Quality Management
Concept of Quality
The word Quality has different meaning, concept and perception to different people.
Some of them are
- Fitness to use
- Compliance with specified requirement
- Free from defects, imperfection and contamination
- Degree of excellence
- Level of Customer satisfaction by way of value for the money paid etc
Dimensions of Quality
5. Durability useful life of the product eg. Washing machine to last trouble free
for min 5 years or till technology changes eg. CRT to LCD/LED TVs
6. Aesthetics appearance, feel, smell, taste of products and packing eg. Colour
and shape of refrigerator
7. Perceived quality Quality associated with reputation and brand name eg.
TATA products or V Guard are expected have good product quality
1. Tangibles
Physical appearance of sales and service office, its facilities and its employees, their
communication capability. Eg. Airline ticketing office
2. Convenience
Accessibility and availability of services nearer to the customer eg. Nokia mobile
phone service centres
3. Assurance
Competence and ability to convey trust and confidence of service personnel eg. A
Samsung service technician is capable to service all Samsung household products
4. Reliability
Dependency of service providers for delivery of service eg Courier Service co,
logistic companies, Authorized service centres of LG
5. Time
Speed at which service is provided- quick response time to a service request. 24 Hr
on the road mobile servicing units of Hyundai Motors
6. Responsiveness
Willingness to help customers in unusual circumstances to solve their problems eg.
insurance companies settling complex claims quickly, air lines providing ticket for
emergency travel
7. Courtesy
The way customers are treated by employees of service providers eg. cabin crew of
air line, single window service in banks
Definition:
Quality is defined as the Totality of characteristics of an entity that bear on its ability
to satisfy the stated and implied needs.
Characteristics
means features of product or services as stipulated in product catalogue or service
contract
Entity
- refer to a particular product or service
Stated need
- needed to be satisfied as per catalogue or contract
Implied needs
- needs expected to be satisfied e.g. Washing soap
Cost of Quality
Failure cost
Cost of appraisal
& prevention
Quality of conformance - %
3. Appraisal cost
Cost to determine the degree of compliance to quality
o Incoming inspection cost for purchased materials and parts
o In process inspection cost for manufactured and assembled parts
o Documentation cost paper work and record keeping of inspection reports
o Equipment cost of inspection and testing equipments, lab etc
o Evaluation cost evaluation of stocked materials and products for determine
deterioration.
4. Prevention Cost
Cost of activities to prevent failure and to reduce appraisal cost
o Quality planning cost of preparing quality procedures and systems
o New product review cost of reliability engineering and product review
o Process capability planning and review to ensure that the manufacturing
processes and systems gives the required quality.
o Quality Audit cost of auditing existing systems
o Supplier quality evaluation evaluate and improve supplier quality during
contract period.
o Training cost for training of employees on quality
5. Hidden Cost
Cost that cannot be quantified or measured accurately
o Loss of revenue and profit from existing and new customers
o Lower market share market being taken over by competitors
o Loss of corporate image affecting credit and investment flow
o Unreported failure cost due to fear of punishment
o Poor employee morale having to a work again and again
o Life cycle cost higher cost of stocking & supplying spares and services in the
entire product life
Quality Assurance
Standards
Process Quality &
control Monitoring Design
documents
Maintenance
Materials /labor
Cost of Q.A.
Cost of machines
Low process High process
Capability capability
(High rejection) (Low rejection)
- The two costs associated with Operations are Cost of QA and Cost of Machines
(Process).
- With increasing cost of a machine (ie Cost of Process Choice),the cost of QA
decreases and cost of production increases.
- The sum of these costs is the Total cost which takes a U shape.
- Hence the machine to be purchased (Investment in Process Choice) should be of
such Process Capability (quality of output) such that the total cost is minimum.
Eg. An ordinary drilling machine that cost Rs 1 lakh only is needed for drilling open
dimension holes where as high quality drilling machine costing Rs 4 lakh is needed for
drilling precision holes.
