Qcar L1-L16 PDF
Qcar L1-L16 PDF
Qcar L1-L16 PDF
2
BITS Pilani, Pilani Campus
3
Evolution of Quality Control
3
BITS Pilani, Pilani Campus
4
Evolution of Quality
Control – contd..
• Middle ages and up to 1900 – contd..
– Volume of production was limited
– Controlling the quality of the product was thus
embedded in the philosophy of the worker because
pride in workmanship was widespread
– Sculptures, Goldsmith etc.
• Early 1900 to about 1920: Foreman Quality
control
– Industrial revolution resulted in concept of mass
production
– Resulted in principle of specialization of labour and
hence individual not responsible for producing whole
product
4
BITS Pilani, Pilani Campus
5
Evolution of Quality
Control – contd..
• Early 1900 to about 1920 – contd..
– Lacked in sense of accomplishment in their work, but
become skilled at a particular accomplishment
– But people are grouped together, where a supervisor
who directed that operation had the task of ensuring
that quality was achieved
• 1920 to 1940 : Inspection quality control phase
– Products and process were complicated and also the
volume increased
– Resulting in increase in number of workers under a
foreman and hence prevents him from keeping close
watch
5
BITS Pilani, Pilani Campus
6
Evolution of Quality
Control – contd..
• 1920 to 1940 : Contd..
– Hence inspectors were designated to check quality of
the product after certain operations comparing it with
standards. If any problem, checked product is reworked
or discarded
– Foundation of statistical aspects of quality control were
developed
– W.A Shewhart developed control charts in 1924
– H.F. Dodge and H.G. Romig was working on acceptance
sampling plan and its application was started in late
1920s
6
BITS Pilani, Pilani Campus
7
Evolution of Quality
Control – contd..
• 1940 to 1960 : Statistical quality control
– Production requirements increased during World war II
and hence 100% inspection was not feasible
– ASQC was formed in 1946 and developed sampling
inspection plans for military purposes MIL-STD-105A
– E. Deming visited Japan and lectured on importance of
statistical quality control in 1950
– J.M. Juran visited Japan in 1954 and impressed upon the
strategic role that management plays in achieving quality
program
– In 1959, Inspection and Quality control handbook was
updated with multi-level continuos sampling plan as well
as topics in life testing and reliability
7
BITS Pilani, Pilani Campus
8
Evolution of Quality
Control – contd..
• 1960 to 1970 : Total quality control
– Gradual involvement of other departments and
management personnel in quality control process
– Attitude of quality is responsibility of inspection
department started changing
– Concept of ‘zero defect’ which emphasized productivity
through worker involvement emerged
– Use of Quality circles was started in Japan, which is based
on participative style of management
• 1970 to present : Total quality control
Organization wide
– Involves participation of everyone in the company from
the operator to supervisor to manager to CEO
8
BITS Pilani, Pilani Campus
9
Evolution of Quality
Control – contd..
• 1970 to present : contd..
– A quality system was evolved which is
• Agreed on companywide operating work structure
• Documented integrating technical and managerial procedures
• Guiding the coordinated actions of men, machine and
information of the company
• To assure customer quality satisfaction and economical cost of
quality
– Expanded use of Ishikawa diagram
– G.Taguchi introduced the concept of quality
improvement through design of experiments
– With more advertisements on comparing competitors
brought quality into limelight
9
BITS Pilani, Pilani Campus
10
Evolution of Quality
Control – contd..
• 1970 to present : contd..
– Top management saw the need for the marriage of quality
philosophy to production in all phases from determination
of customer needs, product design and customer service
– Training program in SQC methods for all workers
– With the wide use of computers, lot of quality control
software came and hence the emphasis on vendor quality
control, product quality audit were placed
– In this phase, customer will reign supreme as the
determiner of acceptable level of quality
– Industries need to adjust to this or lose market share
10
BITS Pilani, Pilani Campus
11
Quality - Definition
11
BITS Pilani, Pilani Campus
12
How to define quality?
12
BITS Pilani, Pilani Campus
13
How to define quality? – contd..
• Diameter
• Length
• Pressure
• Wall Thickness
13
BITS Pilani, Pilani Campus
14
How to define quality? – contd.
• Taste, Smell
14
BITS Pilani, Pilani Campus
15
How to define quality? – contd.
Place B Place A
• Time
• Reliability
15
BITS Pilani, Pilani Campus
16
How to define quality? – contd.
• Honesty
• Courtesy
16
BITS Pilani, Pilani Campus
17
Quality characteristics
18
BITS Pilani, Pilani Campus
19
Some Terminologies
• Defect • Defective
– Associated with quality – Severity of one or more
characteristic that does not defects in a product may
meet certain standards cause it to be unacceptable
– According to ANSI (The
American National Standard
Institute), a defect is a
departure of quality
characteristic from its
intended level or state that
occurs with a severity
sufficient to cause an
associated product or
service not to satisfy
intended normal or
reasonable foreseeable
usage requirements
19
BITS Pilani, Pilani Campus
20
Some Terminologies – contd..
• Standard / specification • Standard
– Refers to a precise statement that – A prescribed set of conditions
formalizes the requirements of or requirements of general or
the customer, it may relate to a
product, process or a service broad application, established
• Specification by authority or agreement, to
– A set of conditions and be satisfied by a material,
requirements of specific and product, process, procedure,
limited application that provide a
detailed description of the test method etc and/or the
procedure, process, material, physical, functional,
product or service for use performance or conformance
primarily in procurement and
manufacturing characteristic thereof
– Standards may be referenced in – A physical embodiment of a
specification
unit of measurement (a
caesium block clock)
20
BITS Pilani, Pilani Campus
21
Aspect of Quality
21
BITS Pilani, Pilani Campus
22
Aspects of Quality
• Quality of design
– Implies that the product or service must minimally possess to
satisfy the requirements of the customer
– Design must be simplest and least expensive too
– Influenced by such factors as the type of product, cost, profit policy,
demand, availability of parts and materials, product safety etc.
– Eg: Customer requirement for yield strength of a cable is
100kg/cm2
– The parameters that influence the yield strength would be selected
to minimally satisfy this requirement
– In practice, the product is over designed with a safety factor k =
1.25, so the cable will be designed for 125 kg/cm2
– Increase in designed quality level will lead to increase in cost at an
exponential rate
22
BITS Pilani, Pilani Campus
23
Aspects of Quality – contd.
23
BITS Pilani, Pilani Campus
24
Aspects of Quality – contd..
24
BITS Pilani, Pilani Campus
25
Aspects of Quality – contd.
25
BITS Pilani, Pilani Campus
26
Aspects of Quality – contd..
• Quality of conformance
– Implies that the manufactured product or service
rendered must meet the standards selected in the
design phase
– With respect to manufacturing, it is concerned with the
degree to which quality is controlled from the
procurement of raw material to the shipment of
finished goods
– It consists of 3 broad areas:
• Defect Prevention: deals with means to prevent the occurrence
of defects and is usually achieved using SPC techniques
• Defect Finding: Done through inspection, test and statistical
analysis of data from the process
• Defect analysis and rectification
26
BITS Pilani, Pilani Campus
27
Aspects of Quality – contd..
• Quality of performance
– Is concerned with how well the product functions or
service performs, when put to use
– It measures the degree to which the product or service
satisfies the customer
– It is a function of both quality of design and quality of
conformance
– If a product does not function well enough to meet the
expectation of a customer or standards, then
adjustment need to be done in the design or
conformance phase
• Relation between QOD, QOC, QOP
27
BITS Pilani, Pilani Campus
28
Aspects of Quality – contd..
Quality of
conformance
Quality of Quality of
design performance
28
BITS Pilani, Pilani Campus
29
Aspects of Quality – contd..
29
BITS Pilani, Pilani Campus
30
Quality Control
30
BITS Pilani, Pilani Campus
31
Quality control
31
BITS Pilani, Pilani Campus
32
Quality control – contd..
32
BITS Pilani, Pilani Campus
33
Quality control – contd..
33
BITS Pilani, Pilani Campus
34
Quality control – contd..
34
BITS Pilani, Pilani Campus
35
Quality control – contd..
36
BITS Pilani, Pilani Campus
37
Quality Improvement
Is a never ending process to reduce both the variability of process and the
production of nonconforming items
Process control deals with identification and elimination of special causes that
force a system to go out of control, while quality improvement relates to the
detection and elimination of common causes
38
BITS Pilani, Pilani Campus
39
Quality Improvement – contd..
39
BITS Pilani, Pilani Campus
40
Quality Improvement – contd..
40
BITS Pilani, Pilani Campus
41
Examples of quality problems
41
BITS Pilani, Pilani Campus
42
Examples of quality problems
42
BITS Pilani, Pilani Campus
43
Quality Circle and Quality Improvement
Team
• Quality circle • Quality Improvement team
– Informal group – Formal group
– Participation is voluntary – Participation is mandatory
– People from same area or – People from different
dept. department
– Consists of operators, – Work towards improvement
supervisors and managers of product or process from
– Work towards improvement quality perspective
of product, process or their – Team gets dismantled after
personal well being completion of project
– Team never gets dismantled
– Productivity improvement
tool
43
BITS Pilani, Pilani Campus
44
Total Quality System – contd..
44
BITS Pilani, Pilani Campus
45
References
• Text book
• www.iisd.org
• YouTube video
https://www.youtube.com/watch?v=guJ4I3O
8DeU
https://www.youtube.com/watch?v=Jxn2sdZ
TLH0
45
BITS Pilani, Pilani Campus
46
Thank You
46
BITS Pilani, Pilani Campus
47
2
BITS Pilani, Pilani Campus
49
Benefits of Quality Control
• Cannot be realized immediately, but on long-term
perspective
• Benefits
– System is continually evaluated and modified to meet the
changing needs of customers and hence a mechanism
exists to rapidly modify product or process
– QC improves productivity as it reduces the scrap and
rework
– QC reduces cost in the long run and helps productivity and
cost reduction go hand in hand
– Improved delivery dates, due to reduction in lead time for
producing parts and SubAssembly are reduced
– Helps to stay competitive
3
BITS Pilani, Pilani Campus
50
Responsibility for Quality
4
BITS Pilani, Pilani Campus
51
Total Quality System
• Quality system is defined as the collective plans, activities
and events that are provided to ensure that a product,
process or service will satisfy customer needs
• System approach integrates the various functions and
responsibilities of different units and provides a
mechanism to ensure that org. goals are met through
proper coordination of other departments
• Elements of total quality system (ANSI/ASQC STD Z1.15)
– Policy, planning organization and administration
• A quality policy is developed inline with org. goals
• Quality manuals are created, which gives the detailed procedure and
costs
• ANSI-American National Standards Institute
5
BITS Pilani, Pilani Campus
52
Total Quality System – contd..
7
BITS Pilani, Pilani Campus
54
Total Quality System – contd..
• Elements of total quality system
(ANSI/ASQC STD Z1.15) – contd..
– Corrective action
• Problems are to be detected, categorized and
systematically documented, similar to that of the
remedial actions
• Specific procedures for corrective action for entities like
materials, vendors, process, equipment should be
developed
– Employee selection, training and motivation
• Guidelines should be established to select people for
particular jobs and job manuals to train people
• Recognition of superior effort and use of motivational
program help reassure employees of the support of mgt.
8
BITS Pilani, Pilani Campus
55
Total Quality System – contd..
9
BITS Pilani, Pilani Campus
56
Quality Cost
10
BITS Pilani, Pilani Campus
57
Quality Costs
11
BITS Pilani, Pilani Campus
58
Quality Costs
• Appraisal costs
– Are those associated with measuring, evaluating or
auditing products, components, purchased materials to
determine the degree of conformance to the specified
standards
– It includes inspection and test of incoming material,
product inspection and testing, cost of calibrating and
maintaining measuring instruments etc.
– These costs occur during or after production, but before
the product is released to the customer
– It decreases with time
13
BITS Pilani, Pilani Campus
60
Quality Costs – contd..
15
BITS Pilani, Pilani Campus
62
Quality Costs – Examples
16
BITS Pilani, Pilani Campus
63
Quality Costs – Examples
• Prevention cost
17
BITS Pilani, Pilani Campus
64
Quality Costs – Examples
• Appraisal cost
18
BITS Pilani, Pilani Campus
65
Measuring Quality Costs
19
BITS Pilani, Pilani Campus
66
Measuring Quality Costs – contd..
• 4 indices – contd..
– Cost base index
• Quality costs / dollar of manufacturing costs
• Includes direct labour, material and overhead costs
• More stable than labour base index because its not affected by
price fluctuations or changes in level of automation
• Important for middle management
– Sales base index
• Quality costs / dollar of sales
• Used by top management
• Not a good measure for short-term, but for long-term strategic
decisions, because sales lag behind production and are
subjected to seasonal variations
• Changes in selling price also affect this index
20
BITS Pilani, Pilani Campus
67
Measuring Quality Costs – contd..
• 4 indices – contd..
– Unit base index
• Quality costs / unit of production
• Used only when output from different production lines are
similar
• For dissimilar ones, the product lines have to be weighted and
standardized product measure computed
– For all these indices, a change in denominator causes
the value of the index to change, even if quality cost do
not change
– Cost of direct labour decreases due to productivity
improvement, labour base index decreases, and it
should not be treated as increased quality costs
21
BITS Pilani, Pilani Campus
68
Effect of Quality on..
• Productivity
– Quality decreases productivity is a misconception
– Quality relies on “make it right first time”, which
increases productivity
– Quality reduces waste (less scrap and rework) and
hence valuable resources can be utilized for production
of defect free goods.
• Effect on cost
– Prevention and appraisal cost
• Increases with initial improvements in productivity and due to
adequate process control procedures are installed
• With time, a reduction in appraisal costs is observed
• These costs are called costs of conformance to quality
requirements
22
BITS Pilani, Pilani Campus
69
Effect of Quality on..
23
BITS Pilani, Pilani Campus
70
Effect of Quality on.. – contd..
24
BITS Pilani, Pilani Campus
71
Management of Quality
• Achieved through functions of planning, organizing,
staffing, directing and controlling
• Eg. Product development
Management function Product phase Action to be taken
25
BITS Pilani, Pilani Campus
72
Management of Quality – contd..
Management Product phase Action to be taken
function
27
BITS Pilani, Pilani Campus
74
28
BITS Pilani, Pilani Campus
75
Case study- BBC
29
BITS Pilani, Pilani Campus
76
Environment impact group
• Waste
• Utilities (energy & water)
• Transport
• Supply chain and procurement
• Property
30
BITS Pilani, Pilani Campus
77
Challenges
31
BITS Pilani, Pilani Campus
78
Action
Waste
• Complete a comprehensive audit of waste throughout the BBC,
concentrating on paper, tapes and toner cartridges;
• Recycle at least 25% of office waste by 2003, rising eventually to 55%.
Utilities
• Each building manager to select an energy efficiency initiative, to be
run as a pilot programme;
• Reduce carbon dioxide emissions by 1% in the next year
Transport and travel
• Reduce car travel at the London White City site by 20% over four years,
by encouraging public transport use, cycling, teleworking and walking;
• Target the business units which require the most intensive transport
use.
32
BITS Pilani, Pilani Campus
79
33
BITS Pilani, Pilani Campus
80
Trends in Quality
• Quality certification
• Quality as project
• Quality in role of strategy
• Quality and sustainability(people and QMS)
34
BITS Pilani, Pilani Campus
81
35
BITS Pilani, Pilani Campus
82
36
BITS Pilani, Pilani Campus
83
References
• Text book
• www.iisd.org
• GEMI practice
• Curkovic, Sime, and Robert Sroufe. "Total
quality environmental management and
total cost assessment: An exploratory
study." International Journal of production
economics 105.2 (2007): 560-579.
• YouTube video
37
BITS Pilani, Pilani Campus
84
Homework
38
BITS Pilani, Pilani Campus
85
Chapter - 2
39
BITS Pilani, Pilani Campus
86
Some Quality Gurus
40
BITS Pilani, Pilani Campus
87
Deming’s contribution
• He emphasized on the role of management
• His approach is not a quick fix, rather a plan of
action to achieve long term goals
• Heart of his philosophy is that management and
workers should speak a common language – the
language of statistics
• His 14 points will form a framework for action and
it needs to be installed for the program to be
successful
• He identified the key components for continuous
improvement / “System of profound knowledge”
– Knowledge of the system and the theory of
optimization
– Knowledge of the theory of variation
– Exposure to the theory of knowledge
– Knowledge of psychology
41
BITS Pilani, Pilani Campus
Deming’s contribution – 88
contd..
• “System of profound knowledge” – contd..
– Knowledge of the system and the theory of
optimization
• Management needs to understand that optimization of
total system is the objective and not optimizing the sub
system to have sub-optimal total system
– Knowledge of theory of variation
• Management should understand special causes and
common causes in the variation of process
• Special causes can be controlled by operator or engineer
• Common causes should be reduced by management
only, which makes the system stable and in control
42
BITS Pilani, Pilani Campus
Deming’s contribution – 89
contd..
• “System of profound knowledge” – contd..
– Exposure to the theory of knowledge
• Experience and intuition are not of value to
management unless they can be interpreted and
explained in the context of a theory
• A data-analysis oriented approach to problem solving is
needed to suggest remedial action based on results
– Knowledge of psychology
• It is required to understand the behaviour and
interactions of people, and also the interactions of
people with their work environment
• People are motivated by intrinsic and extrinsic factors
– Job satisfaction and motivation: Intrinsic
– Reward and Recognition: Extrinsic
• Management should create the right mix of these factors
43
BITS Pilani, Pilani Campus
90
Thank You
44
BITS Pilani, Pilani Campus
91
contd..
