Data of blowing air pressure & percent defective

of plastic tank 200 98.5 99.5 100 100

97.5
Date Air pressure Percent 95.5
93.5
91
(kgf/cm2) Defective 175 88
Oct-01 8.6 0.889 84
2 8.9 0.884 79
150 75
3 8.8 0.874 73
4 8.8 0.891 65.5
5 8.4 0.874 125
6 8.7 0.886 55.5

In nos

In %
7 9.2 0.911 100 50
8 8.6 0.912
9 9.2 0.895
75 66
10 8.7 0.896
33
11 8.4 0.894
45
12 8.2 0.864 50 25
13 9.2 0.922
14 8.7 0.909 25 20
15 12
15 9.4 0.905 10 8 6 5 4 4 2 2
16 8.7 0.892 1
0 0
17 8.5 0.877

Stores

Others

production
Quality
Personnel
Service
Materials

Finance
Marketing

Manufacturing
Development
Maintenance

Engineering

Information
Production

Research &
18 9.2 0.885

Systems

Factory

Planning
Plant
19 8.5 0.866
20 8.3 0.896
21 8.7 0.896
22 9.3 0.928
Dept
23 8.9 0.886
24 8.9 0.908
25 8.3 0.881
26 8.7 0.882
27 8.9 0.904

2 .Stratification 3. Pareto Diagram 4. Cause & Effect Diagram

1. Check Sheet
28
29
30
8.7
9.1
8.7
0.912
0.925
0.872

7 QC TOOLS + Additional Tools

C o r r e la t io n S t u d y f o r B o lt In s t a lla t io n

36

34

32

30
T o r g u e (F t. L b s .)

28

26

24

22

20
60 65 70 75 80 85
A ir P r e s s u r e (p s i)

5. Histogram 6. Scatter Diagram 7. Graphs & Control Charts

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
1
Why these tools selected for MYB?
 Systematic method to collect the data
 Stratifying the data in various ways
 Identify the vital few from trivial many (Prioritization to have
focused effort)
 Identifying the various causes of the problem, understanding
clear linkage between primary, secondary causes etc.
 Understanding the behavior of the process
 Understanding the relation between parameters (output
characteristics with process and input)

Identifying the real root cause of the problem

All these tools are simple and effective to solve problems. It is data
based approach and supports to take appropriate decision

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
2
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
1. Check Sheet
 Purpose:
 Simplification of data collection process
 Reduce recording errors
 Ease of analysis

How to use?
• Have clearly defined objectives, e.g
– Control & monitor the process / product,
– Analyze non-conformity,
– Inspection,
• Stratify data w.r.t. source,
• Measurement should be reliable,
• Find right way to record data,

 Types
 For continuous/ variable data

 For attribute data

 For concentration of defects

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
Check Sheet

1) Check sheet for production process distribution

Eg : Specification for machining parameter is 30.5 +/-
0.006.
How do you ask operator to record this data?

Is it easy for operator to record each data like 30.503,

30.505 etc?
NO
We need to simplify the data collection method and also
we should be able to see data distribution.

How can we do that?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
5
Check Sheet
Check sheet for production process distribution Spec: 30.5 +/-0.006
Spec Deviation Checks Frequency
1 5 10 15 10
-10
-9
-8
-7
-6 1
-5 3
-4 7
-3 10
-2 11
-1 8
30.500 0 6
1 4
2 2
3 2
4 1
5 0
6 0
7
8
9
10

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
6
Check Sheet
Check sheet for production process distribution – Application

 When we want to know the variation in the data

 If we want to know the distribution of the data (if
we want to plot histogram of the data)
 When stratification is required and we want to use
only 1 check sheet – we can use different colors
in the same check sheet, so that data can be
further analyzed.

Can you think of application in your work area?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
7
Check Sheet
2. Defective item check sheet

Defects

Defectives

If it is suspected that there is a shift to shift variation, is the design

of check sheet good?
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
8
Check Sheet
Defective Item check sheet – Application

 Generally used in final inspection of the process

 Helps to calculate number of defects and defectives
at the end of the shift easily
 It will not indicate the cause of the defect, it can
provide information which type of defect is
occurring more, so that improvement actions can
be directed towards that
Note: For the data collected in the example given, it is
impossible to know shift to shift variation at later stage,
stratification is required to be considered at designing the
checklist.
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
9
Check Sheet

3. Defect location check sheet

External defects like scratch, dent mark etc are
found some time on specific area of the
product. This check list helps identify the
location where more defects are produced and
hence it helps to detect the cause of the
problem.
It helps to know the distribution of the defect
within the component.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
10
Check Sheet

Instead of saying just number of defects, if you specify location E3, E1, F2,
F1 & A3 are having defects, you can take go towards the root cause easily
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
11
Check Sheet

4. Defect cause check sheet

If we want to stratify based on the causes like
operator wise, machine wise, shift wise along
with the type of defect occurred, it is possible
apply defect cause check sheet.