2. Quality Management System essentially means that quality is built into the
system of various Management processes, Production processes and
Procurement processes so as to ensure the required quality to the end product is
maintained and improved on a continuous basis.
3. ISO 9000 QMS is different from inspection or quality control because
inspection is a post production activity.
1. System Management
2. System Methodology
3. System Maintenance
5. Contract review compliance as per order placed by all customers w.r.t. time,
delivery, quality, performance
6. Design control to ensure that products are produced as per the latest design which
are modified from time to time to meet changing customer needs.
7. Document and data control maintenance of all data and records for traceability of
the reason for any defect which can appear at a later date while in use.
10. Product identification and traceability documentation for tracing the source and
reasons for any defect found at a later date in the product
11. Process control ensure production machines are capable of giving the required
quality
12. Inspection and testing clearly defined inspection and testing methods and
procedures
13. Control of non-confirming product because of very high cost, certain parts are
not rejected due to deviation in certain dimensions, but other parts are modified to give
the desired performance this is a non-confirming product documentation and
performance monitoring and control of such products
14. Handling, storage, packaging and delivery a documented system for all these
activities
15. Training documented requirements of employee training and training imparted
1. ISO 9000
These standards provide guidelines for selection of ISO 9000 family of standards (ISO
9001, 9002 or 9003) as applicable to the particular business.
2. ISO 9001
This Standard provides Quality standards as applicable to the companies and
businesses engaged in Design, Development, Production, Installation (including
testing and commissioning) and Servicing. Almost all engineering industries come
under ISO 9001 series. Eg. BHEL, Tata Motors, Maruti, HMT
3 ISO 9002
This document gives Quality standards as applicable to the companies and business
engaged only in production, installation/supply and servicing. (No product design
and development is undertaken by the company)
Most of the process based industries in petroleum, fertilizer etc where the company by
itself do not carry out any product design and product development work, come under
this series.
E.g. Urea produced by FACT and other companies are same in quality and
composition. They do not carryout designing of Urea. ISO 9002 is applicable to
FACT Urea Plant
4 ISO 9003
These standards relate to final inspection and testing system and procedures for both
in-house production as well as purchased parts and materials.
5 ISO 9004
These standards relate to internal Quality Assurance and Quality Audit.
2. Appoint Consultant
- Consultant will help preparation of Quality Manual and other ISO 9000
documents.
- He will provide overall guidance for implementation
- Identify executives and fix responsibility for each of the 20 elements of ISO
9000.
- These Element Owners will ensure ISO implementation and maintenance as per
the Company Quality Manual.
8. Review present system
9. Prepare Documentation
The following ISO 9000 Quality Documents are prepared by MR, Implementation
Team, Consultant, Element Owners and Departmental Executives
- Install new system in all departments with effect from certain date
11. Internal Audit
- Company has to register with the ISO 9000 Certifying Agency for certification
by paying necessary fees.
- Every six months, Surveillance Audit is done by the certifying body auditor.
- If the system is working OK, and there are no major Non- Conformities,
company continues to remain as ISO 9000 certified.
Advantages of ISO 9000
Following are the major advantages of implementing ISO 9000 system
1. Quality is built into the system, i.e. the quality is assured if the systems are
correctly followed.
7. Increases profit.
8. Bring about attitudinal change in the employees, resulting in their participation
and better performance.
9. Each employee feels his importance and role in the operations of the company.
- ISO 9000 is a top down management activity. Top management and all
executives should be fully committed to the system to make it a success. Any
laxity on their part will make the system ineffective.
2. Lack of resources
- This is mainly because of the human nature to resist changes. This needs careful
handling by bringing in high awareness in each and every employee at all
levels.