• “System of profound knowledge” – contd..
– Exposure to the theory of knowledge
• Experience and intuition are not of value to
management unless they can be interpreted and
explained in the context of a theory
• A data-analysis oriented approach to problem solving is
needed to suggest remedial action based on results
– Knowledge of psychology
• It is required to understand the behaviour and
interactions of people, and also the interactions of
people with their work environment
• People are motivated by intrinsic and extrinsic factors
– Job satisfaction and motivation: Intrinsic
– Reward and Recognition: Extrinsic
• Management should create the right mix of these factors
2
BITS Pilani, Pilani Campus
93
Deming’s 14 points for
management
• Focus is on Management
• He expects a fundamental change in the style of
management and corporate culture
• Corporate culture should be in such a way that the
workers feel comfortable enough to recommend
changes
• Point 1: “Create and publish to all employees a
statement of the aims and purposes of the company”
– Management must demonstrate constantly their commitment
to this statement
– Emphasis is on long-term strategic plans and also on mission
statements that need to be understood by employers,
consumers, vendors and investors too.
– Can be achieved by
• Product Improvement cycle
• Constancy and consistency of purpose 3
BITS Pilani, Pilani Campus
Deming’s 14 points for 94
management
4
BITS Pilani, Pilani Campus
95
Deming’s 14 points for
management
5
BITS Pilani, Pilani Campus
96
Deming’s 14 points – contd..
• Point 1 – contd..
– Product Improvement cycle (Figure)
• It integrates the phases of QOD, QOC, QOP
• All customer needs must be quantified
• Company must focus on the customer and not on the
competitor
– Constancy and consistency of purpose (Figure)
• Implies setting a course (say all dept. follow the quality
improvement) and keeping to it
• Allocates resources for long-term planning, employee
training and education, which ensures that it can
maintain a competitive position for years to come
• Should not look into short term profits
6
BITS Pilani, Pilani Campus
97
Deming’s 14 points – contd..
• Point 1 – contd..
– Constancy and consistency of purpose – contd..
• Company should not deviate from long-term objectives
• i.e. it should accept that variability exists in any
operation and hence it should try to reduce
• Doing your best is not good enough, you have to know
what to do, then do your best
• Point 2
– Learn the new philosophy, top management and
everybody
• Quality consciousness must be everything to everyone
• Acceptable level of defect should be abandoned and one
should work towards reducing defects continuously and
also address the need of customer
7
BITS Pilani, Pilani Campus
98
Deming’s 14 points – contd..
• Point 3
– Understand the purpose of inspection, for
improvement of processes and reduction of cost
• Inspection does not increase quality, it just separates out
acceptable and non acceptable items
• It does not address the root cause of problem
• Hence emphasis should be placed on defect prevention
rather than defect detection
– Drawbacks of mass inspection
• Even 100% inspection will not eliminate all the defectives if
more than one person is responsible for inspection
• Reason being that, it is only human to assume that others
will find what you have missed
• Also the inspector fatigue should be taken into account
8
BITS Pilani, Pilani Campus
99
Deming’s 14 points – contd..
• Point 4
– End the practice of awarding business on the basis of price
tag alone
• Companies should review the bidders approaches to quality
control, and if possible should know what is the fraction of non
conforming items and the stability of process etc.
• Generally contract should be based on effective price per unit
• Eg. 16Rs. With 8% defective and 17Rs with 2% defective
– Principle of vendor selection
• While selecting a vendor, the total cost should be taken into
account
• Vendors and buyers should work as a team to choose methods
and materials to improve customer satisfaction
• It is desirable to have reduced no. of suppliers
• Having multiple suppliers results in mistrust between vendor
and buyer and hence no long-term relationships
• Price not quality is the driving factor
• Also results in variability in the incoming quality and hence
increased cost due to changes in setup, no volume discounts etc.
9
BITS Pilani, Pilani Campus
100
Deming’s 14 points – contd..
• Point 5
– Improve constantly and
forever the system of
production and service
• Focus is on defect
prevention and
process improvement
which are carried out
by the use of statistical
methods
• Process improvement
should be based on
Deming’s cycle which
consists of 4 stages:
Plan Do Check and Act
(PDCA)
10
BITS Pilani, Pilani Campus
101
Deming’s 14 points – contd..
• Point 5
– Plan
• Opportunities for
improvement are
recognised and
operationally defined
• Since customer
satisfaction is
important, degree of
difference between
customer needs
satisfaction and
process performance
are analysed
• Goal is to reduce the
difference and the
possible relationship
between the variables
in the process and
their effect are
hypothesized
11
BITS Pilani, Pilani Campus
102
Deming’s 14 points – contd..
• Point 5 – contd..
– Do
• Cause of action
developed in
planning is put
into action
• If necessary trial
runs or prototype
are conducted
• Feedback is
obtained from
process and
customer
12
BITS Pilani, Pilani Campus
103
Deming’s 14 points – contd..
• Point 5 – contd..
– Check
• Analysing the result
• Statistical methods
will be used
– Act
• Decision is made
regarding
implementation if the
results of the check
stage are positive and
if negative alternative
plans are developed
13
BITS Pilani, Pilani Campus
104
Deming’s 14 points – contd..
• Point 5 – contd..
– Variability reduction
and loss function
• Deming’s philosophy calls
for abandoning the idea
that everything is fine if
specifications are met
• Based on this Taguchi
formalized loss function
in 1960
• i.e. It explains that
economic loss accrue
with any deviation from
the target value
• However the loss
increase in a non linear
relationship with larger
deviation from target
14
BITS Pilani, Pilani Campus
105
Deming’s 14 points – contd..
• Point 6
15
BITS Pilani, Pilani Campus
106
Deming’s 14 points – contd..
• Point 6
– Institute training
• Employee training is integral for individual performance in
the extended process setting
• When employees are hired, they should be carefully
instructed in the company’s goals and this helps them to
understand their responsibilities for meeting customer
needs
• A employee should know what is to be done and its
importance in the entire process, which will make them
take pride in their work
• It make them feel secure, have better morale and hence
improve their productivity
• Training should be ongoing
• According to Deming, it is necessary for all employee of the
organization to be trained in SQC tools
16
BITS Pilani, Pilani Campus
107
Deming’s 14 points – contd..
• Point 7
17
BITS Pilani, Pilani Campus
108
Deming’s 14 points – contd..
• Point 7
– Teach and institute leadership
• Supervisors are the vital links between
management and the workers
• To be effective leader, the supervisors
must think in terms of helping workers do
a better job and manage them
• Supervisors should be trained in SQC and
should create a atmosphere, which
improves employee morale, promotes
team work and help achieve the overall
annual goal of quality improvement
18
BITS Pilani, Pilani Campus
109
Deming’s 14 points – contd..
• Point 8
– Drive out fear, create
trust. Create climate for
innovation
• Fear results in physical
or psychological
disorders, poor morale
and also results in lack of
job security and hence
reduces productivity
• An employee should feel
free to know about the
process, about his
responsibility and
should be able to suggest
something, which helps
in building trust and a
climate
19 for innovation
BITS Pilani, Pilani Campus
110
Deming’s 14 points – contd..
• Point 9
– Optimize toward the aims
and purposes of the
company, the efforts of
teams, group, staff areas
• Focus is to remove
organizational barriers –
both internal and external
• Internal barriers will result
in impeding the flow of
information /
communication between
the different level of
employees and also
between departments
• External barrier is the lack
of flow of information
between the customers,
investors and vendors,
which should be
eliminated
20
BITS Pilani, Pilani Campus
111
Deming’s 14 points – contd..
• Point 10
– Eliminate
exhortations for the
workforce
• Never set a target
arbitrarily as it has
demoralizing effect
• Set the goal, which
should be feasible and
the management
should specify the way
to achieve it
• Goals should not be
intuitive or arbitrary,
but should be based
on information from
every level
21
BITS Pilani, Pilani Campus
112
23
BITS Pilani, Pilani Campus
114
Deming’s 14 points – contd..
• Point 12
– Remove barriers
that rob people the
pride of
workmanship
• Quality is achieved in
all components of the
extended process,
when the employees
are satisfied and
motivated and when
they understand their
role in the context of
organization goals
24
BITS Pilani, Pilani Campus
115
Deming’s 14 points – contd..
• Point 12 – contd..
– Factors that cause loss of pride
• Lack of treatment of employee with dignity
• Lack of communication of company’s mission to all
levels
• Assigning blame to the employees for not meeting
company goals though the output has problems due to
system
– Performance classification system
• Inconsistencies in performance evaluation and there
should not be categorization of employee from a similar
sample
• Teamwork should be given importance, when
conducting performance appraisal, those promoting
teamwork should be placed in higher category
25
BITS Pilani, Pilani Campus
116
26
BITS Pilani, Pilani Campus
117
31
BITS Pilani, Pilani Campus
122
Crosby’s Philosophy – contd..
• 14 step plan – contd.
– Quality improvement team
• Representation from each department or division
form the team
• Responsible for bringing suggestions to actions
32
BITS Pilani, Pilani Campus
123
Crosby’s Philosophy – contd..
• 14 step plan for
quality improvement –
contd..
– Quality measurement
• Identifies the status of
quality
• Identifies area where
corrective action is
needed and quality
improvement efforts
should be directed
• Results should be
placed in highly
visible charts 33
BITS Pilani, Pilani Campus
124
34
BITS Pilani, Pilani Campus
125
35
BITS Pilani, Pilani Campus
126
Crosby’s Philosophy – contd..
• 14 step plan for quality improvement – contd..
– Corrective action
• Needs open communication and action discussion of
problems
• Discussions exposes other problem not identified before
and helps to determine procedures to eliminate them
• Attempts to resolve problem should be made as they arise
36
BITS Pilani, Pilani Campus
127
Crosby’s Philosophy – contd..
• 14 step plan for
quality improvement
– contd..
– Adhoc committee for
the zero defects
program
• Concept of zero defect
to be communicated
for all employee
• Committee is
responsible for getting
top management
commitment
37
BITS Pilani, Pilani Campus
128
Crosby’s Philosophy – contd..
• 14 step plan for
quality improvement
– contd..
– Supervisor training
• All levels of
management must be
aware about the steps
of quality
improvement
programs
• They should be
explaining the
program to
employees
38
BITS Pilani, Pilani Campus
129
Crosby’s Philosophy – contd..
• 14 step plan for
quality improvement
– contd..
– Zero defects day
• Philosophy of zero
defects should be
established
companywide and
should originate on
one day
• Management should
foster this type of
quality culture and
should try to motivate
the people
39
BITS Pilani, Pilani Campus
130
Crosby’s Philosophy – contd..
• 14 step plan for
quality
improvement –
contd..
– Goal setting
• Employees along
with their
supervisors should
set specific
measurable goals,
which creates a
attitude among
people to achieve it
40
BITS Pilani, Pilani Campus
131
Crosby’s Philosophy – contd..
• 14 step plan for quality
improvement – contd..
– Error cause removal
• Employees are asked to
identify reasons that
prevent them meeting
zero defect goal – They
should not make
suggestion, but list the
problem
• A separate functional
group should come up
with procedures for
removing it
– Recognition
• Award programs should
be based on recognition
rather than money, who
have met or exceeded 41
their goals BITS Pilani, Pilani Campus
132
Crosby’s Philosophy – contd..
42
BITS Pilani, Pilani Campus
133
Thank You
43
BITS Pilani, Pilani Campus
134
3
BITS Pilani, Pilani Campus
137
Juran’s Philosophy – contd..
• Quality planning
– Occurs at various levels of
organization, each having a
distinct goal
– Strategic quality
management, where broad
quality goals are set, and a
management chooses a
plan of action and allocates
resources to achieve the
goals
– Operational quality
management, where
departmental goals are set,
inline with strategic goals
– At worker level, clear
assignment of task to each
worker to contribute to
departmental goal
4
BITS Pilani, Pilani Campus
138
Juran’s Philosophy – contd..
• Quality control
– Goal is to run the process
effectively, such that plans
are enacted
– Any deficiency in planning
process, results in process
operating at high level of
chronic waste
– If unusual symptoms are
sporadically detected,
quality control will try to
identify the cause behind
the variation and remedial
action taken to bring it back
to control
– Objective is to eliminate the
cause of sporadic spike and
bring process output to
zone of quality 5
BITS Pilani, Pilani Campus
139
Juran’s Philosophy – contd..
• Quality improvement
– Deals with continuous
improvement of product
and process
– It requires an action on
upper and middle
management
– Deals with actions like
creating new designs,
changing methods or
procedures, new
equipment investment
etc.
– Results in reducing the
chronic waste and hence
the cost of quality
6
BITS Pilani, Pilani Campus
140
Quality trilogy – in detail
• Quality planning
– Identify the customer – both external and internal,
which is similar to that of Deming’s extended process
– Determine the customer needs. This decides the long
term survival of customers
– Develop product features that respond to customer
needs
– Establish quality goals that meet the needs of
customers and suppliers like and do so at a minimum
cost. Total cost from an organization point of view
should be minimum
– Develop a process that can produce the needed
product features
– Prove process capability: To establish whether the
process is adequate for making a product that
conforms to design specifications 7
BITS Pilani, Pilani Campus
141
Quality trilogy – in detail: Contd..
• Quality control
– Choose control subjects
• Identify the product characteristics that is to be controlled in
order to make the product conform to requirements
• Selection is done by identifying the characteristics that has more
impact on customer
– Choose units of measurement
– Establish measurement
• Procedures for taking measurements are defined
• Involves identifying what equipments to be use, how to use, who
will be responsible, how to make measurements an when to
calibrate, what training is needed etc.
– Establish standards of performance
• Should be based on customer requirements
– Measure actual performance
• Concerned with measurement of actual process output; to know
the operation level of process
8
BITS Pilani, Pilani Campus
142
Quality trilogy – in detail: Contd..
• Quality control – contd..
– Interpret the difference
• Compare actual and standard and if the process is stable,
the difference may not be significant
– Take action on the difference
• Take remedial action for the difference
• It is management’s responsibility to suggest a remedial
cause of action
• Quality Improvement
– Prove the need for improvement
• Identify problems and convert it into dollar figures to
attract management attention
• Convince the management and involve the management to
make a change in the process
– Identify specific projects for improvement
• Prioritize the problems, based on Pareto analysis as only 9
limited resources will be available
BITS Pilani, Pilani Campus
Quality trilogy – in detail:
143
Contd..
• Quality Improvement – contd..
– Organize to guide the projects
• Establish a clear organization structure
• Assign authority and responsibility at all levels of
management to facilitate this
• Establish precise responsibilities: guidance for overall
improvement program, guidance for each individual
project and diagnosis and analysis for each project
– Organize for diagnosis – for discovery of causes
• A “Diagnostic Arm”, a group of persons brought together
to determine the causes of the problem
• Organization need to enlist the right people to ensure
that the required tools and resources. This is done by
“Steering Arm”
– Find the causes
• Involves data gathering and analysis to determine the
causes
• Involves studying the symptoms, hypothesising the
causes and validating it 10
BITS Pilani, Pilani Campus
144
Quality trilogy – in detail: Contd..
• Quality Improvement – contd..
– Provide remedies
• Diagnostic step is based on cause and effect relationship
• Remedies may be for problems that are controllable by
management and those that are controlled by operators
• It may involve change in methods or equipment, which may
require substantial financial investment
– Prove that the remedies are effective under operating
condition
• Check whether the process can be implemented and will it have
benefits
• Breakthrough process requires overcoming resistance to change
• Proposed procedure may require new equipment and operators
may have to be trained
• Management commitment is vital
– Provide control mechanisms to hold the gains
• Once remedial action implemented, it need to be sustained, 11
which needs a control system like audit in some departments
BITS Pilani, Pilani Campus
145
3 philosophies compared
Criteria Deming Crosby Juran
Definition of - Deals with - “Conformance -“Fitness for
quality predictable to requirements use”
uniformity of - Emphasis on - explicitly
product customer needs related to
- Emphasis on use implicitly meeting the
of SQC - Emphasize on needs
- Quality of product zero defects, the
depends on quality requirement
of process should be met
Management - Highly - Focussed on - Scales
commitment emphasized creating quality management
- Points 1, 2 and 14 culture which commitment at
are aimed at needs all levels
management management
commitment 12
BITS Pilani, Pilani Campus
146
3 philosophies compared –
contd..
Criteria Deming Crosby Juran
Strategic - Top management - Need for - Need for
approach to involves pursuing all quality quality council
quality 13 points in a improvement and also
system structured cycle team emphasized on
problem-solving
steering and
diagnostic arms
Measurement - Depends on - Quality is free
of quality deriving a dollar and it is that un-
value for the total quality that costs
cost of quality
Never ending - Follow Plan-Do- - Repeats the Believes in
process of Check-Act cycle and cycle of quality breakthrough
improvement repeats planning, control sequence 13
and improvement BITS Pilani, Pilani Campus
147
3 philosophies compared –
contd..
