had a thickness that was normally distributed with a
mean of 400 and a standard deviation of 4.
If the setting is accurate, what proportion of the rubber
extruded will be within specifications?
S. Chopra/Operations/Quality
40
Process Capability: Sigma Capability
Sigma capability is the number of standard deviations
from the mean to the closest specification limit.
 Sigma capability of extrusion process =

Susan has asked operators to take a sample of 10 sheets
of rubber each hour from the extruder and measure the
thickness of each sheet. Based on the average
thickness of this sample, operators will decide whether
the extrusion process is in control or not. Given that
Susan plans 3-sigma control limits, what upper and
lower control limits should she specify to the
operators?
Impact of Mean Shift
 If
a bearing is worn out, the extruder produces a mean
thickness of 403 when the setting is 400. Under this
condition, what proportion of defective sheet will the
extruder produce? Assuming the control limits in (2),
what is the probability that a sample taken from the
extruder with the worn bearings will be out of control?
On average, how many hours are likely to go by before
the worn bearing is detected.
S. Chopra/Operations/Quality
42
Why 6-Sigma? Rapid Detection
 What
if extrusion is to become a 6-Sigma process?
– Target mean =
– Target standard deviation =

Process improvement has resulted in the extrusion
process having a mean of 400 and a standard
deviation of 1.667. What should the new control
limits be? What is the proportion of defectives
produced?
S. Chopra/Operations/Quality
43
Improving Process Capability
 Return
to the case of the worn bearing in (3) where
extrusion produces a mean thickness of 403 when the
setting is 400. Under this condition, what proportion
of defective sheets will the extruder produce (for the
6-sigma process)? Assuming the control limits in (5),
what is the probability that a sample taken from the
extruder with the worn bearings will be out of
control? On average, how many hours are likely to go
by before the worn bearing is detected.
S. Chopra/Operations/Quality
44
Key Learning Objectives: SPC
 Specification
limits: Voice of the customer
 Process capability is a measure of the quality delivered
(external): links VoP with VoC
 Improving capability may require variability reduction
and/or mean shift
 Control limits used to verify if process is in control
(internal), i.e., is maintaining capability: Voice of the
process
 Higher process capability reduces defectives and
speeds up detection of assignable cause
S. Chopra/Operations/Quality
45