4. Capital Investment
- Capital Investments are essential for upgrading inspection and testing facilities
and some production facilities to meet the quality requirements. With out capital
investment, ISO 9000 cannot be maintained in long run
Components of TQM
Leadership and
Management Total
Commitment Customer
Satisfaction
Physical
Human Resources,
Resources & Systems and
Team work Techniques
TQM Model
Fig. 15.4 - 1
1. Continuous Improvement
3. Employee empowerment
4. Team Approach
5. Fact based decisions
6. Knowledge of quality tools
7. Suppliers/Vendors Quality
8. Quality at the Source
1. Continuous Improvement
TQM aim at continuous improvement of materials, methods, machines and
people even if they are presently working well. The term Kaizen refer to
continuous improvements even in small steps, a technique perfected by Japan to
achieve global competitiveness.
3. Employee empowerment
Motivate and encourage innovation. Give workers the authority to make
changes in their operational area and make them responsible for the
improvement of performance.
4. Team Approach
Use synergy from team work for problem solving
6. Develop internal strength through its employees and systems to overcome bad
times and competition.
A number of tools are available to aid management to interpret and analyze data and
take decisions. Some of them are
1. Flow Charts
Okay Okay
Flaws Defective
- Very useful to analyze long processes with several activities/ procedures and
decision points to make the process more productive
2. Check Sheets
It gives a format to record certain data in a structured manner.
It is a simple tool used for problem identification and analysis
example
Rejection data of 4 parts with Batch Qty of 1000 each
Date Part A Part B Part C Part D
Rejections rejections rejections Rejs
01.01.10 3 5 2 -
02.01.10 1 2 2 2
03.01.10 4 2 1 -
04.02.10 1 5 - -
Etc
The above data immediately draw your attention to Part B which has more rejection
and need corrective actions
3. Histogram
Histogram is a graphical representation of distribution. It will provide clear visual idea
of distribution and range.
It will also highlight unusual values in a given data.
Eg. Histogram of Machine Breakdown causes A to F and their frequency of
occurrence.
F
r
e
q
u
e
n
c
y
A B C D E F
Machine Break down causes
The histogram shows, C is a major cause of break down and F is a minor cause
4. Pareto Analysis
Pareto analysis is the technique of focusing on the most important problem area and
resolve that problem on priority.
F
r
e
q
u
e
n
c
y
C D B E A F
Machine Break down causes
Popular 80-20 rule is based on Pareto concept, which says 80 % of problems is
contributed by 20% of the items and vice-versa.
Eg. 80 % of product defects come from 20% of causes.
A Pareto diagram is used for Pareto analysis.
Eg. Pareto diagram for machine breakdown . Major causes for machine break down
are 2 (C&D) out of total 5 causes. They are to be tackled immediately to reduce
break down time substantially.
5. Scatter Diagram
Scatter diagram show the degree and direction of relationship between two
variables.
E.g. The relation ship between % of rejection while machining a component and the
humidity of the room. Scatter diagram shows rejections are higher when humidity is
higher. Action is to reduce humidity to reduce rejection.
*
* *
% of *
rejection
*
*
* *
% Humidity
* UCL
* * * * *
Quality * * * * * Mean
* * * * *
Parameter * *
Eg. Hole Dia LCL
* *
Lot No
Lots where the value lies between UCL & LCL shows the process is under control.
Variations are random or natural deviations .
When the value goes beyond UCL and LCL, the process is not under control and the
variations are due to certain causes
7. Cause and Effect Diagram or Fish bone diagram
It is a method to search for possible causes for a problem or process deviations
It is a systematic problem solving tool.
Possible Sources and Causes of problems are explored and examined through group
discussions by all concerned.
Detailed analysis is made by asking who, what, where, when, why and how to find
the source and root cause of the problem.
Eg. Poor Surface finish (Effect) in a Grinding operation
- It is not only applicable in production but also in other functional areas like
finance, marketing, sales and service, HR, purchase etc.
- i.e. 100 % inspection is not possible, it is time consuming, expensive and non-
productive
- Hence Statistical Quality Control (SQC) is used to ensure that the products
and parts have the required quality.