Criteria Deming Crosby Juran
Education - Emphasizes on - Emphasizes on - His stress on
and training training of new quality culture, education is
employee and also where education implicit
re-training on is a part of it
continuous basis for
existing people
Chapter - 3
Quality Management –
Practices, Tools and Standards
16
BITS Pilani, Pilani Campus
150
17
BITS Pilani, Pilani Campus
151
Management Commitment
• Should have strategic plans with long-
term focus
• Should satisfy the customer and
investor by delivering quality products
• Can be accomplished by having a
‘Quality Program’ like TQE, TQM or Six
Sigma etc.
• Eg. Ford Motor Company follow Total
Quality Excellence (TQE) program
18
BITS Pilani, Pilani Campus
152
Total Quality Excellence
• (TQE)
A system that integrates
all aspects of the company
and its suppliers into
strategic decision making
on different functions of a
company
• It is based on Ford’s
vision of “being a low cost
producer of the highest
quality products and
services which provide
the best customer value”
19
BITS Pilani, Pilani Campus
153
TQE – Concepts, Functional
Areas
• Key Concepts
– Quality should be defined by customer
– Quality is achieved through prevention
of problems and not detection
– Vendor-Vendee concept, (Internal and
External Customer)
– All employees, suppliers, dealers and
part of the company (Deming’s
extended process)
– Quality improvement is never ending
20
BITS Pilani, Pilani Campus
154
TQE – Contd..
• Functional Areas includes
Engineering, Delivery and
Commercial Performance
• Embraces the system approach
• Relies on partnership between
manufacturer and supplier
– Ford developed ‘Q101 Preferred
Quality Award’ for suppliers
• Emphasize on Process control, Customer
feedback and Documentation including
management developed documentation
for design and change control 21
BITS Pilani, Pilani Campus
155
Q 101 Ford’s Preferred Supplier
Award
22
BITS Pilani, Pilani Campus
156
Basic concepts of TQM
23
BITS Pilani, Pilani Campus
157
Old vs. TQM Approach
24
BITS Pilani, Pilani Campus
158
25
BITS Pilani, Pilani Campus
159
Total Quality Management
(TQM)
• Management Commitment
– Must exists at all levels of the company
– Creates policy at top level, implements at
middle level and quality mgt tools and
techniques are used at operation level
• Customer
– Satisfying customer needs is the driving
force
– Direct feedback using a data driven
approach is the best way to identify
customer needs
– Key principle is that are customers are
both internal and external
26
BITS Pilani, Pilani Campus
160
TQM – Contd..
Customer surveys are used to find the
discrepancies between expectation and
satisfaction and taking measures to
eliminate is called ‘Gap Analysis’
• Process
– Vendors are part of the process
– Cross functional team, with authority to
solve problem and improve the process
– Use of technical tools and techniques and
management tools is necessary to improve
continuously 27
BITS Pilani, Pilani Campus
161
TQM – Contd..
• People
– Involving employees in the decision making
process (Empowerment)
– Ownership and motivation are necessary to
make employee get a sense of pride
• Communication
– Open channel of communication
– Emphasis is on Information sharing between
departments
– Change from role of management of
coordination and control to that of coaching
and caring 28
BITS Pilani, Pilani Campus
162
TQM – Contd..
• Culture
– Links the human element and the
company’s vision
– They are the beliefs, values, norms and
rules that prevail within an company
– Culture should embrace a ‘participative
style’ of management
• Vision
– Tells about what company wants to be
– It is general and outlines the scope and
purpose of organization
– Should be motivational and make people
work towards the goal 29
BITS Pilani, Pilani Campus
163
TQM – Contd..
• Mission
– Derived from vision
– More focused and defines the areas of
concentration
– No mention about time frame
• Quality Policy
– Framed by senior management, it is the
company’s road map
– Indicates what to be done, but different from
procedures and instructions
– Eg. ‘quality is the basic business principle at
Xerox’
– Outcome is the performance standard 30
BITS Pilani, Pilani Campus
164
Vision and Mission Statement
• Vision
– Be recognized as the best education provider
in Mechanical Engineering
• Mission
– The Mechanical Engineering Group is
committed
• to provide a high quality (world-class) education
in Mechanical Engineering
• to conduct strong research programs
• to foster a close partnership with industry and
– The mission is guided by a continuous
improvement philosophy
31
BITS Pilani, Pilani Campus
165
Vision and Mission – contd..
32
BITS Pilani, Pilani Campus
166
Vision and Mission – contd..
33
BITS Pilani, Pilani Campus
167
Six Sigma Quality
• Developed by Motorola
• Made popular by General Electric (GE)
• Some of the companies that has adopted 6 σ are
– Wipro
– CTS
– Satyam
– General Motors
– Ford India Limited
• Uses quantitative goals, based on the measure
of process variation (σ)
• The objective is to minimize the variation from
the existing level (if it is 3 σ) to half
• The concept of 6 σ is based on the concept of
normal distribution 34
BITS Pilani, Pilani Campus
168
Variation – A Simple Test
37
BITS Pilani, Pilani Campus
171
Six Sigma Quality – contd..
• For a process output given by normal
distribution
– Mean is the measure of location of process
– The no. of non-conforming product will be
2700ppm (1350ppm on each tail)
– Assume a product with 1000 processes or
parts, then the no. of defects per product will
be 2.7!!! (only 7 out 100 will be defect free)
• For defect free product, the process
spread given by (+/- 3 σ) has to be
significantly less than that between USL
and LSL 38
BITS Pilani, Pilani Campus
172
A Normal distribution
•
39
BITS Pilani, Pilani Campus
173
Six Sigma – Contd..
• Focus of Six sigma is to reduce the process
variation such that the spec. limits are 6 σ from
the mean
• If process is stable and remains centered, the
proportion of defective will be 0.001ppm on
each tail (Ideal)
• Even for a shifts by 1.5 σ from the mean, no. of
defects will be 3.4ppm on each tail
• For the example above there will be 0.0034
defect per product and the yield will be 99.66%
• Requires a fundamental change in management
philosophy and culture
40
BITS Pilani, Pilani Campus
174
Quality Function
Deployment (QFD)
• A planning tool that focus on designing
quality into a product
• A system that involves cross functional
team and takes care of complete cycle of
product development
• Helps to translate customer needs into
technical requirements of products
• Called as House of Quality, Matrix
product planning, Customer drive
41
BITS Pilani, Pilani Campus
175
Questions answered by QFD
• 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? 42
BITS Pilani, Pilani Campus
176
QFD “house of quality”
43
BITS Pilani, Pilani Campus
177
QFD - Process
Customer Rating
45
BITS Pilani, Pilani Campus
179
QFD – Process (Contd..)
46
BITS Pilani, Pilani Campus
180
Target goals
47
BITS Pilani, Pilani Campus
181
QFD – Process (Contd..)
• Co-relationship matrix
– Forms the roof of HOQ
– 4 levels of relationship – strong
positive, positive, negative, strong
negative
– Used for determining the trade-off
between ‘Hows’
48
BITS Pilani, Pilani Campus
182
x = Design Trade-offs
49
BITS Pilani, Pilani Campus
183
QFD – Process (Contd..)
50
BITS Pilani, Pilani Campus
184
Benchmarking
51
BITS Pilani, Pilani Campus
185
QFD – Process (Contd..)
• Technical Competitive Assessment
– Similar to that of customer assessment of
competitor
– Done by a technical staff of the company
– Used to set objective values for the
technical descriptors
52
BITS Pilani, Pilani Campus
186
Technical Competitive
Assessment
53
BITS Pilani, Pilani Campus
187
QFD – Process (Contd..)
• Relationship matrix
– A mechanism to analyse how each
technical descriptors will help in
achieving ‘What’
– Represented by 0 to 5, where 0 is no
relationship 1 – low, 3 – medium and 5 -
High
54
BITS Pilani, Pilani Campus
188
55
BITS Pilani, Pilani Campus
189
QFD – Process (Contd..)
56
BITS Pilani, Pilani Campus
190
U – Us
M – Maytag
W – Whirlpool
G – GE
F – Frigidaire
A - Amana
57
BITS Pilani, Pilani Campus
191
QFD – The real meaning
• It is used to take the ‘voice of customer’
and translate it into a set of product and
process parameters and can be deployed
through a 4-phase process
– Product planning
– Part / Subsystem deployment
– Process planning
– Production planning
58
BITS Pilani, Pilani Campus
192
QFD – The real meaning – contd.
technical
requirements
component
characteristics
process
operations
quality plan
59
BITS Pilani, Pilani Campus
193
QFD - Advantages
Thank You
62
BITS Pilani, Pilani Campus
196
2
BITS Pilani, Pilani Campus
198
Six Sigma – Contd..
• Focus of Six sigma is to reduce the process
variation such that the spec. limits are 6 σ from
the mean
• If process is stable and remains centered, the
proportion of defective will be 0.001ppm on
each tail (Ideal)
• Even for a shifts by 1.5 σ from the mean, no. of
defects will be 3.4ppm on each tail
• For the example above there will be 0.0034
defect per product and the yield will be 99.66%
• Requires a fundamental change in management
philosophy and culture
3
BITS Pilani, Pilani Campus
199
Quality Function
Deployment (QFD)
• A planning tool that focus on designing
quality into a product
• A system that involves cross functional
team and takes care of complete cycle of
product development
• Helps to translate customer needs into
technical requirements of products
• Called as House of Quality, Matrix
product planning, Customer drive
4
BITS Pilani, Pilani Campus
200
Questions answered by QFD
• 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? 5
BITS Pilani, Pilani Campus
201
QFD “house of quality”
6
BITS Pilani, Pilani Campus
202
QFD - Process
Customer Rating
8
BITS Pilani, Pilani Campus
204
QFD – Process (Contd..)
9
BITS Pilani, Pilani Campus
205
Target goals
10
BITS Pilani, Pilani Campus
206
QFD – Process (Contd..)
• Co-relationship matrix
– Forms the roof of HOQ
– 4 levels of relationship – strong
positive, positive, negative, strong
negative
– Used for determining the trade-off
between ‘Hows’
11
BITS Pilani, Pilani Campus
207
x = Design Trade-offs
12
BITS Pilani, Pilani Campus
208
QFD – Process (Contd..)
13
BITS Pilani, Pilani Campus
209
Benchmarking
14
BITS Pilani, Pilani Campus
210
QFD – Process (Contd..)
• Technical Competitive Assessment
– Similar to that of customer assessment of
competitor
– Done by a technical staff of the company
– Used to set objective values for the
technical descriptors
15
BITS Pilani, Pilani Campus
211
Technical Competitive
Assessment
16
BITS Pilani, Pilani Campus
212
QFD – Process (Contd..)
• Relationship matrix
– A mechanism to analyse how each
technical descriptors will help in
achieving ‘What’
– Represented by 0 to 5, where 0 is no
relationship 1 – low, 3 – medium and 5 -
High
17
BITS Pilani, Pilani Campus
213
18
BITS Pilani, Pilani Campus
214
QFD – Process (Contd..)
19
BITS Pilani, Pilani Campus
215
U – Us
M – Maytag
W – Whirlpool
G – GE
F – Frigidaire
A - Amana
20
BITS Pilani, Pilani Campus
216
QFD – The real meaning
• It is used to take the ‘voice of customer’
and translate it into a set of product and
process parameters and can be deployed
through a 4-phase process
– Product planning
– Part / Subsystem deployment
– Process planning
– Production planning
21
BITS Pilani, Pilani Campus
217
QFD – The real meaning – contd.
technical
requirements
component
characteristics
process
operations
quality plan
22
BITS Pilani, Pilani Campus
218
QFD - Advantages
25
BITS Pilani, Pilani Campus
221
Innovation & Cont. Improvement
26
BITS Pilani, Pilani Campus
222
Benchmarking (BM)
27
BITS Pilani, Pilani Campus
223
Objective of Benchmarking
28
BITS Pilani, Pilani Campus
224
Benchmarking Process
29
BITS Pilani, Pilani Campus
225
Benchmarking - Requirements
30
BITS Pilani, Pilani Campus
226
Why Benchmarking?
31
BITS Pilani, Pilani Campus
227
Spider Chart
32
BITS Pilani, Pilani Campus
228
Quality Audit
▪ Purpose
▪ Suitability Quality Audit: Evaluation of Quality
program against reference standard
▪ Conformity Quality Audit: Evaluation of operations,
activities within the quality system with respect to
procedures 33
BITS Pilani, Pilani Campus
229
Quality Audit – Contd..
▪ Types
▪ System Audit
▪ Most extensive
34
BITS Pilani, Pilani Campus
230
Quality Audit – Contd..
▪ Types – contd..
▪ Process Audit
▪ In depth evaluation of one or more processes
▪ Takes less time
▪ More focussed and less costly
▪ Used generally in process industries like
chemical industries
▪ Done when there is need for process
improvement or when there is unexpected
output
35
BITS Pilani, Pilani Campus
231
Quality Audit – Contd..
▪ Types – contd..
▪ Product Audit
▪ Assessment of final product or service on its ability
to meet customer needs
▪ Involves conducting periodic test on products or
obtain info. from customers
▪ Separate from decision on product acceptance or
rejection and hence not part of the inspection system
36
BITS Pilani, Pilani Campus
232
Quality Audit – Contd..
▪ Audit Quality
▪ Influenced by independence and objectivity of the
auditor
▪ Depends on being a external or internal auditor
▪ For suitability audit, external is most preferable
▪ Methods
▪ Location oriented
▪ All functions are audited in that particular location
▪ Examines the action and interaction of the elements
in the quality program at that location and may be
used to identify discrepancies between location
▪ Function oriented
▪ Examine and evaluate activities related to a particular
element or function within a quality program at all
locations
▪ Successive visits to each location is necessary to
complete the audit
▪ Utility of quality audit is derived only when remedial
actions are taken
37
BITS Pilani, Pilani Campus
233
Tools for Continuous
Quality Improvement
• For solving the problems and involves data collection,
analysis, hypothesis and validation
• Called as 7 QC tools
– Check sheets
– Histogram
– Flow chart
– Pareto chart
– Cause and effect diagram
– Control chart
– Scatter diagram
38
BITS Pilani, Pilani Campus
234
What are tools and techniques?
• Tools
– Tools and techniques are practical methods, skills,
means or mechanisms that can be applied to particular
tasks
– Used to facilitate positive change and improvements
– Narrow in focus and is usually used on its own
– Examples of tools are 7 QC tools
• Technique
– It has a wider application than a tool.
– Needs more thought, skill and training to use
– It can be thought of as a collection of tools
– Eg., statistical process control (SPC) employs a
variety of tools such as charts, graphs and histograms
39
BITS Pilani, Pilani Campus
235
Why to use tools and
techniques?
• Plays a key role in a company-wide
approach to continuous improvement. They
allow:
– Processes to be monitored and evaluated;
– Everyone to become involved in the
improvement process;
– People to solve their own problems;
– A mindset of continuous improvement to be
developed
40
BITS Pilani, Pilani Campus
236
Requirement for tools and
techniques
• Requires attention to be paid to a number of
“critical success factors” to make their use and
application effective and efficient. Some of these
are:
– Full management support and commitment;
– Effective, timely and planned training;
– A genuine need to use the tool or technique;
– Defined aims and objective for use;
– A co-operative environment;
– Backup and support from improvement facilitators.
41
BITS Pilani, Pilani Campus
237
7 QC Tools – contd.
• Check sheets
– A systematic record
keeping or data
collection
– Observations are
recorded as they
happen and reveals
pattern or trends
42
BITS Pilani, Pilani Campus
238
7 QC Tools – Contd..
• Histograms
– Displays large amount
of data that are
difficult to interpret in
raw form
– Provides summary of
data and also reveals
whether the process is
centered, the degree of
variation etc.
– Used to identify
process capability
relative to customer
requirements
43
BITS Pilani, Pilani Campus
239
7 QC Tools – Contd..
• Flow charts
– Shows the sequence
of events
– Mostly used in
manufacturing and
service operations
to describe working
procedures
– Valuable process
information can be
obtained in addition
to identifying
problematic areas
44
BITS Pilani, Pilani Campus
240
7 QC Tools – contd.
• Pareto Diagram
– helps to prioritize
the problem by
arranging them in
decreasing order of
importance
45
BITS Pilani, Pilani Campus
241
7 QC Tools – Contd..
• Cause and Effect diagram
–Called as Ishikawa diagram for fish
bone diagram
–Explores possible causes of
problem due to men, machine,
material and method
–To identify root cause, each cause
may be further broken down
46
BITS Pilani, Pilani Campus
242
7 QC Tools – contd.
47
BITS Pilani, Pilani Campus
243
7 QC Tools – Contd..
• Control charts
– Distinguishes special (assignable) causes of
variation from common causes
– Used to monitor and control a process on a ongoing
basis
– Plots a selected quality characteristic, found from
subgroups of observations, as a function of sample
number
– Central line on the control chart is the average value
of the characteristic
– Two limits, UCL and LCL are used to detect whether
the process goes out of control
48
BITS Pilani, Pilani Campus
244
7 QC Tools – contd.
49
BITS Pilani, Pilani Campus
245
7 QC Tools – Contd..
• Scatter Plots
– Shows the
relationship between
two variables
– Used as a follow up to
cause and effect
analysis to find
whether a stated
cause has impact on
quality
50
BITS Pilani, Pilani Campus
246
Chapter - 4
Fundamentals of Statistical
Concepts & Techniques in Quality
Control and Improvement
51
BITS Pilani, Pilani Campus
247
Basic Terminologies
• Population • Sample
– Set of all items that possess – A subset of population
a certain characteristic of – Eg. Selecting 200 plastic
interest cups from the week 23
– Eg. Average thickness of the output
plastic cups produced in • Statistic
week no. 23 (10,000) – A characteristic of a
• Parameter sample, which is used to
make inferences on the
– Is a characteristic of a population parameters
population, which describes that are unknown
it – Eg. Average thickness
– Eg. Average thickness of of 200 plastic cups is
10,000 cups 1mm
52
BITS Pilani, Pilani Campus
248
Basics of Probability
53
BITS Pilani, Pilani Campus
249
Example
54
BITS Pilani, Pilani Campus
250
Basics of Probability – Contd..