F
r Normal
e Distribution
q
u
e
n
c
y Mean + 3
Diameter
The measurements will have a Mean Value, Standard deviation and Range as given
blow
Mean x x / n
Standard deviation = (x x ) 2
/n
Standards
Process Quality
&
control Monitoring
Design
drawings
Maintenance
- It means the out put of the process confirm to the intended design of the
process output (eg. a grinding machine gives x micron surface finish and y
micron concentricity).
- Since it is not possible to inspect 100 % items, the quality is assured statistically
from samples taken at intervals during the process
- Statistical Process Control is a tool to evaluate the output of the process and
ensure that it is statistically acceptable.
- Different Control Charts ( x bar, R, c, p chart etc) are used for process control
3. Acceptance Sampling
- It is used to inspect and accept parts and raw materials purchased from suppliers
- It is also used for inspection and acceptance of finished product from assembly
line (a post production process)
- OC (Operating Characteristics) Curve is used to accept Purchased Parts
- AOQ (Average Outgoing Quality) graphs are used for quality analysis and
control of finished products produced eg. Electric Fans going out of a company
Types of Variations
1. Random variations or Chance Variations
o Variations occur at random
o They are caused by complex minor causes
o do not result in significant deviation from the nominal value eg dia 20.01,
19.99, 20.02 etc
2. Assignable variations
- variations are relatively large e.g. dimensional deviations of a machined part
(dia 20.2, 20.8. 19.5 etc), chemical composition deviation of chemical product.
- Causes of variations can be assigned to one or more of the following causes
o Due to workmen
o Due to machine
o Due to materials
o Due to tools
o Due to environment like humidity, dust etc
o Or combination of any of the above
Process Variations are represented and controlled using different Control Charts
Control Charts
Types of control charts
1. Control chart for Variables: variables are based on measurements like
dimensions, temperature, rpm, surface finish, chemical composition etc (x
bar and R Chart)
2. Control charts for Attributes: attributes are based on Numbers. The
population is classified into two classes, right or wrong, good or bad,
without errors or with errors
e.g 92 nos of invoices without error and 8 with error in 100 invoices checked
(p chart)
e.g. 32 number Mobile Phones with total 5 or more defects vs 9968 number
with less then 5 total defects. (c chart) in a population of 10000
If the cycle time for a process is more (eg. 5 to 10 mts), then one sample is taken for
inspection in every 30 to 60 mts in a shift. Control Charts are prepared based on this
inspection data
1. x Chart (x bar chart)
Standard deviation = (x x ) i
2
/n
* (Assignable cause)
Qu
UCL
ali * *
ty * * *
Pa * * x Mean
* *
ra * * *
m * *
et * * LCL
er * (Assignable Cause)
123 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Sample Number
Interpretation of Control Chart
The variations in dimensions of samples which fall with in the UCL and LCL are due to
random variations and are acceptable
The variations in dimensions of samples which fall outside the UCL and LCL due to
assignable variations reasons for which are to be investigated and corrected to
bring the process under control ie. within acceptable limits .
A new Process Control Chart is drawn and this represents the process.
As the process continues for the next shift and subsequent days, samples are taken at
certain intervals.
If the values are within New UCL & LCL, we can cay that the process is under control
with 99.75 % confidence level (corresponding to +-3 sigma level)
Where
A2 is constant for the sample size n at a certain confidence level ( value of
constants are available in A2 table for different sample sizes and confidence levels
refer any OM Book)
* (Assignable cause)
Qu
UCL
ali * *
ty * * *
Pa * * * x Mean
* *
ra * * *
m * *
et * * LCL
er * (Assignable Cause)
123 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Lot Number
The variations of Lot numbers falling within the UCL and LCL are Chance variations
which are acceptable and the process is under control
Variations of the lot numbers falling out side the UCL and LCL are Assignable
Variations the reasons for which are to be investigated to put the process under
control.