• Events
– Simple events cannot be broken into other events
– Compound events are made up of two or more simple
events
– Complementary of an event, say A, implies the
occurrence of everything except A. i.e. P(Ac) = 1 – P(A)
• Laws
– Additive law defines the probability of the union of 2 or
more events (say A & B), i.e. implies A may happen, B
may happen or both
– P(A u B) = P(A) + P(B) – P(A n B)
55
BITS Pilani, Pilani Campus
251
Simple and Compound Events
56
BITS Pilani, Pilani Campus
252
Basics of Probability – Contd..
• Laws – contd..
– Multiplicative law defines the probability of the
intersection of 2 or more events (say A & B), i.e.
implies all the events in the group occurs
– P(A n B) = P(A).P(B | A) = P(B).P(A | B)
– P(B | A) represents conditional probability, (i.e.,
probability that B occurs if A has)
• Independence
– Two events A & B are said to be independent, if the
outcome of one has no influence on outcome of
other
– P(B | A) = P(B) and hence P(A n B) = P(A).P(B)
57
BITS Pilani, Pilani Campus
253
Basics of Probability – Contd..
• Mutually Exclusive
– Two events A & B are said to be mutually exclusive, if
they cannot happen simultaneously.
– Probability of Intersection P(A n B) = 0 and
probability of union P(A u B) = P(A) + P(B)
– For mutually exclusive, the events A & B are
dependent. If A & B are independent, the additive
rule will be P(A or B or both) = P(A) + P(B) –
P(A).P(B)
58
BITS Pilani, Pilani Campus
254
Example
59
BITS Pilani, Pilani Campus
255
Example
60
BITS Pilani, Pilani Campus
256
Example
Suppose the operations that produce the length and the width
are not independent. If the length does not satisfy the
requirement, it causes an improper positioning of the part
during the width trimming and thereby increases the chances
of nonconforming width. From experience, it is estimated that
if the length does not conform to the requirement, the chance
of producing nonconforming widths is 60%. Find the
proportion of parts that will neither conform to the length
nor the width requirements.
61
BITS Pilani, Pilani Campus
257
Thank You
62
BITS Pilani, Pilani Campus
258
• Continuous variable
– Variable that can assume any value on a continuous scale
within a ‘range’ Eg. Viscosity of a resin
• Discrete variable
– Variable that can assume a finite number of values are called
discrete Eg. No. of defect in a shirt
– They are classified as acceptable or not
– Continuous characteristic can also be viewed as discrete. Eg.
Diameter of a hub in a tire
• Accuracy
– Refers to the degree of uniformity of the observations around
a desired value, such that on average, target is realized
• Precision
– Refers to the degree of variability of observation
4
BITS Pilani, Pilani Campus
262
Measures of Scale
• Nominal Scale
– Data variables are simply labels to identify an attribute
– Eg. Critical / Major / Minor
• Ordinal Scale
– Data has the properties of nominal data
– Data ranks or orders the observation
– Eg. Grades, 1- outstanding, 5 – poor
• Interval Scale
– Data has the properties of ordinal data and a fixed unit of
measure describes the interval between observations
– Eg. Temp. of water in diff stages of cooling during 2 hrs
interval
• Ratio Scale
– Data has the properties of Interval data and a natural
zero exists for measurement scale Eg. Wt. of cement bag:
100 kg, 100.2kg.
5
BITS Pilani, Pilani Campus
Basic Statistical Concepts 263
www.phdcomics.com
1 2 2 3 4 4 5 6 7
…
BITS Pilani, Pilani Campus
268
Distribution
N = 500
50
40
30
20
10
1 2 3 4 5 6 7
X X
= X=
N n
population sample
10 scores: 8, 4, 5, 2, 9, 13, 3, 7, 8, 5
ξΧ = 64
ξΧ/n = 6.4
Mode
Median
Mean
BITS Pilani, Pilani Campus
Central Tendency and Skew
274
Mode
Median
Mean
• Standard Deviation/Variance
– the average deviation of scores from the mean of the
distribution
– takes all scores into account
– less influenced by extreme values
BITS Pilani, Pilani Campus
Standard Deviation 278
mean mean
low variability high variability
small SD large SD
standard deviation = = =
- 3.4 11.56 2
-1 +1
26
BITS Pilani, Pilani Campus
284
Measures of Dispersion
• Standard Deviation
– Mostly used measure of dispersion and has the same unit as the
observation
– Measures the variability of the observation around the mean
– Population Standard Deviation, Sample Standard Deviation
• Inter Quartile Range
– Lower / First quartile (Q1) is the value such that 1/4th of the
observations fall below it and 3/4th fall above it (Q1 = 0.25 (n+1))
– Vice Versa for Third Quartile (Q3) (Q1 = 0.75 (n+1))
– Difference between 3rd quartile and 1st quartile (IQR = Q3 – Q1)
– Larger the value of IQR, greater the spread of data
– To find IQR, the data are ranked in ascending order and then Q1
and Q3 are calculated
28
BITS Pilani, Pilani Campus
286
9.2, 8.9, 8.7, 9.5, 9.0, 9.3, 9.4, 9.5, 9.0, 9.1
29
BITS Pilani, Pilani Campus
287
30
BITS Pilani, Pilani Campus
288
Measures of Skewness &
Kurtosis
• Skewness coefficient (V1)
– Describes the asymmetry of the dataset about the mean or
indicates the degree to which distribution deviates from
symmetry (formulae)
– Negatively skewed: V1= -ve, Mean < Median
– Positively skewed: V1= +ve, Mean > Median
– Not skewed: V1= 0, Mean = Median
• Kurtosis coefficient (V2)
– Is a measure of peakness of the dataset (formulae)
– Is also a measure of heaviness of the tails of distribution
– For normal distribution (Mesokurtic), V2 = 3
– Leptokurtic, More peaked, V2 > 3
– Platykurtic, Less peaked, V2 < 3
31
BITS Pilani, Pilani Campus
289
32
BITS Pilani, Pilani Campus
A sample of 50 coils to be used in an electrical circuit is
290
randomly selected, and the resistance of each is
measured (in ohms). Calculate the skewness coefficient
and kurtosis coefficient.
33
BITS Pilani, Pilani Campus
291
Thank You
34
BITS Pilani, Pilani Campus
292
• Sample data can be described with histograms, while population data are
described by probability distribution
• For discrete random variables, the probability distribution shows the value
that the variable can take and their corresponding probabilities
– P(xi) ≥ 1 for all i, P(xi) = P(X = xi); i = 1, 2,…
– Sum of all P(xi) = 1
– Some examples of discrete random variables are the number of defects
in an assembly, the number of customers served over a period of time,
and the number of acceptable compressors.
• Continuous random variable can take a infinite number of values and
hence probability distribution is expressed by Mathematical function
– f(x) >= 0 for all x, where P(a ≤ x ≤ b) = b∫af(x)dx
– Integration from minus infinity to plus infinity is one
– Almost all variables for which numerical measurements can be
obtained are continuous in nature: for example, the length of a pin, the
diameter of a bolt, the tensile strength of a cable, or the specific gravity
of a liquid.
2
BITS Pilani, Pilani Campus
294
Variables and Random Variables
Earlier, we saw that just like in algebra, we can use a variable to represent some
quantity, such as height.
i.e. It is just like a variable in
statistics, except each outcome has
now been assigned a probability.
1 2 3 4 5 6
0.3 0.2 0.1 0.25 0.05 0.1
“The probability
that…
3
BITS Pilani, Pilani Campus
295
Is it a discrete random variable?
4
BITS Pilani, Pilani Campus
296
Probability Distributions vs Probability Functions
There are two ways to write the mapping from outcomes to probabilities.
1 2 3 4
0.1 0.2 ?0.3 0.4
The table form that you know and love.
Advantages of distribution:
Probability for each outcome more explicit.
? 5
BITS Pilani, Pilani Campus
Example 297
{ HHH, ?
HHT,
HTT, Probability Function
HTH,
THH,?
THT,
TTH, ?
TTT }
6
BITS Pilani, Pilani Campus
Cumulative Distribution Function
298(CDF)
0 1 2
?
0.25 ?
0.75 1 ?
7
BITS Pilani, Pilani Campus
CDF 299
F(x)
p(x)
9
BITS Pilani, Pilani Campus
301
10
BITS Pilani, Pilani Campus
302
Discrete Distributions
• The discrete class of probability distributions deals with those random
variables that can take on a finite or countably infinite number of values.
• Hyper geometric distribution
– Useful in sampling from a ‘finite’ population without
replacement, where the outcomes are success or
failure
– If we consider, getting a nonconforming item as
success, the probability distribution of nonconforming
item (x) is given by P(x) = Dcx . (N-D)c(n-x) / Ncn
• D: no. of defects in population, x: no. of defects in
sample
• N: Size of population, n: Size of sample
– Mean µ = E(x) = nD/N
– Variance σ2 = Var(x) = nD/N(1 – D/N)((N-D)/N-1))
11
BITS Pilani, Pilani Campus
303
12
BITS Pilani, Pilani Campus
304
Discrete Distributions
• Binomial distribution
– Useful in sampling from a ‘large’ population
without replacement, or to sample with
replacement from a finite population
– Probability of success (p) on any trial is
assumed to be a constant
– Let x denote the no. of successes, if n trials are
conducted, probability of x successes is given
by
• P(x) = ncx . px (1-p)n-x , x = 0,1, 2..
– Mean µ = E(x) = np
– Variance σ2 = Var(x) = np(1 – p)
13
BITS Pilani, Pilani Campus
Example 305
14
BITS Pilani, Pilani Campus
306
15
BITS Pilani, Pilani Campus
307
Discrete Distributions
• Poisson distribution
– Used to model the no. of events that happen
within a product unit, space or volume or time
period. Eg. No. of machine breakdown per
month
– Probability distribution function of the no. of
events (x) is given by
• P(x) = e-λ . λx / x! , x = 0,1, 2..
– Mean or average no. of events is given by λ
– Mean µ = Variance σ2 = λ
– It is used as an approximation to the binomial,
when ‘n’ is large and ‘p’ is small, such that np = λ
is a constant or average no. of defects per unit is
constant
16
BITS Pilani, Pilani Campus
308
17
BITS Pilani, Pilani Campus
309
Discrete vs Continuous Distributions
You all know the distinction between discrete and continuous variables:
• Discrete: hair colour, shoe size, IQ, cars passing in the next hour, …
• Continuous: height, weight, time, …
Random variables are discrete or continuous when the outcomes are discrete or
continuous.
20
BITS Pilani, Pilani Campus
312
Continuous Distribution
• Normal Distribution
– Most widely used
– Probability density function of a normal random variable is f(x)
= 1/√2πσ exp[-(x - µ)2/ 2σ2], -∞ < x < ∞
– µ = population mean, σ = population std. deviation
– Change in mean changes the location of distribution. As µ
increases, distribution shifts right and vice versa
– As variance increases, the spread about mean increases
– Normal distribution is symmetric and Mean = Median = Mode
– Proportion of population values that fall in the range of µ +/- σ
is 68.26%, µ +/- 2σ is 95.44%, µ +/- 3σ is 99.74%,
– Shape of the density function changes for diff. values of µ & σ
and hence it needs standardization
21
BITS Pilani, Pilani Campus
313
22
BITS Pilani, Pilani Campus
Continuous Distribution – 314
Contd..
• Normal Distribution
– Area within certain limits can be found by
looking up tabulated values for a ‘Std. normal
distribution’
– Standardized normal random variable z = x -
µ / σ and z is the no. of standard deviations a
raw value x is from mean and has mean of 0
and variance of 1
– Z-value can be positive or negative and at
mean it is zero
– Density function (formula)
– Cumulative density function (formula)
– Sample problem 23
BITS Pilani, Pilani Campus
315
• Exponential Distribution
– Used in reliability analysis to describe the time to failure
of a component or system
– P.D.F is given by f(x) = λ exp[- λ x], x ≥ 0, λ = failure rate
– Represents a constant failure rate and is used to model
failures that happen randomly and independently
– Mostly used in the useful life period of a life cycle of a
product
– Standard deviation = Mean (µ) = 1 / λ
– Exponential cumulative distribution function (formula)
– Function is ‘memory less’ as the probability of
component’s life exceeding (s+t) time units, given that it
has lasted ‘t’ time units, is same as probability of the life
exceeding ‘s’ time units (eqn.)
– Sample problem
25
BITS Pilani, Pilani Campus
317
Thank You
27
BITS Pilani, Pilani Campus
319
1 2 3 4
0.1 0.2 ?0.3 0.4
The table form that you know and love.
Advantages of distribution:
Probability for each outcome more explicit.
? 2
BITS Pilani, Pilani Campus
Example 321
{ HHH, ?
HHT,
HTT, Probability Function
HTH,
THH,?
THT,
TTH, ?
TTT }
3
BITS Pilani, Pilani Campus
CDF 322
F(x)
p(x)
5
BITS Pilani, Pilani Campus
324
6
BITS Pilani, Pilani Campus
325
Discrete Distributions
• The discrete class of probability distributions deals with those random
variables that can take on a finite or countably infinite number of values.
• Hyper geometric distribution
– Useful in sampling from a ‘finite’ population without
replacement, where the outcomes are success or
failure
– If we consider, getting a nonconforming item as
success, the probability distribution of nonconforming
item (x) is given by P(x) = Dcx . (N-D)c(n-x) / Ncn
• D: no. of defects in population, x: no. of defects in
sample
• N: Size of population, n: Size of sample
– Mean µ = E(x) = nD/N
– Variance σ2 = Var(x) = nD/N(1 – D/N)((N-D)/N-1))
7
BITS Pilani, Pilani Campus
326
8
BITS Pilani, Pilani Campus
327
Discrete Distributions
• Binomial distribution
– Useful in sampling from a ‘large’ population
without replacement, or to sample with
replacement from a finite population
– Probability of success (p) on any trial is
assumed to be a constant
– Let x denote the no. of successes, if n trials are
conducted, probability of x successes is given
by
• P(x) = ncx . px (1-p)n-x , x = 0,1, 2..
– Mean µ = E(x) = np
– Variance σ2 = Var(x) = np(1 – p)
9
BITS Pilani, Pilani Campus
Example 328
10
BITS Pilani, Pilani Campus
329
11
BITS Pilani, Pilani Campus
330
Discrete Distributions
• Poisson distribution
– Used to model the no. of events that happen
within a product unit, space or volume or time
period. Eg. No. of machine breakdown per
month
– Probability distribution function of the no. of
events (x) is given by
• P(x) = e-λ . λx / x! , x = 0,1, 2..
– Mean or average no. of events is given by λ
– Mean µ = Variance σ2 = λ
– It is used as an approximation to the binomial,
when ‘n’ is large and ‘p’ is small, such that np = λ
is a constant or average no. of defects per unit is
constant
12
BITS Pilani, Pilani Campus
331
13
BITS Pilani, Pilani Campus
332
Discrete vs Continuous Distributions
You all know the distinction between discrete and continuous variables:
• Discrete: hair colour, shoe size, IQ, cars passing in the next hour, …
• Continuous: height, weight, time, …
Random variables are discrete or continuous when the outcomes are discrete or
continuous.
16
BITS Pilani, Pilani Campus
335
Continuous Distribution
• Normal Distribution
– Most widely used
– Probability density function of a normal random variable is f(x)
= 1/√2πσ exp[-(x - µ)2/ 2σ2], -∞ < x < ∞
– µ = population mean, σ = population std. deviation
– Change in mean changes the location of distribution. As µ
increases, distribution shifts right and vice versa
– As variance increases, the spread about mean increases
– Normal distribution is symmetric and Mean = Median = Mode
– Proportion of population values that fall in the range of µ +/- σ
is 68.26%, µ +/- 2σ is 95.44%, µ +/- 3σ is 99.74%,
– Shape of the density function changes for diff. values of µ & σ
and hence it needs standardization
17
BITS Pilani, Pilani Campus
336
18
BITS Pilani, Pilani Campus
Continuous Distribution – 337
Contd..
• Normal Distribution
– Area within certain limits can be found by
looking up tabulated values for a ‘Std. normal
distribution’
– Standardized normal random variable z = x -
µ / σ and z is the no. of standard deviations a
raw value x is from mean and has mean of 0
and variance of 1
– Z-value can be positive or negative and at
mean it is zero
– Density function (formula)
– Cumulative density function (formula)
– Sample problem 19
BITS Pilani, Pilani Campus
338
• Exponential Distribution
– Used in reliability analysis to describe the time to failure
of a component or system
– P.D.F is given by f(x) = λ exp[- λ x], x ≥ 0, λ = failure rate
– Represents a constant failure rate and is used to model
failures that happen randomly and independently
– Mostly used in the useful life period of a life cycle of a
product
– Standard deviation = Mean (µ) = 1 / λ
– Exponential cumulative distribution function (formula)
– Function is ‘memory less’ as the probability of
component’s life exceeding (s+t) time units, given that it
has lasted ‘t’ time units, is same as probability of the life
exceeding ‘s’ time units (eqn.)
– Sample problem
21
BITS Pilani, Pilani Campus
340
• Sampling Distribution
– Sample mean is most widely used estimator and
hence it is important to know the sampling
distribution of mean, which is given by Central
Limit Theorem (CLT)
– Suppose we have a population with mean ‘µ’ and
standard deviation ‘σ’. If random sample of size
‘n’ are selected from this, and if sample size is
large, following holds good: (CLT)
• Sampling distribution of the sample mean will be
approximately normal
• Mean of this sampling distribution of mean ‘µx’ will be
equal to population mean ‘µ’
• Std. deviation is σx = σ / √n
24
BITS Pilani, Pilani Campus
343
Inferential Statistics – Contd..
• Sampling Distribution
– The degree to which a sampling distribution of
a sample mean approximates a normal
distribution becomes greater as sample size ‘n’
becomes larger
– For population distribution that is symmetric
and uni-modal, sample size as small as 4 or 5
yield sample means that are approximately
normally distributed
– Variability of the sample mean, as measured by
standard deviation decreases as sample size
increases
25
BITS Pilani, Pilani Campus
344
Thank You
26
BITS Pilani, Pilani Campus
345
• Sampling Distribution
– Sample mean is most widely used estimator and
hence it is important to know the sampling
distribution of mean, which is given by Central
Limit Theorem (CLT)
– Suppose we have a population with mean ‘µ’ and
standard deviation ‘σ’. If random sample of size
‘n’ are selected from this, and if sample size is
large, following holds good: (CLT)
• Sampling distribution of the sample mean will be
approximately normal
• Mean of this sampling distribution of mean ‘µx’ will be
equal to population mean ‘µ’
• Std. deviation is σx = σ / √n
3
BITS Pilani, Pilani Campus
348
Inferential Statistics – Contd..
• Sampling Distribution
– The degree to which a sampling distribution of
a sample mean approximates a normal
distribution becomes greater as sample size ‘n’
becomes larger
– For population distribution that is symmetric
and uni-modal, sample size as small as 4 or 5
yield sample means that are approximately
normally distributed
– Variability of the sample mean, as measured by
standard deviation decreases as sample size
increases
4
BITS Pilani, Pilani Campus
349
5
BITS Pilani, Pilani Campus
350
Estimation of Product and Process
Parameters
• Can be done using Point Estimation and
Interval Estimation
6
BITS Pilani, Pilani Campus
351
Estimation – contd.
• Point Estimation
– Consists of a single numerical value that is used
to make an inference about an unknown
product or process parameter
– Eg. Estimate the length of a shaft produced in
certain month (25mm)
– A point estimator is said to be ‘unbiased’ if the
expected value or mean of sampling
distribution is equal to the parameter
– A point estimator is said to have minimum
variance, if its variance is smaller than that of
any other point estimator for the parameter
under consideration
7
BITS Pilani, Pilani Campus
352
Estimation – Contd..
• Interval Estimation
– A range of interval is determined, such that
there is some desired level of probability
that the true parameter value is contained
within it
– Called as confidence interval
– Say 2 end points (L, U), probability of
parameter ‘γ’ being contained in the interval
is some value (1 – α)
– P(L ≤ γ ≤ U) = 1 – α, which represents a 2-
sided confidence interval
8
BITS Pilani, Pilani Campus
353
Estimation – Contd..
• Interval Estimation
– If a large number of such confidence intervals
were constructed from large no. of independent
samples, then 100(1 – α)% of these intervals
would be expected to contain the parameter
value ‘γ’
– It can also be one sided. An interval of the type
L <= γ, such that P(L <= γ) = 1 – α is a one sided
lower 100(1 – α)% confidence interval for γ
– Context of the situation influence the type of
confidence interval selected
9
BITS Pilani, Pilani Campus
354
Estimation of Parameters
• Confidence interval about the mean
(Unknown)
– Case 1: Variance Known (Population)
• Based on Central Limit Theorem, the
sampling distribution of this mean is
approximately normal. A 100(1 – α)%
two-sided confidence interval for µ is
given by
X − z / 2 n
X + z
2
n
10
BITS Pilani, Pilani Campus
355
11
BITS Pilani, Pilani Campus
356
Estimation of Parameters
• Confidence interval about the mean
(Unknown)
– Case 2: Variance Unknown (Population)
• A random sample of size ‘n’ is chosen and the
sample mean X and sample variance is
calculated. The sampling distribution of
X - µ / (s / √n) follows ‘t’ distribution with
(n-1) degree of freedom, which is similar to
normal distribution.
s s
X − t( / 2),( n−1) X + t( / 2),( n−1)
n n
• Sample problem
12
BITS Pilani, Pilani Campus
357
1
n n 2
= 1 2 2 2
ν = no. of degrees of freedom s12 s2 2
n n
1 + 2
n1 −1 n2 −1
16
BITS Pilani, Pilani Campus
361
18
BITS Pilani, Pilani Campus
363
Estimation of Parameters
– Contd..
• Confidence interval for difference between 2
proportion
– A sample size of n1 selected with parameter p1
and a sample size of n2 selected with parameter
p2 pˆ1 (1 − p
ˆ1 ) ˆ 2 (1 − p
p ˆ2 )
( p1 − p2 ) − z
ˆ ˆ +
2 n1 n2
p1 − p2
ˆ1 (1 − p
p ˆ1 ) ˆ 2 (1 − p
p ˆ2 )
ˆ1 − p
(p ˆ 2 ) + z +
2 n1 n 2
19
BITS Pilani, Pilani Campus
364
Estimation of Parameters
– Contd..
• Confidence Interval for Variance
– A random variable x from a normal distribution with mean µ
and variance σ2 (both unknown)
– An estimator of σ2 is the sample variance s2
– Sampling distribution of s2 follows a ‘chi-square’ distribution
with (n-1) degrees of freedom given by
(n-1)s2 / σ2 = χ2n-1
– A chi square distribution is skewed to the right and a
100%(1-α) interval for population variance is given by
( n − 1) s 2 ( n − 1 ) s 2
2
2 / 2,n −1 12− / 2, n −1
– Sample problem
20
BITS Pilani, Pilani Campus
365
21
BITS Pilani, Pilani Campus
366
Estimation of Parameters
– Contd..
• Confidence Interval for ratio of two variances
– Random variable x1, x2 from a normal distribution with mean
µ1, µ2 and variance σ12, σ22 (both unknown)
– Random sample of size n1, n2 are chosen and the variance was
found to be s12, s22
– The ratio of these 2 variances will follow an ‘F’ distribution
with (n1-1) degrees of freedom in numerator and (n2 – 1) d.o.f
in denominator given by
s12 / 12
~ F( n1 −1),( n2 −1)
s2 / 2
2 2
22
BITS Pilani, Pilani Campus
367
23
BITS Pilani, Pilani Campus
368
Thank You
24
BITS Pilani, Pilani Campus
369
Chapter - 6
2
BITS Pilani, Pilani Campus
371
Introduction to Control Charts
• Emphasis is on process control and improvement
• A Control chart is a
– Graphical tool
– For monitoring the activity of an ongoing process
– It is called as ‘Shewhart’ chart
– Values of quality characteristic is plotted along ‘ordinate’
– Sample number or subgroups (in order of time) is plotted
along ‘abscissa’
– Example for quality characteristic includes Average length,
Average tensile strength, Average service time etc.
– Quality characteristic can be categorized as
• Variable: Numerical values can be obtained
4
BITS Pilani, Pilani Campus
373
More about Control Charts –
Contd..
• A Control chart has
– A centre line, represents the average value of the characteristic
– It indicates, where the process is centred
– UCL and LCL, used to make decision regarding the process
– If the points plot within the control limits and do not exhibit
any identifiable pattern, then process is in statistical control
– If a point plots outside the control limits or if an identifiable
random pattern exists process is out of statistical control
– Benefits
• Indicates when something may be wrong, so that corrective
action can be taken
• Pattern of plot can help in diagnosing the possible causes and
hence possible remedial action
• Helps in estimating the process capability of the process
6
BITS Pilani, Pilani Campus
375
Causes of Variation – Contd..
7
BITS Pilani, Pilani Campus
376
Statistical Basis for Control
Charts
• Centre line is based on the mean of a process
and is found by taking the average of the values
in sample
• Can also be desirable target or standard value
• Role of Normal distribution
– Value of the statistic plotted on a control chart are
assumed to have an approximately normal
distribution
– Any sample from population distribution that is
uni-modal and symmetric, the central limit theorem
states that if the plotted statistic is a sample
average, it tend to have a normal distribution
8
9
BITS Pilani, Pilani Campus
378
Statistical Basis – Contd..
• Why 3σ limits?
– Chosen in such a way that probability of the sample points
falling between them is ‘1’, if the process is in statistical
control
– Normal distribution theory states that a sample statistic will
fall within the limits 99.74% of the time, if the process is in
control
• Most common basis for deciding whether a process is out
of control is “the presence of a sample statistic outside the
control limits” and also it depends on other rules
• A control chart is a means of online process control
• If the control limits are calculated from current data, then
it tells whether the process is presently in control or not
• If the control limits are calculated from previous data,
based on a process that was in control, it helps to find
whether the process has drifted out of control
11
12
13
14
15
16
17
18
OC Curve
19
20
21
22
23
24
26
27
28
29
30
31
32
33
34
35
36
37
38
The length of industrial filters is a quality characteristic of interest. Thirty samples, each of
size 5, are chosen from the process. The data yields an average length of 110 mm, with the
process standard deviation estimated to be 4 mm.
(a) Find the warning limits for a control chart for the average length.
(b) Find the 3ó control limits. What is the probability of a type I error?
(c) If the process mean shifts to 112 mm, what are the chances of detecting this shift by the
third sample drawn after the shift?
(d) What is the chance of detecting the shift for the first time on the second sample point
drawn after the shift?
(e) What is the ARL for a shift in the process mean to 112 mm? How many samples, on
average, would it take to detect a change in the process mean to 116 mm?
39
40
Thank You
41
BITS Pilani, Pilani Campus
410
Chapter - 8
• Attribute
– It is a quality characteristic for which a numerical value is not
specified
– It is measured on a nominal scale, i.e., it does or does not meet
certain guidelines
– It is categorized according to the scheme or levels
• Nonconformity
– A quality characteristic that does not meet certain standards or
specification is said to be non conformity
– Eg. Length of a bar 40 ± 0.5. Both 40.6 and 42 are nonconformity
• Nonconforming item
– A product with one or more non-conformities such that it is unable
to meet the intended standards and is unable to function as
required
– It is possible to have several non-conformities on a product without
being classified as a nonconforming item
x
ˆ
p =
where x = no .of nonconforming item in sample, n
n
= sample size
10
g = no. of samples
pˆ i xi
CLp = p = i =1 = i =1
xi = no. of non conforming items g ng
– Variance is calculated from
p (1 − p )
Var ( pˆ ) =
n 12
13
No. of defectives
14
pˆ x i i
CLp = p = i =1
= i =1
0.129
g ng
p (1 − p ) 0.129 +
UCLp = p + 3 . 3 [0.129(1 – 0.129)/50]
n
= 0.271
p (1 − p )
LCLp = p − 3 .
n 0.129 -
3 [0.129(1 – 0.129)/50]
=0 15
BITS Pilani, Pilani Campus
425
p chart – contd..
16
17
18
19
20
21
22
25
26
Day 1 2 3 4 5 6 7 8 9 10
n 500 550 700 625 700 550 450 600 475 650
D 5 6 8 9 7 8 16 6 9 6
p 0.010 0.011 0.011 0.014 0.010 0.015 0.036 0.010 0.019 0.009
pbar 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.014 0.014
UCL 0.029 0.028 0.027 0.028 0.027 0.028 0.030 0.028 0.030 0.027
LCL 0 0 0 0 0 0 0 0 0 0
27
0.060
0.050
0.040
0.030
p
0.020
0.010
0.000
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Day
28
29
30
31
32
n
– Mean of these sample size is i
n= i =1
g g
34
36
37
38
39
40
– n = sample size
g
np = np (1is
– Standard deviation of number of nonconforming − p)
41
UCLnp = np + 3 . np (1 − p )
LCLnp = np − 3 . np (1 − p )
If LCL is negative, then it is taken as zero
Standard given
– If standard is specified for number of non conforming
items is given as npo, then the limits are given as
CLnp = npo
UCLnp = npo + 3 . npo (1 − po )
LCLnp = npo − 3 . npo (1 − po ) 42
No. of defectives
43
44
45
46
47
CLc = co
UCLc = co + 3 . co
LCLc = co − 3 . co
50
51
52
53
ni
i =1
LCLu = u − 3 .
u
ni
As the sample sizes increases, control limits draws
closer
If ni = 1, all formula will be equal to that of c - chart
55
56
1 500 14 10 1.4
2 400 12
3 650 20
4 500 11
5 475 7
6 500 10
7 600 21
8 525 16
9 600 19
10 625 23 57
58
59
60
(U-chart)
‘c’ and ‘u’ chart treat all types of non conformities
equally, regardless of their degree of severity
– Eg. Monitor A has trouble in colour while Monitor B has 5
scratch marks on its surfaces – here Monitor A defect is
more serious than Monitor B’s
An alternative approach is to assign “weights” to non
conformities according to their relative degree of
severity
The quality rating system, which rates demerits /
unit is called U-chart and is often helpful in service
applications
Classification of non conformities
– Classification is based on degree of seriousness, ANSI
standard A3 classifies in the following manner:
• Class 1 defects: Very serious and defects lead directly to
‘severe’ injury or to catastrophic economic losses
61
(U-chart) – Contd..
Classification of non conformities – contd..
• Class 2 defects: serious and defects that lead to
‘significant’ injury or to significant economic losses
• Class 3 defects: major and defects that cause ‘major’
problems with normal use of product or service rendered
• Class 4 defects: minor and defects that cause ‘minor’
problems with normal use of product or service rendered
Once classification of defects or non conformities has
been established, demerits per unit are assigned to each
class
Definition of classes and no. of classes are not rigid and
varies with respect to organization and the problem
environment
Assigned weights for defects is also user dependent, while
ANSI standard uses 100, 50, 10, 1 for the above
mentioned classes
62
(U-chart) – Contd..
Construction
– Let us have 4 categories of non conformities
– Let the sample size be ‘n’
– Let c1, c2, c3 and c4 denote the total no. of non
conformities in the sample for 4 categories
– Let w1, w2, w3 and w4 denote the weights
assigned to each category
Assumption
– Non conformities in each category is
independent of defect in other categories
– Occurrence of non conformities in each
category is given by Poisson distribution
For a sample size of ‘n’, total no. of demerits is given by
D = w1c1 + w2c2 + w3c3 + w4c4
63
(U-chart) – Contd..
Demerits per unit of the sample are given by
D w1c1 + w2 c2 + w3c3 + w4 c4
U= =
n n
U is the linear combination of independent
poisson random variable
Centre line of the U chart is U = w1u1 + w2u2 + w3u3 + w4u4
– u1 ,....u4 average no. of non conformities per
unit in respective classes g
and are calculated
as
i =1
ci
u = g
ni
i =1
64
(U-chart) – Contd..
Estimated standard deviation of U is given by
w1 u1 + w2 u2 + w3 u3 + w4 u4
2 2 2 2
ˆU =
n
Control limit of the U chart is
UCLU = U + 3.ˆU
LCLU = U − 3.ˆU
If LCL is negative, then take LCL as zero
65
66
67
68
Thank You
69
BITS Pilani, Pilani Campus
479
BITS Pilani
Pilani Campus
2
Introduction 481
3
BITS Pilani, Pilani Campus
Introduction – Contd.. 482
5
BITS Pilani, Pilani Campus
Preliminary decisions 484
6
BITS Pilani, Pilani Campus
Preliminary decisions – contd..
485
• Sample size
– Size is normally between 4 and 10 and in industry it
will be 4 to 5
– Larger the sample size, better the chance of detecting
small shifts
– Based on factors like cost of inspection or cost of
shipping a nonconforming item to the customer etc.
• Frequency of sampling
– Depends on the cost of obtaining information
compared to the cost of non detecting a non
conforming item
– As process is brought to control, frequency of
sampling decreases
7
BITS Pilani, Pilani Campus
Preliminary decisions – Contd..
486
8
BITS Pilani, Pilani Campus
Control chart format 487
9
BITS Pilani, Pilani Campus
Control chart for Mean and
488Range
• Development of chart
– Using a pre-selected scheme and sample size record
measurements of the selected quality characteristic
– For each sample, calculate the sample mean and range
Ri = xmax - xmin x1 + x 2 + + x n
x=
n
– Obtain and draw the centre line and trial control limit
• Find the mean of all sample mean (Formula)
• Find the mean range of all samples (Formula)
– 3σ control limits for Mean chart is given by
ˆ
Z / 2 x = Z / 2 X 3 X = X 3
n n
– For normally distributed population, the distribution of the
statistic’s relative range (w) = R / σ and it is dependent on
sample size ‘n’
10
BITS Pilani, Pilani Campus
Control Chart for Mean (Xbar)
489
11
BITS Pilani, Pilani Campus
Control chart for Mean and
490Range –
Contd..