* (Assignable cause)
R
UCL
a
n
g R Mean
e
LCL
* (Assignable Cause)
123 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Lot Number
E.g.
A quality inspector took 5 lots at 45 mts interval from a continuous glue
manufacturing process. Each time he took 4 samples and recorded Glue setting time
as below Lot No
Sample 1 2 3 4 5
1 12.11 12.15 12.09 12.12 12.09
2 12.10 12.12 12.09 12.10 12.14
3 12.11 12.10 12.11 12.08 12.13
4 12.08 12.11 12.15 12.10 12.12
Also given the following factors at 3 (taken from standard tables for A2, D3 and D4
Sample size n A2 D3 D4
3 1.02 0 2.57
4 0.73 0 2.28
5 0.58 0 2.11
6 0.48 0 2.00
x = x / K = 12.11
x = Sample mean
R = sample range = Difference between the maximum and minimum values is one lot
or Sample
Ranges = 0.03, 0.05, 0.06, 0.04, 0.05
R = Average of sample ranges
= R / K = 0.046
R Chart
D3 = 0, D4 = 2.28
Range Mean = R = 0.046
LCL = D3 R = 0
UCL = D4 R = 0.105
All values are with in UCL and LCL in both the charts
Hence the process is under control with 99.75 % confidence level.
p Chart
p chart is to monitor number of defectives generated in a process.
Formula for p chart
Sample size =n
Number of lots =K
Number of defectives in each lots = c1, c2, c3 etc
p = pi
K
p(1-p)
UCL = p + 3
n
p(1-p)
LCL = p - 3
n
Alternately
Mean (center line) = n p = ci
K
UCL = n p + 3 np(1-p)
UCL = n p - 3 np(1-p)
Example
In a telephone company, an inspector took 100 Telephone bills each month from the
bills send to customers in the past 20 months. The number of defects identified in each
sample is as follows.
Sample 1 2 3 4 5 6
No. of defectives 4 10 12 3 9 11
Samples 7 8 9 10 11 12 13
Defectives 10 22 13 10 8 12 9
Samples 14 15 16 17 18 19 20
Defectives 10 21 10 8 12 10 18
Construct a process control chart that will describe 99.74 % of variations in the
process when it is under control.
Is the billing process under control?
Answer
Fraction defectives p = 0.04, 0.10 etc
No of samples K = 20
Sample size n = 100
Centre line p = p =
2.2
= 0.11
K 20
p(1-p)
UCL = p + 3
n
0.11(1-0.11)
= 0.11 + 3 = 0.20
100
p(1-p)
LCL = p - 3
n
0.11(1-0.11)
= 0.11 - 3 = 0.02
100
UCL=0.20
p
0.11
LCL0.02
c Chart
When the goal is to control the total number of defects in one unit or piece, we use c
chart to control the process
Eg. Number of defects in a hotel room. In a hotel room defects can be dust on TV, tap
leak, door very tight etc.
The goal set by the management is that the total number of defects in any room should
not be more than 3.
If they have say 100 rooms, it is not possible to inspect all the rooms.
Management can stipulate that House keeping process should be under control at 3
sigma level (99.74 % confidence level). An inspector can prepare c chart, UCL and
LCL and certify the house keeping is under control
Number of lots =K
Number of defects in each sample =c
c = c
K
UCL = c + 3 c
LCL = c - 3 c
UCL
c c
LCL
Lot number
Acceptance Sampling
Process Control Cart is used to ensure that the production process is under control
giving the required quality.
Where as Acceptance Sampling is for the parts that are purchased, or Finished
Products that go out for sale. Eg. 10000 bearings, 50000 Oil Seals
6. Need to inspect the parts more quickly to prevent inventory hold up during
inspection.
3. Also used for acceptance of group assemblies (e.g. gear box) in mass
production of automobiles
5. Widely used for material in the form of gas, liquid, stored in heaps, tanks or
moving in conveyor etc.