– Mean of w is given by d2
– Estimate of the process standard deviation is
ˆ = R /d 2
– (UCL, LCL) = X 3 R = X A2 R
d2 . n
12
BITS Pilani, Pilani Campus
Control Chart for Range (R) 491
13
BITS Pilani, Pilani Campus
Control chart for Mean and
492Range –
Contd..
• Plot the values of the range on the chart and find
whether points are in statistical control
– An R chart is analysed before X-bar chart to determine
out of control situations, as R chart reflect process
variability, which should be brought to control.
– If R chart shows out of control, then the X-bar chart is
meaningless
• Delete the out of control points for which remedial
action has been taken to remove special causes and
the remaining samples are used to obtain revised
limits
• A point of interest is about the point that falls
below the LCL, when LCL is greater than zero
14
BITS Pilani, Pilani Campus
Control chart for Mean and
493Range –
Contd..
• These points are desirable because they indicate
unusually small variability, within the sample
and might be due to special causes
• This condition helps in further reducing our
process variability
• Implement the control chart
Why two charts?
• X bar chart monitors the mean between sample
values
• R chart monitors the variation within sample.
15
BITS Pilani, Pilani Campus
Example 494
16
BITS Pilani, Pilani Campus
Example – contd.. 495
17
BITS Pilani, Pilani Campus
Example – contd.. 496
18
BITS Pilani, Pilani Campus
Example – contd.. 497
20
BITS Pilani, Pilani Campus
Example – contd.. 499
21
BITS Pilani, Pilani Campus
500
22
BITS Pilani, Pilani Campus
501
23
BITS Pilani, Pilani Campus
Standardized control chart
502
ni X i
• Mean of sample average X = i =1g
n i
24 i =1
BITS Pilani, Pilani Campus
Standardized control chart
503– Contd..
(n − 1) si
2
i
ˆ = i =1
g
(n
i =1
i − 1)
25
BITS Pilani, Pilani Campus
Standardized control chart
504– Contd..
R
CLR = d 2 . 0 (ˆ = )
d2
UCLR = R + 3 . R = d 2 . 0 + 3. d 3 0 = D2 . 0
LCLR = R − 3 . R = d 2 . 0 − 3 . d 3 0 = D1. 0
27
BITS Pilani, Pilani Campus
Interpretation and Inferences
506
• Caution
– Sample plots may fall outside the limit, even
though no special causes are present
– Reason being that desirable standards may
not be consistent with the process conditions
– It is easy to meet a desirable target value for
process mean than it is for process
variability (Range)
28
BITS Pilani, Pilani Campus
Interpretation and Inferences
507 –
Contd..
• Interpretation
– Difficult and needs thorough knowledge
about different process parameters on
quality characteristic
– When R-chart is brought to control, many
special causes for the Xbar chart are
eliminated as well
– Xbar chart monitors the centering of the
process and a jump indicates process
average has jumped
29
BITS Pilani, Pilani Campus
Interpretation and Inferences
508– Contd..
30
BITS Pilani, Pilani Campus
Control chart patterns and
509corrective
action
• A ‘non-random identifiable’ pattern in the
plot of a chart might provide reason to look
for special cause in a process
• There are about 15 typical patterns
identified by Western Electric company and
9 of them have been discussed here
• Natural Patterns
– No identifiable arrangement of the plotted point
exists
– No point fall outside the control limits
– Majority of the points are near the centre line and
few points close to control limits
– Demonstrates the presence of stable system of
common causes
31
BITS Pilani, Pilani Campus
Control chart patterns – contd..
510
32
BITS Pilani, Pilani Campus
Control chart patterns – Contd.
511
33
BITS Pilani, Pilani Campus
Control chart patterns– Contd..
512
34
BITS Pilani, Pilani Campus
Control chart patterns – Contd..
513
• Trending pattern
– Differs from gradual shift in level, that trends do not stabilize or
settle down
– Represents changes that steadily increase or decrease
– For X-bar chart, can be due to tool wear, deterioration of
equipment, build up of debris on jigs and fixtures, change in
temperature etc.
– For R-chart, it may be due to improvement in operator skill due to
on job training, decrease due to fatigue etc.
35
BITS Pilani, Pilani Campus
Control chart patterns – Contd..
514
• Cyclic patterns
– Characterized by repetitive periodic behaviour in the system
– Cycles of low and high points will appear on the control chart
– X-bar chart may exhibit because of rotation of operators, periodic
change of temperature or humidity, seasonal variation in incoming
components
– R-chart may exhibit this pattern because of operator fatigue and
getting energized in subsequent breaks, a difference between shifts,
periodic maintenance of equipments etc.
– If samples are taken so infrequently, only the high and low points will
be represented
36
BITS Pilani, Pilani Campus
Control chart patterns – Contd..
515
• Wild patterns
– Can be classified as Bunches and Freaks
– Cluster of several observation that are different
from other points and special causes are
associated with these points
– Freaks
• are caused by external disturbances that influence one or
more samples
• They are points that are too small or large with respect to
control limits and fall outside the control limits and
hence easy to identify
• Care should be taken that no measurement or recording
error is associated with that freak point
• Some special causes may be sudden, very short-lived
power failures, use of new tool for a brief test period etc.
37
BITS Pilani, Pilani Campus
Control chart patterns – Contd..
516
38
BITS Pilani, Pilani Campus
Control chart patterns and
517corrective
action – Contd..
• Wild patterns – Contd.
– Bunches
• Cluster of several observation that are different from other
points
• Possible causes may be use of new vendor, use of a different
machine, use of new operator etc., for a short time period.
39
BITS Pilani, Pilani Campus
Control chart patterns and
518corrective
action – Contd..
• Mixture patterns
– Effect of two or more population in the sample
– Characterized by points that fall near the control limits, with
absence of points near the centre line
– Might be due to material from two different vendors, different
production method, two or more machine being represented
40
BITS Pilani, Pilani Campus
Control chart patterns and
519corrective
action – Contd..
• Stratification patterns
– Is also due to presence of two or more population distribution
– Output is combined or mixed and samples are selected from it
– Majority of the points fall close to centre line, with very few points
near the control limits
– Can be misinterpreted as indicating unusually good control
41
BITS Pilani, Pilani Campus
Control chart patterns and
520corrective
action – Contd..
• Interaction patterns
– Occurs when the level of one variable affects the
behaviour of other variables associated with the quality
characteristic
– Interaction pattern can be detected by changing the
scheme for rational sampling
– Example, low pressure and low temperature may
produce a desirable effect on output characteristic
– Effective sampling method would involve controlling
the temperature at several high values and then
determining the effect of pressure on output
characteristic for each temperature value
– If the R-chart shows the sample range to be small, then
information regarding the interaction could be used to
establish desirable process parameter settings
42
BITS Pilani, Pilani Campus
Control chart patterns – Contd..
521
• Interaction patterns
43
BITS Pilani, Pilani Campus
Control chart for Mean and Standard
522
deviation (X-bar and s chart)
44
BITS Pilani, Pilani Campus
Control chart for Mean and Standard
523
deviation (X-bar and s chart)
1 ( n − 2) !
2 2
2
c4 = ( n − 3)
( n − 1) !
2
45
BITS Pilani, Pilani Campus
Control chart for Mean and Standard
524
deviation (X-bar and s chart) – Contd..
g
• No Given Standards s i
– Centre line of a s-chart is CLs = s =
i =1
g
s
where g is the no. of samples and = c
ˆ
4
si is the standard deviation of the ith sample
– UCLs= s + 3. s = s + 3. . 1 − c4
2
c4
s
LCLs = s − 3. . 1 − c4 = B3 s
2
c4 46
BITS Pilani, Pilani Campus
Control chart for Mean and Standard
525
deviation (X-bar and s chart) – Contd..
• No Given Standards
• X-bar chart
– Centre line is similar to that of X-bar chart
– Control limits are given by
s
UCLX = X + 3. = X + A3 s
c4 n
s
LCLX = X − 3. = X − A3 s
c4 n
• S-chart is constructed first as the standard deviation of
X-bar is dependent on ‘s’ and if the s-chart is not in
control, any estimate of the standard deviation of X-bar
chart is unreliable
47
BITS Pilani, Pilani Campus
Control chart for Mean and Standard
526
deviation (X-bar and s chart) – Contd..
• Given Standard
– If the target standard deviation is given as σ0 then, centre line of
a s-chart is
CLs = c4 0
– Hence control limits are given by
UCLs = c4 0 + 3 s = c4 0 + 3. 0 . 1 − c4 = B6 0
2
– X-bar Chart
• Target value is specified as Xo then control limits are given
by CL = XX 0
3
UCLX = X 0 + A 0 , where A =
n
LCLX = X 0 − A 0 48
BITS Pilani, Pilani Campus
Example 527
49
BITS Pilani, Pilani Campus
528
50
BITS Pilani, Pilani Campus
Control chart for individual
529 units – MR
Chart
• Chart for sample size is 1, i.e. for individual units
• Reasons for sample size to be ‘1’
– The rate of production is low
– Testing process may be destructive and the cost of the item is
very high
– If every manufactured unit is inspected
• The value of the quality characteristic is expressed as ‘X’
• Variability of the process is estimated from the ‘Moving
Range’, that are found from successive observations
– Moving Range of 2 observations is simply the difference
between them
– Moving Range are co-related, because they use common rather
than individual values and hence the pattern of MR chart must
be interpreted carefully
51
BITS Pilani, Pilani Campus
Control chart for individual
530 units – MR
Chart – Contd..
• Also, it can’t be assumed that X-values for the individual
units are normally distributed as in other control charts
• Hence checking of the distribution of individual values
is done first, by using frequency histogram to verify the
shape, skewness, kurtosis of the distribution
• No Given Standards
– An estimate of the process standard deviation is
– Where MR is the average of moving range of successive
MR
observation ˆ =
– If there are ‘g’ observations, there will be (g -1) moving rangesd 2
– Centre line is given by CL = MR
MR
52
BITS Pilani, Pilani Campus
Control chart for individual
531 units – MR
Chart – Contd..
– Control limits are given by
MR
UCLMR = MR + 3. = D4 MR
d2
MR
LCLMR = MR − 3. = D3 MR
d2
– For n = 2, D4 = 3.267, D3 = 0 and hence UCL = 3.267MR and LCL
=0
• X-Chart
– Centre line of X-chart is given by CL X = X
– Control limits are given by
MR
UCLX = CL + 3 = X + 3.
d2
MR
LCLX = CL − 3 = X − 3.
d2
53
BITS Pilani, Pilani Campus
Example 532
54
BITS Pilani, Pilani Campus
Example 533
55
BITS Pilani, Pilani Campus
Example 534
56
BITS Pilani, Pilani Campus
Example 535
57
BITS Pilani, Pilani Campus
Control chart for individual
536 units – MR
Chart – Contd..
• Given Standards
– If standard values are specified as Xobar and σ0, then,
CLX = X o
UCLX = X o + 3. o
LCLX = X o − 3. o
– Assuming n = 2, (difference between 2 values only)
CLMR = d 2 o = 1.128 o
UCLMR = D4 d 2 o = 3.685 o
LCLMR = D3d 2 o = 0
58
BITS Pilani, Pilani Campus
Example 537
59
BITS Pilani, Pilani Campus
538
60
BITS Pilani, Pilani Campus
539
Thank You
61
BITS Pilani, Pilani Campus
540
Chapter - 11
Reliability
2
BITS Pilani, Pilani Campus
542
Introduction
• What is Reliability?
– “Reliability is the probability of a product performing its
intended function for a stated period of time under certain
specified conditions”
• Four aspects of Reliability
– Reliability is a probability related concept
– Functional performance of the product has to meet certain
stipulations
• Product design will ensure development of product that meet
or exceed stipulated requirements
• Eg. Strength of cable = 1000Kg, operation it should be more
– Reliability implies successful operation over a certain
period of time
– Operating or environmental condition under which product
use takes place are specified.
• Example: Reliability of a cable is given as having a
probability of successful performance of 0.90 in
withstanding 1000Kg of load for 2 years under dry 4
condition BITS Pilani, Pilani Campus
544
Life cycle curve
• Most product goes through three distinct phases from
product creation to wear out
• A life cycle curve is a plot between failure rate λ and time
• Also called as ‘Bathtub curve’ (Figure)
• Consists of three phases, namely: Debugging or infant
mortality, Chance failure / useful life time, Wear out / aging
• Debugging phase exhibits a drop in the failure rate as initial
problems identified during prototype testing are ironed out
• In chance failure phase, the failure rate is constant and here
failure occurs randomly and independently. Also, it is called
as useful period
• In wear out phase, an increase in failure rate is observed, as
the product approaches its end of their useful life as parts
age and wear out
5
BITS Pilani, Pilani Campus
545
Life cycle curve – contd..
6
BITS Pilani, Pilani Campus
546
Probability distributions
8
BITS Pilani, Pilani Campus
548
Probability distributions – contd..
9
BITS Pilani, Pilani Campus
549
Probability distributions – contd..
10
BITS Pilani, Pilani Campus
550
Probability distributions – contd..
• Exponential distribution – contd..
– Reliability function R(t) for the exponential distribution is
shown in the figure
– At time 0, reliability is 1 and
t
it decreases exponentially
with time
= 1 − e dt = e
−t −t
13
BITS Pilani, Pilani Campus
553
Probability distributions – contd..
• Weibull distribution
– “It is used to model the time to failure of
products that have a varying failure rate
– Hence a candidate to model the debugging or
wear out phase
– It is a three parameter distribution where
density function is given by
t −
−1
t −
f (t ) = e − , t
– The parameters are:
• Location parameter is given by
• Scale parameter is (− )
• Shape parameter is ( 0 )
( 0 ) 14
BITS Pilani, Pilani Campus
554
Probability distributions – contd..
• Weibull distribution – contd..
– The probability density function varies for
different values of these parameters (Figure)
– Weibull distribution reduces to exponential
distribution, when = 0 and β = 1
– For reliability modelling, the location parameter
=0
– For α = 1 and β = 1, the failure rate decreases
with time and can therefore be used to model
components in debugging phase”
– For α = 1 and β = 3.5, the failure rate increases
with time and so can be used to model products
in the wear out phase”. In this case, the weibull
distribution approximates normal
15
BITS Pilani, Pilani Campus
555
Probability distributions – contd..
17
BITS Pilani, Pilani Campus
557
18
BITS Pilani, Pilani Campus
558
System Reliability
• System reliability
– “Reliability of the product (made up of a number of
components) is determined by the reliability of each
component and also by the configuration of the system
consisting of these components”
– Product design, manufacture, maintenance influence
reliability, but design has a major role
– One common approach for increasing the reliability of the
system is through “redundancy in design”, which is usually
achieved by placing components in parallel.
– As long as one component operates, the system operates
• Systems with components in series
– For the system to operate, each component must operate
– It is assumed that the components operate independently
of each other (Failure of one component has no influence on
the failure of any other component)
A B C
19
BITS Pilani, Pilani Campus
559
Series System Example
20
BITS Pilani, Pilani Campus
560
System Reliability – contd..
21
BITS Pilani, Pilani Campus
561
22
BITS Pilani, Pilani Campus
562
System Reliability – contd..
• Systems with components in series – contd..
– Use of the Exponential Model
• If the system is in chance failure phase, a constant
failure rate could be justified based on which we can
calculate failure rate, mean time to failure and system
reliability
• Suppose the system has ‘n’ components in series
• Each component has exponentially distributed time-to-
failure with failure rates given by
• The system reliability is given by , − − − −
1 2 n
i
i =1 i = = constant
• When all components have same failure rate, If
then
1
MTTF =
n
26
BITS Pilani, Pilani Campus
566
System Reliability – contd..
• Systems with components in parallel –
contd..
– The probability of failure of each component is
given by Fi = 1-Ri.
– System fails only if all the components fail and
hence the probability of system failure
n
is
Fs = (1 − R1 )(1 − R2 ) − − (1 − Rn ) = (1 − Ri )
i =1
= 1 − (1 − e )
n
− i t
Rs
i =1
• Time to failure of the system is not
exponentially distributed
28
BITS Pilani, Pilani Campus
568
System Reliability – contd..
• Systems with components in parallel –
contd..
– Use of Exponential model
• In the special case, where all the components have the
same failure rate the system reliability is
( ) Rs = 1− (1 − e )
n
−t n
Rs = 1 − 1 − e − i t
i =1
29
BITS Pilani, Pilani Campus
569
System Reliability – contd..
• Complex system
– A complex system is one which has
components that are both in series and in
parallel
– Assumption
• Components operate independently
• Time to failure of each component is assumed
to be exponentially distributed
– The above described methods are used for
calculating the reliability and failure rate
30
BITS Pilani, Pilani Campus
570
Complex system – Example
31
BITS Pilani, Pilani Campus
571
Complex system – Example
32
BITS Pilani, Pilani Campus
572
Complex system – Example
33
BITS Pilani, Pilani Campus
573
System Reliability – Home work
• A company assembles a product from four major
components arranged as follows
B
A C D
• The components are purchased from different
vendors, who have supplied the following
reliability data:
Vendor
Component 1 2 3
A 0.94 0.95 0.92
B 0.86 0.80 0.90
C 0.90 0.93 0.95
34
D 0.93 0.95 0.95
BITS Pilani, Pilani Campus
574
System Reliability – contd..
• System with standby components
– “In a stand by configuration one or more
parallel components wait to take over
operation upon failure of the currently
operating component”
– It is assumed that only one component in
parallel configuration is operating at any
given time
– Hence the system reliability is higher than
for comparable systems with components
in parallel
– In parallel systems, all components are
assumed to be operating simultaneously
35
BITS Pilani, Pilani Campus
575
Standby system – Example
36
BITS Pilani, Pilani Campus
576
System Reliability – contd..
• Standby components – contd..
– A standby system with a basic component and
two standby components in parallel (Figure) is
shown
Am
As
As
– Typically a failure sensing mechanism triggers
the operations of a stand by component when
the currently operating component fails
37
BITS Pilani, Pilani Campus
577
System Reliability – contd..
• System with standby components – contd..
– Use of Exponential Model – contd..
• If the time to failure of the components is assumed to
be exponential with failure rate , the number of failure
in a certain time ‘t’ adheres to a Poisson distribution
with parameter t
• Hence probability of ‘x’ failures in time ‘t’ is given by
P ( x) =
(t ) x
e t
x!
• For a system that has a basic component in parallel
with one standby component, the system will be
operational at time ‘t’ as long as there is no more than
one failure. Therefore, the system reliability would be :
Rs = e − t
+e − t
(t ) 38
BITS Pilani, Pilani Campus
578
System Reliability – contd..
• System with standby components –
contd..
– Use of Exponential Model – contd..
• For a system (stand by) with a basic
component and two standby components, the
system will be operational if the number of
failures is less than or equal to 2, then
2!
• For n components on stand by, the reliability
and mean time to failure is given by
− t
Rs = e 1 + t +
( t )2
+
( t )3
+ ....... +
(t )n
MTTF =
n +1
2! 3! n!
BITS Pilani, Pilani Campus
579
Problems
• System with standby components –
Exercise Problem 14
• A standby system has a basic unit with four standby
components. The time to failure of each component
has an exponential distribution with a failure rate of
0.008 per hour.
– For a 400 hour operation period, find the reliability of the
standby system.
– What is its mean time to failure of the above system?
– Suppose all five components are operating simultaneously
in parallel, what would be the system reliability?
– What would be the mean time to failure for the parallel
system?
40
BITS Pilani, Pilani Campus
580
Problems – contd..
• System with standby components –
Exercise Problem 14 Solution
− t
R = e 1 + t +
(t )2
+
(t )3
+ ....... +
(t )n
MTTF =
n +1
s n!
2! 3!
• Rs = exp[-0.008(400)] [1 + 0.008(400) +
(0.008(400))2 / 2 + (0.008(400))3 / 6 +
(0.008(400))4 / 24 = 0.7806
41
BITS Pilani, Pilani Campus
581
Problems – contd..
• System with standby components –
Exercise Problem 14 Solution
• If all five units were operating in parallel,
system reliability would be
– Rs = 1 – [1 – exp(-0.008(400))5] = 0.18786
• In that case, MTTF is
– (1/0.008) (1+ ½ + 1/3 + ¼ + 1/5) = 285.4167
hours
42
BITS Pilani, Pilani Campus
582
BITS Pilani
Pilani Campus
Chapter – 13
45
BITS Pilani, Pilani Campus
585
Introduction – contd..
• Few more examples include personal services,
education, banking, medicinal, financial, public
utilities, transportation etc.
• Service industries provide a tangible product and an
intangible component that affects customer
satisfaction
• Two parties are involved – one that assists or provides
the service (Vendor) and the party receiving service
(Vendee)
• Service functions are found in manufacturing also
done by staff personnel – provides expertise to the
operating departments
– Includes, customer services, accounting, payroll, R&D etc.
• Differences in the manufacturing and service sector
– A detailed differences are tabulated
46
BITS Pilani, Pilani Campus
586
Differences in manufacturing and
service sector
• Manufacturing • Service
– Product is tangible – Tangible and intangible
– Back orders are possible – Cannot be stored, if not
– Producer or company is used
the only party involved – Producer and consumer
in the making of the are both involved in the
product delivery of service
– Product can be resold – Service cannot be resold
– Formal specifications – No formal specification
provided by customer given by customer
– Customer acceptance of – Customer satisfaction is
product is quantifiable difficult to quantify
47
BITS Pilani, Pilani Campus
587
Service quality characteristics
• Quality of service can be broken down into two categories:
– Effectiveness deals with meeting the desirable service
attributes that are expected by the customer.
– Efficiency concerns the time required for the service to be
rendered
• Service quality characteristics can be grouped under 4
categories
– Human factors and behavioral characteristics
• Influenced by the attitude and behaviour of the provider
• Includes eagerness to help, thoughtfulness, complacency, courtesy
etc
• Can be developed through training while some are inherent in
individual
• Hence proper screening of employees and assignment of job is
important
– Behavioral characteristics
• Attitude of the customers is beyond the organization control
• Customer mind-sets are often a function of what they expect to
receive
48
BITS Pilani, Pilani Campus
588
Service quality characteristics –
contd..
• Behavioral characteristics – contd..
• Customer expectations can be influenced by advertisement
and reputation
• In turn, behavioural pattern gets affected
• Measurement of attitudes and behavioural characteristics
is not as simple and well-defined
• Timeliness characteristics
• Service that is not used in a given span of time cannot be
stored
• Timeliness with which a service is performed is critical to
customer satisfaction
• Characteristics related to timeliness are categorized by the
service phase, which they are associated
• Categories might include the time to order the service, the
waiting time before the service is performed, the time to
serve, and the post service time
49
BITS Pilani, Pilani Campus
589
Service quality characteristics –
contd..
• Service nonconformity characteristics
• A nonconformity is a deviation from the ideal level
• Eg. - No. of errors per 100 vouchers, no. of data entry errors
per 1000 keystrokes, No. of complaints per 50 guests etc
• Target performance level should be zero nonconformities
• Goal is to achieve this through quality improvement measures
• Quality characteristics in this category are well defined and
more readily measured
• Facility related characteristics
• Physical characteristics of faculty associated with service and
delivery can affect customer satisfaction
• Eg. – Ambience in the restaurant, Cyber cafe in a petrol bunk,
Appearance of waiter etc.
• Characteristics are not as clearly defined and measurable but
better than behavioural
50
BITS Pilani, Pilani Campus
590
Service quality characteristics –
Examples
• Facility related
characteristics
51
BITS Pilani, Pilani Campus
591
Service quality characteristics –
Examples
• Behavioral
characteristics
52
BITS Pilani, Pilani Campus
592
Service quality characteristics –
Examples
• Timeliness characteristics
• Service nonconformity characteristics
54
BITS Pilani, Pilani Campus
594
Measuring service quality
• Characteristics • Measures
– No. of customer complaints,
– Human factors and
No. of compliments based
behavioural on behavioural factors of
persons in service
– Time to process a
– Timeliness transaction, Waiting time to
receive the baggage, etc
– No. of billing errors in
mobile phone usage, No. of
– Service nonconformity errors in data entry of
marks in database etc.
– No. of complaints due to
insufficient legroom in bus,
– Facility related Lack of lab facilities in a
institute etc.
55
BITS Pilani, Pilani Campus
595
Measuring service quality – contd..
• In terms of ease of quantification and measurement,
categories can be ranked as service nonconformity,
timeliness, facility related and human and behavioural
factors
• Success of many service functions is determined by the
interaction between the provider and the customer
• Because of subjectiveness of quality characteristics,
measurement and evaluation of service quality is difficult,
rather defining the measurement unit itself is problematic
• Many factors influence the behavioral aspects and are
outside the influence of the company and they cannot be
predicted and may cause large performance variation
• Eg. Family life, mental outlook, unforeseen personal events
affect employee behaviour
• To counteract these performance variations in human
behavior, procedures that generate representative statistics
of performance can be devised
56
BITS Pilani, Pilani Campus
596
Measuring service quality – contd..
• Randomly choosing samples of performance from the time
interval under consideration is one way to eliminate the bias
• In situation where we know that behavioral patterns vary
greatly based on the time period (for instance, if error rates
are high in the first and eighth hours of an 8 hour work per
day), we can select sampling plan that adequately reflect this
(stratified sampling)
• Another difficulty is that significant difference exists
between individuals.
• Thus, even though the scheme of stratified sampling is used
to select appropriate samples that reflect the performance of
an individual, it is not obvious whether this same scheme can
be collectively applied to group of individuals
• Individuals vary in their peak performance periods, as
people work best in early morning and others at night and
sampling plan can be designed to reflect them
57
BITS Pilani, Pilani Campus
597
Techniques for Evaluating Service
Quality
• In the service sector, ergonomic, anthropometric, and
behavioral characteristics are important as are the physical
characteristics of the service systems and timeliness with
which the service is provided
• Descriptive statistics that provide numerical measures and
the corresponding graphical methods can be used to
describe distributions of service quality characteristics and
their summary measures
• Eg. Distribution of waiting time before seen by physician can
be shown by frequency histogram and trend chart to find
busy time
• Understanding variability in service quality characteristics is
important to the control and improvement of service quality
• In addition to variation due to equipment, process and
environment, performance variation is also due to person to
person, project to project variation etc.
• Use of Control charts are used to monitor service processes
and to determine status of statistical control
58
BITS Pilani, Pilani Campus
598
Techniques for Evaluating Service
Quality – contd..
• Variable control chart are used to measure
quantifiable characteristics like time to serve a
customer in a restaurant, temperature maintained in
a plane etc.
• Attribute control chart are used to service
nonconformity characteristics like proportion of
billing errors, while ‘c’ and ‘u’ charts are applied to
service nonconformity, facility-related and behavioral
characteristics
• Sampling techniques in service operations
– 100% sampling
– Convenience sampling
– Judgment sampling
– Probability sampling
59
BITS Pilani, Pilani Campus
599
Techniques for Evaluating Service
Quality – contd..
• 100% sampling
– Used when the cost of external errors or nonconformities is high
– Cost of sampling and inspection is high, but better than a
nonconformity found by a customer
• Convenience sampling
– Samples are chosen by the ease of drawing them and are
influenced by the subjectivity of the individual
– Eg. Selecting a thin file or selecting the top file during audit
• Judgment sampling
– Samples are chosen based on expert opinion
– Can also create bias and hence should be cautious in making
statistical inferences
• Probability sampling
– Has a statistical basis and is the most preferable
– Each item has an equal chance of being selected and done using
random number tables 60
BITS Pilani, Pilani Campus
600
A model for service quality
• Key concepts
– Customer satisfaction is a function of the perceived quality
of service, which is a measure of how actual quality
compares to expected quality.
– External and Internal factors affect customer perceptions of
quality
– External factors
• Are not directly under the control of a service organization
• Includes the social values and lifestyles of customers,
knowledge of service, and the services offered by the
competitors
– Internal factors
• Are directly under the control of service organization
• Includes the image management, client management and
service delivery system
61
BITS Pilani, Pilani Campus
601
A model for service quality – contd..
62
BITS Pilani, Pilani Campus
602
A model for service quality – contd..
• Components of service
– Core service and peripheral service and the benefits are
both tangible and intangible
– Eg. For a travel agency providing bus service
• Core service is transport of people between cities
• Peripheral service are comfort and safety of the bus stations,
meal services, pick up and drop etc.
– Peripheral service play a major role in influencing the
customer to select one company over another
• Factors involved in service quality
– Timeliness of the service, Adequacy of the service, price of
the service
– Evidence of good service spread through mouth
– Delivery of service must change with customer expectations
63
BITS Pilani, Pilani Campus
603
Applications
• Discussion is about application of quality control and
improvement techniques in the service sector
• Different applications discussed are
– Administrative operations
– Banking
– Education
– Food industry
– Federal, state and local government
– Health care
– Insurance
– Public utilities
– Transportation
64
BITS Pilani, Pilani Campus
604
Applications – contd..
• Education
– Amount of knowledge retained is a function of the individual
receiving the service and the service delivery process
– Controlling quality in education is unique because it involves
transformation of human qualities and values into learning
and retaining knowledge
– Behavioral characteristics
• A major influence on quality is the competency and behavioral
characteristics of the teachers
• Knowledge of their subject is a must, but is not enough and
delivery process is crucial
• Enthusiasm and the ability to stimulate interest and curiosity in
students are traits of a good teacher
• There are not easily quantified, rather they are subjective
• Other measures of teacher quality can be technical background,
certification and recognition by other professional associations,
research and publication in the area of expertise etc.
65
BITS Pilani, Pilani Campus
605
Applications – contd..
• Education – contd..
– Facility related characteristics
• Science and technology requires current laboratory facility and
equipment
• An adequate library with holding of books and periodicals
complements the delivery process
• Modern computer facilities are especially pertinent today
– Timeliness characteristics
• Proper scheduling of classes, tests, assignments etc.
• Lead time for release of grade sheets, degree certificates
• Proper completion of syllabus, and returning of graded answer
books etc.
– Service nonconformity characteristics
• Proportion of students passing the board exams or in a class, No.
of complaints about the library, or lab facilities, No. of billing
errors in mess and tuition fee bills 66
BITS Pilani, Pilani Campus
606
Applications – contd..
• Education – contd..
– The model
• Customers in educational model has conflicting objectives
• Parents want quality education at lower costs
• Employers want graduates exposed to cutting-edge technology
• Board of trustees want to minimize the operational expenditure
of the university
• Others include tax payers, alumni, etc.
• Students are of course customers, but also acts as co-producers
• They play a significant role in determining the quality of end
product through the degree and intensity of their effort and
motivation
• Students are considered as “work-in-process” inventory
• Finished product is the graduate, while the vendor is the high
schools and junior colleges
• Barriers in the system are due to communication between
departments, schools and colleges and flow of information is
curtailed, which prevents the improvement of quality of service
67
BITS Pilani, Pilani Campus
607
Applications – contd..
• Education – contd..
– Role of quality
• In education, it is possible to achieve certain degree of
control can be achieved over the customers – students by
selectively choosing the students based on GPA/marks,
scores on standardized tests, Interview etc.
• This directly affects the degree of control on quality of
output
• Characteristics that measure the quality of output have
tangible and intangible components
• Tangible measures may be proportion of graduates
receiving job offers or average salary of graduates with job
offers etc.
• Intangible measure may be satisfaction and fulfillment that
comes from obtaining a quality education – not easily
quantifiable
• Potential of quality control and improvement methods are
pretty complex
68
BITS Pilani, Pilani Campus
608
Applications – contd..
• Education – contd..
– Quality tools
• Evaluating the teaching performance by students – does it really
measure the ability and devotion of the teacher?
• Response of student may be motivated by his grade in the class!!
• Solution is carefully planned evaluation form – questionnaire
• Control chart for average score and range of the score can
indicate trends or special causes
• Control charts can also be used to evaluate facility and service
delivery characteristics
• Eg. No. of students unable to register for a certain course – c chart
helps in identifying a extra section for that course
• Methods of improvement could be handled by Pareto analysis and
cause-and-effect diagram
69
BITS Pilani, Pilani Campus
609
Applications – contd..
• Administrative operations
– Involves a high volume of paperwork and hence opportunity for error
is high
– To improve the efficiency, a through analysis of the flow of paperwork
is necessary and should eliminate redundant steps
– Mistakes can also be reduced by redesign of form – where data are
entered, Eg., employee filling travel expense form
– Factors that affect are: Form design, Information required, sequence
of actions for processing the desired service function and the
individuals involved
– Form design should be simple and layout should be such that
individuals processing the information can conduct their operations
sequentially and also flipping of pages to be removed
– Personnel training influences efficiency and error rate
– Stability of this service functions can be checked through control
charts – ‘p’ charts and ‘c’ charts can be used
– Eg. No. of errors per 100 purchase orders, Proportion of bills that
have errors in mail order company
– Tools for improvement are Pareto analysis, cause and effect analysis,
experimental design
70
BITS Pilani, Pilani Campus
610
Applications – contd..
• Banking
– Involves a high volume of activity and necessary to apply quality
control and improvement procedures
– Automation helped in performing tasks quickly and timely manner,
but risk of errors are high
– Behavioral characteristics like courtesy, communication have a major
impact on customer satisfaction
– Timeliness characteristics like waiting time, transaction processing
time also of concern to customer and processing time varies based on
nature of transaction
– Perfect accuracy is expected as far as service nonconformity
characteristics and hence 10% of bank employee work on detection
and correction of errors
– Attribute sampling plan and process control techniques play a major
role
– One important activity is use of MICR which has impact on the quality
of checks
– The requirements of these magnetic characters and the process
through which they are created are very stringent
71
BITS Pilani, Pilani Campus
611
Applications – contd..
• Banking – contd..
– For example, the characters has to lie within certain regions of the
check area, the ink should be uniformly distributed, properties of ink
etc plays a major role
– If the reader cannot sense the characteristics, the check need to be
processed manually thereby increasing the processing cost
– Hence incoming quality of printed checks and encoding process by
which the check amount is entered requires careful control
– Sampling plans such as ANSI/ASQC Z1.4 can be used to measure the
quality of documents like checks, forms etc.
– Example
• Say MICR Checks are printed in a batch of 30000 each
• Table 10-8, sample size code is K assuming inspection level I
• Assume normal inspection and AQL of 0.4%, Table 10-9 gives sample
size is 125, with a rejection number of 2
– Process control methodology can be used for check processing and
other clerical operations – use of ‘p’ chart and ‘c’ chart
– A measure of process capability is obtained from the center line and
the control limits of the attribute charts
72
BITS Pilani, Pilani Campus
612
Applications – contd..