Eg. 100 nos are taken from a supply of 2000 bearings and fully inpsected
If > 2 nos are defectives in the sample of 100, the whole of 2000 bearing are rejected
Double sampling plan will specify two acceptance nos c1 and c2 for the initial
sample and single acceptance no c3 for combined samples
If the defectives are <= c1 , lot is accepted.
If the defectives are > c2, the lot is rejected
If the number of defectives falls between c1 and c2, a second sample is taken and
inspected.
If the defectives of combined sample is <= c3 , lot is accepted or else rejected.
Multiple Sampling
if the defectives are <= lower limit c21, the lot is accepted
defectives > upper limit c22, lot is rejected
If they fall in between c21 and c22, a 3rd sample is taken and inspected.
This process is continued when ever number of defects fall between upper and lower
limits and stopped when an accept or reject verdict is possible
Advantage of double and multiple sampling is that the sample size can be smaller
compared to single sampling resulting in saving in time and cost of inspection.
Multiple sampling is adopted for purchases made from reputed companies eg. SKF
Bearings, Intel Mother board, Intex keyboard etc
Important feature of any sampling plan is how to discriminate between good lot and
bad lot through sampling.
100 % inspection only can clearly discriminate between good and bad parts
In any sampling plan, some bad lots may get accepted and good lots may get rejected
Features
X Axis - Lot quality (fraction defective)
Y Axis - Probability of Accepting the lot
75%
50%
25%
LTPD
AQL
= 0.10
Go
o d Indifferent Bad
.03 .06 .09
Lot Quality (Fraction Defective)
- Each OC curve is for a specific Sampling Plan with specific Sample Size (n) and a
specific Acceptance Number (c)
- Any point on the OC curve will show the probability of accepting a lot with
certain fraction defectives in the whole lot.
Sampling Terms
In an OC curve, the AQL line meet the curve at certain point, which indicate the
probability of rejecting a good lot containing acceptable quality.
Consumers (buyers) risk is the probability that a lot containing defectives exceeding
the LTPD will be accepted.
It means that the end products of the buyer will contain some defective parts which
will be passed on to the end consumer with that probability
probability Reject
Accept
of acceptance
In 100 % inspection of the lot, the probability of accepting lots with defects =<0.01%
defect is 100 % and >0.01% is nil.
This is because the full lot with defect =< 0.01% is accepted, if more > 0.01%, the lot
is rejected
AOQ curve is used to ensure that the quality of out going finished product is with in
the Average Quality specified by the company before they are sent to distributors.
0.02
0.05 0.1 0.15 0.2 0.25
In coming Fraction defective
In statistical terms, Six Sigma quality means less than 3.4 defects per million
opportunities for defect in any process, products or services ( compared to 3 sigma,
ie, 2.4 defects in 1000 Nos)
F
r
e
q
u
e
n
Mean
c
y
-6 +6
Quality parameter
(dia of a shaft)
Six Sigma is essential for global companies, that manufacture goods for world
market. For them, 3 sigma limit is totally inadequate to meet customer satisfaction
and competitiveness.
Eg. HP produces 3 lap tops per minute in their US plant and annual production is
about 9 lakhs
At 3 sigma quality (99.76 % confidence level) in the final assembly alone, the number
of defective laptops per year is 2160.
If one takes cumulative defects from 3 sigma quality at component level, the
defective laptops will several thousands and it is not acceptable to HP.
Hence they go for Sigma quality level i.e only 3 defective lap tops per million
produced
- Champion Six Sigma Leader, who identify and rank potential projects, help
select and evaluate candidates, manage program resources, serve as advocate
for the program
- Master Black Belts They are specialists in Six Sigma Program having extensive
training in statistics and use of quality tools. They train Black belts
- Black Belts they are project team leaders responsible for implementation of
process improvement project after getting trained in six sigma program and
demonstrated mastery in the program.