• Food industry
– Several quality characteristics differentiate unit of food products and
impact product acceptability
– More clearly an attribute is defined and measured, the easier it is to
judge the quality of the food
– Some quality characteristics are readily measurable – proportion of
fat in milk, % of foreign material in a cereal, etc.
– Other characteristics are not as easily quantifiable – smell of a ripe
fruit, taste of a particular brand of strawberry etc.
– Requirements for the level of quality are set by customer –
wholesaler, distributor etc.
– Specifications and requirements are also influenced by federal
agencies like Food and Drug Administration, FPI, BIS etc.
– Total quality systems approach is required in which the level of
quality is monitored at each step from raw materials to processing to
packing and delivery
– For perishable foods - packaging, delivery and storage of the food
items are important
– Variety of quality attributes exists in food industry
73
BITS Pilani, Pilani Campus
613
Applications – contd..
• Food industry – contd..
– A consumer may combine specific attributes to arrive at a single
measure of quality, generally referred to as sensory quality
– Sensory quality includes attributes as appearance, flavor and
kinesthetic associated with the product
– Appearance may be color, size, shape while flavor refers to smell and
taste. Kinesthetic or muscle sense includes attribute as texture and
viscosity
– Amount of force required to initiate a saliva flow is a measure of
these two attributes
– Requirements for the level of quality are set by customer –
wholesaler, distributor etc.
– These characteristics are not usually under strict government control
and depends on customer preference
– Another class of quality characteristic are quantitative in nature and
deals with amount of particular component in the product, Eg.
Proportion of protein, fat etc.
– These characteristics are easy to measure and control chart for mean
and range can be used
74
BITS Pilani, Pilani Campus
614
Applications – contd..
• Food industry – contd..
– Certain quality characteristics cannot be readily judged by senses of
the consumer as they cant detect the presence of toxic items
– This is where the government plays an important role
– Tools that can be used are:
– Xbar and R charts for % of water, protein etc, while c-chart for the
micro organism or bacteria count while p chart can be used in
poultry and egg to monitor the egg shell shape and texture
• Federal, State and Local government
– Provides variety of services like recreational needs, housing,
environmental quality, small business, safety etc.
– Some government agencies provide service directly to consumer and
others are regulatory that deal with upkeep of desired quality levels
for the protection of consumer
– They generate numerous statistics and use frequency or relative
frequency histograms
– Eg. The income distribution families receiving food subsidies from
the government is obtained from relative frequency histogram
– Trend charts can be used to find the total amount allocated by the
government in budget preparation
75
BITS Pilani, Pilani Campus
615
Applications – contd..
• Federal, State and Local government – contd..
– Opportunities to use control charts and sampling
techniques are wide ranging
– Eg. Income tax: to determine whether tax returns to be
audited or the ratio of adjusted gross income to taxable
income can be monitored using p-chart
– In case of agriculture department, proportion of insect-
infested cocoa beans and the proportion of moldy wheat
can be monitored through p-chart in addition to know
about no. of errors in misclassification of a family,
miscounting etc can be monitored through c-chart
– For a environmental protection agency, if certain
standards are mandated by law, the degree of
conformance to those standards can be done using
process capability analysis
– Level of pesticide in the mineral water can be monitored
with Xbar and R charts
76
BITS Pilani, Pilani Campus
616
Applications – contd..
• Health care
– Provision of healthcare services is a national concern
– Consumer also influences the level of quality and also other groups
like Health department is highly influential in setting guidelines for
quality health care
– Unusual circumstances exist in health care. Patients are not always in
a position to judge the technical competency of the service provided
– Neither the administrators also, but only peer reviews by physician
can judge the quality
– In other services, consumer pays for the service rendered, while in
health care its not the case
– Federal government assists certain segments of the population
through Medicare / Medicaid programs
– Program guidelines which limit the maximum amount that may be
paid affect the quality and profitability of services
– Healthcare executives set service-level policies based on providing
adequate service at a reasonable price and plan their execution
77
BITS Pilani, Pilani Campus
617
Applications – contd..
• Health care – contd..
– Another unusual feature is the organizational structure of hospital
administration
– Governing board and administrative officers are separate from the
medical staff
– In private hospitals, board of trustees may be the owners or
representative of the owners and not physician, while in pubic
hospitals, board members are selected from local citizenry, but may
have professional expertise, contributing to effective management of
hospital
– Executive policies set by board are carried out hospital
administration
– Physicians too have authority and they mostly serve as consultants,
who have contracts to provide specific service like surgery
– These two administrative powers must act in harmony for total
quality system, but their objectives conflict
– Best health care is not always profitable as government has used cost
cutting measures such as ‘diagnosis related group’
78
BITS Pilani, Pilani Campus
618
Applications – contd..
• Health care – contd..
– In this system, a patient is classified into a DRG based on the
diagnosis and other characteristics like age, length of stay in the
hospital, surgical procedures etc.
– DRG then defines the maximum amount for which hospitals will be
reimbursed by the federal government, which has direct impact on
the quality of service
– Third party insurers also have an indirect impact on the quality of
service, as patients can go to hospitals approved by the insurer
– Depending on plans, insurers also control patient’s choice of
physicians
– To summarize, hospitals are under extreme pressure to cut costs and
yet give high quality service
– Hence hospitals must come up with effective quality control and
improvement measures for both service and cost
– An error free environment is the objective
– Patients demand better levels of service – clean rooms, home like
atmosphere etc.
79
BITS Pilani, Pilani Campus
619
Applications – contd..
◼ Health care – Quality characteristics and methods of control
◼ Pareto chart and cause and effect diagram is used for quality
improvement techniques reducing waiting time, admit time etc.
• Blood pressure, pulse rate etc. • Run chart
• Time to obtain an appointment
• Response time of ambulances
• Admit time in emergency
• Xbar and R charts
room services
• No. of adverse comments per
week on nurse’s performance
• No. of errors in blood tests per
100 samples • ‘c’ chart
• Proportion of cases with
inaccurate diagnosis
• Proportion of tests performed
incorrectly
• ‘p’ chart
80
BITS Pilani, Pilani Campus
620
Applications – contd..
• Insurance
– Consumer needs insurance for protection from accidents, hazards,
and casualties
– Personal insurance:- insurance for home, automobile, life (medical
insurance) while Group insurance:- Medical insurance to employees
of an organization
– Organizations also purchase insurance from other agencies to
provide their customers a sense of security
– In these cases, service nonconformity quality characteristics are
critical and the goal is to have no errors
– It involves people and hence training and motivation are critical
– Important functions
• Data collection
• Inputting the data for calculation of casualty probabilities
• Estimating payoffs
• Underwriting policies
• Processing claims etc. which requires human operations and
interactions
81
BITS Pilani, Pilani Campus
621
Applications – contd..
• Insurance – contd..
– Timeliness characteristics like correcting a complaint or error,
processing a claim or obtaining damage appraisals from
adjusters should be performed within time limit
– Objective of insurance seller
• Increasing profits is the primary concern here
• Pareto analysis helps to identify vital characteristics of improving
profit
• Cause and effect diagram helps to identify the focus areas
• Cutting operations costs and avoid errors through well
documented procedures (Documentation
• Sellers are interested in reducing risks to their own company by
specifying limitations and exclusions in policies
• But this conflicts with the buyers objective of obtaining
protection against all types of risk
– Statistical techniques are widely used: Premiums are set based
on historical data, sampling methods are used to obtain data
– They are used to estimate and analyze the probability of fire,
accident, amount of loss etc.
82
BITS Pilani, Pilani Campus
622
Applications – contd..
• Insurance – contd..
– Control of sampling and non-sampling errors are vital to the
accuracy of the estimation process
– Generally random sampling is used along with stratified
sampling and cluster sampling techniques
– Non sampling errors include incorrect data transcription,
inaccurate data input to models used for estimation purposes,
use of inappropriate or outdated models for estimation, faulty
computer programming etc.
– Several control charts can also be used:
• P-chart monitor the proportion of unacceptable documents,
proportion of claims not processed or paid etc.
• C- chart can be used to monitor no. of incorrect entries per 100
documents and no. of underwriters per 500 customer accounts
• Xbar and R chart could apply to characteristics as processing time
for claims and the amount paid by the company for policies in
similar category etc.
– Run chart can be used to track the total amount paid on a
monthly basis
83
BITS Pilani, Pilani Campus
623
Applications – contd..
• Personal services
– Providers deal with individual customers on a frequent basis
– Eg. Beauty shops, hotels and motels, laundry, travel agency etc.
– Providers rely heavily on repeat business, and hence should know the
characteristics and attributes that satisfy the customers as well those
that make them hate the service
– Individuals taste and preference vary and hence businesses try to
incorporate flexibility in the design of services
– Eg. Hotel industry, some customer prefer indoor swimming pool,
others may be interested in snack bar while some would like
washbasin to be outside while some inside
– There are 2 groups of factors – those cause customer satisfaction and
those cause customer dissatisfaction
– Dissatisfied customers do not come back and also spread the word
– Reputation is everything and hence customer feedback is important,
which should be obtained by providing incentives
– Facility related quality characteristics that cause service
nonconformities fall in the customer dissatisfaction category
– Eg. Dirty carpet in a hotel room, inadequate temp. control in a saloon
etc.
84
BITS Pilani, Pilani Campus
624
Applications – contd..
• Personal services – contd..
– Eliminating causes of customer dissatisfaction should be 1st step but it is
not enough and other factors that influence the perceived quality level to
exceed expectations is most important
– Eg. A hair dresser providing a head massage after his basic service, while
a beauty parlor providing a free mehendi design
– This added touch may lead the customers feel that they have obtained
more for their money’s worth
– Such customers act as a best source of advertisement for business and
spread the word making way for repeat business as well new customers
– Factors that cause customer satisfaction fall in human factors and
behavioral or timeliness category. Eg. Courteous behavior of the front
desk attendant in a hotel or a minimal waiting time to be seated in a
restaurant etc.
– Xbar and R chart can be used to control – time to register and check in at
a hotel, time for the suit to be dry cleaned etc.
– P-chart can be used for proportion of rooms available for occupancy or
proportion of on-time delivery of pizza etc.
– Demerit charts may be used to control the weighted quality points from
customer report cards
85
BITS Pilani, Pilani Campus
625
Applications – contd..
• Public utilities
– include electric power gas and telephone etc.
– Characteristics
• Unique feature is that companies providing these facilities are
nearly monopolies or oligopolies.
• Are regulated by public service commissions that set the rate
structure for the service and not the customer directly
• Indirectly, though, consumer opinions have an impact on the
negotiated rate structure approved by the public service
commission.
• The standard for quality is imposed by the regulatory agencies.
• Customer expects continuous, uninterrupted service
• Utilities have a guaranteed customer base and hence operating
scenario is different from other service industries
86
BITS Pilani, Pilani Campus
626
Applications – contd..
• Public utilities – contd..
– Recently competition in utility services is the talk of the day.
– Another feature of utilities is that consumer controls the
consumption and generates instantaneous demand, thereby lead
time for demand is zero and hence rely on forecasting
– Providing adequate service under these circumstances is both
science and art. The art is forecasting. (the predictive function).
– Mathematical models are developed to forecast demands.
– There is an art in deciding what model components to consider,
what measures of demand to use. What information from historical
demand to use, and how to weight them.
– These services cannot be stored. It is not feasible to over produce
and store power to meet peak demand at a later point in time while
gas can be stored
87
BITS Pilani, Pilani Campus
627
Applications – contd..
• Public utilities – contd..
– Service nonconformity and facility related characteristics are
critical to electric power and gas industry
– All 4 categories are important for telephone industry
– Xbar and R chart can be used for variables such as the amount
purchased from outside sources, time to restore service after
failure, waiting time to get telephone installed etc.
– Trend chart track electrical power or gas consumption on a
monthly basis for a single customer or group of customer and help
identify patterns in consumption thereby helps in developing the
model for forecasting
– No. of errors in meter readings per 100 customers, No. of customer
complaints per month, No. of human errors per month in a power
plant can be monitored by c-chart
88
BITS Pilani, Pilani Campus
628
Applications – contd..
• Transportation
– All 4 categories of quality characteristics play a major role
– Level of quality is largely influenced by customer and some
government agencies also have regulatory powers
– Eg, Civil Aeronautics Board in US receives customer complaints
regarding air travel and summarizes the findings to the airline
companies
– Airline companies can improve based on the remedial action for
the above complaints, but customers don’t write to CAB, rather
they switch airlines
– Other methods like rating forms should be used to find out the level
of customer satisfaction and expectation
– The response rate may be low and can be increased through free
gifts
– Important quality characteristics is safety of operations and the
fatal and nonfatal accidents should always have a goal of zero
– Prevention of accidents through equipment maintenance, checking
and training of personnel is a priority
89
BITS Pilani, Pilani Campus
629
Applications – contd..
• Transportation – contd..
– Other quality characteristics deal with operation and delivery of
service
– Behavioral characteristics include courtesy of check-in agent, flight
attendants or the cab driver of taxi etc.
– No. of complaints received per month can be monitored by c-chart
– Timeliness characteristics include waiting time at the check-in
counter, waiting time to retrieve baggage, flight delay in arriving at
a destination etc.
– Xbar and R chart may be used to monitor the above measures
– Service nonconformity characteristics include no. of damaged or
lost bags per month, no. of flight cancellation etc which can be
monitored through np-chart
– No. of mistake in reservation per month, no. of airport security
problems per month are monitored through c-chart
90
BITS Pilani, Pilani Campus
630
Thank You
91
BITS Pilani, Pilani Campus
631
BITS Pilani
Pilani Campus
Process Capability
2
633
Process Capability
Process capability represents the performance of a process in a
state of statistical control. It is determined by the total
variability that exists because of all common causes present in
the system.
Nominal
value
Process distribution
Lower Upper
specification specification
20 25 30
Process is capable
Nominal
value
Process distribution
Lower Upper
specification specification
20 25 30
=x – Lower specification
Upper specification – x=
Cpk = Minimum of ,
3 3
We take the minimum of the two ratios because it gives the worst-
case situation.
= =
Cpk = Minimum of
x – Lower specification ,
Upper specification – x
3 3
Process
Cpk = Minimum of 1.53, 0.94 = 0.94 Capability
Index
23
BITS Pilani, Pilani Campus
654
Example
In an assembly operation in a semiconductor manufacturing
company, the lower and upper specification limits are given by
4.8 and 5.2 seconds. A random sample of 25 completion times
gave a mean and standard deviation of 5.12 and 0.06 seconds,
respectively. Can we conclude that the Cp index for this operation
exceeds 1, so as to be considered acceptable by the customer?
Test at a significance level of 0.05.
24
BITS Pilani, Pilani Campus
655
26
BITS Pilani, Pilani Campus
657
Thank You
36
BITS Pilani, Pilani Campus
667
2
BITS Pilani, Pilani Campus
669
Numericals
The amount of a preservative added to dairy products should not exceed certain
levels of 23 ± 6 mg (set by the Food and Drug Administration). Samples of size
5 of processed cheese produced the values of the average and range shown in
Table.
(a) Construct appropriate control charts and determine stability of the process.
(b) If the process is out of control, assuming remedial actions will be taken, estimate
the process mean and standard deviation.
(c) Assuming normality and a target value of 23 mg, determine the indices Cp, Cpk,
Cpm and Cpmk.
(d) What proportion of the dairy products meets government standards, assuming
normality?
(e) Find a 95% confidence interval for Cpk, assuming normality.
(f) Can we conclude that Cpk is less than 1? Use a = 0.05.
Acceptance Sampling
5
BITS Pilani, Pilani Campus
672
Definition
r !( n − r )!
r =c
n!
Pa ( p ) = P ( r = 0 ) + P ( r = 1) + P ( r = 2 ) + + P (r = c) = p r (1 − p )
n−r
r = 0 r !( n − r ) !
Pa ( p )
This tells that once n and c are Ideal OC Curve
we assume that the sample is chosen from an isolated lot of finite size.
Construct an OC curve for a single sampling plan where the lot size is 2000,
the sample size is 50, and the acceptance number is 2.
The average outgoing quality (AOQ) is the average quality level of a series
of batches that leave the inspection station, assuming rectifying inspection,
after coming in for inspection at a certain quality level p.
Taking N as the lot size, n as the sample size, p as the incoming lot quality,
and Pa as the probability of accepting the lot using the given sampling plan,
the average, outgoing quality is given by
The average outgoing quality limit (AOQL) is the maximum value, or peak, of the
AOQ curve. It represents the worst average quality that would leave the inspection
station, assuming rectification, regardless of the incoming lot quality.
If rectifying inspection is conducted for lots rejected by the sampling plan, another
evaluation measure is the average total inspection (ATI). The ATI represents the
average number of items inspected per lot.
For single sampling plans, the average total inspection per lot for lots with an
incoming quality level p is given by
Construct the ATI curve for the sampling plan where N = 2000, n = 50, c = 2.
Consider the calculations for a given value of the lot quality p of 0.02. The probability of
accepting such a lot using the sampling plan is Pa = 0.920. The ATI for this value of p is
Find a single sampling plan that satisfies a producer's risk of 5% for lots that
are 1.5% nonconforming.
Find a single sampling plan that will satisfy a consumer's risk of 10% for
lots that are 8% nonconforming.
Find a single sampling plan that satisfies a producer's risk of 5% for lots
that are 1.8% nonconforming, and a consumer's risk of 10% for lots that are 9%
nonconforming.
Thank You
33
BITS Pilani, Pilani Campus