It helps in analyzing the data cause wise and

improves the understanding of the process

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
12
Contents

Study the data and judge the process behavior.

It is also possible to prepare check sheet along with Cause & Effect diagram
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
13
Various type of check sheets

 Production process distribution check sheet :

Collects the data in easy manner & distribution can be seen

 Defective item check sheet : Defect wise data collected

to prioritize Eg: final inspection Helps to prioritize the
problem to be selected

 Defect Location check sheet:Helps in understanding

the location where more defects are generated – used for
defects like dent, scratch etc – Provides clue for solving the
problem

 Defect Cause check sheet

Shift wise, machine wise, operator wise etc data is gathered
in checklist. Provides clue for solving the problem

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
14
Check points for check sheets preparation

Below items can be added , as necessary

1. The purpose of the checks

2. The items being checked

3. The methods of the checks

4. The dates and times of the checks

5. The person to perform the checks

6. The results
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
15
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
2. Stratification ( Classification)
Grouping the data based on particular features
eg:
Operator wise Skill, experience, age, union, qualification
Material wise Manufacturer, type, date of arrival, date of use
Time wise Day, night, shift, working days & holidays, week
Machine wise Type of machine, machine number

Exercise: Each one write down

the possible application of
check sheet & stratification and
Share with class

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
17
Stratification
Stratification can be used along with other QC tools for
better understanding of process

• Check sheet and Stratification can be used in combination to

collect the data which can provide better understanding of the
process
• Stratification can be used along with histogram for better
understanding
• Stratification can be used along with Pareto diagram
• Stratification can also be applied along with scatter diagram

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
18
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
3. Pareto Diagram
 Developed by V. Pareto, an Italian economist and applied
by Dr. JM Juran, a leading quality consultant.
 Used for prioritization by 80-20 rule.

 Differentiates ‘Vital few’ from ‘Useful many’.

Pareto Principle
 Pareto principle holds good to the present day in various
applications
 ‘ A few causes lead to many defects;

 many causes lead to few defects.’

 The few causes that lead to many defects are the vital few.
 The many causes that lead to few defects are the Useful
many.
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
20
Pareto Diagram

“Get to the biggest problems first”

‘Solve the vital few’

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
21
Pareto Diagram
Preparatory steps for Pareto diagram
1- Decide the problem to be investigated and how to collect the data
Eg – Defective item, losses in monetary terms etc
2. Decide what data is necessary and how to classify them
Eg - Type of defect, location wise, process wise, machine wise, operator
wise etc
3. Summarize the heading which are not frequently occurring as
others.
4. Determine the method of data collection and period of data
collection.
5. Collect the data using check sheet.
6. Prepare Pareto diagram as per the steps given in next slide.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
22
Pareto Diagram–Key Steps & Example
Philosophy: 20% of causes will contribute for 80% of problem
Scale selected: 2492 on left side = 100% in right side
Data Arranged in
Actual Data
Descending order
Defect Quantity DEFECT Quantity Cum % VITAL FEW
Parato Analysis for Defects
USEFUL MANY
Ovality 10 Weld Failure 620 24.88 2492
Rust 65 Weld open 515 45.55
Lenth not 100.00
as per Fin Height 90.00

Nof of defects
spec 20 425 62.60 2000 80.00

Cumilative %
Line mark 350 Line mark 350 76.65 70.00
Weld 1500
Dent 60.00
Failure 620 125 81.66
Diameter Diameter
50.00
Undersize 100 Oversize 120 86.48 1000 40.00
620
Diameter 515
425 350
30.00
Weld open
515 Undersize 100 90.49 500 20.00
125 120 100 65 42
Dent 125 Rust 65 93.10 40 20 20 10 10 10 20 10.00
Chemical 0 0.00
Burr
20 propery 42 94.78

y
ve nt

t
ne ht

pe urr
e

s
ss

h
en

us
k

s
96.39

y
Hardness 10

lit
Thickness 40

nd z e

ze

c
W ilur

er
ar

is
e

es
g

er

pe

va

e
op

B
ei

th
si

fin
m

n
rs

op

kn
Surface

Fa

rs
H

O
er

Su rd

O
d
Burr

pr

e
ic
n
el

a
d

ac
finish 10 20 97.19

O
Fi

Th

H
Li
el

al

rf
as
W

er
Chemical

ic
Lenth not

er
et

m
et

t
no
m
propery 42 as per spec 20 97.99

he
m
ia

ia

h
D
Others 20 Ovality 10 98.39

nt
D

Le
Diameter
Hardness
Oversize 120 10 98.80
Surface Defect
Fin Height
425 finish 10 99.20
Thickness 40 Others 20 100.00
Interpretation: First 4 defects are contributing for 80% of total defects
2492
1. Arrange the data in Descending order 6. Draw horizontal line for 80% and identify the point of
intersection on cumulative graph
2. Mention “Others” at the end (combining all other minor
defects) 7. Draw the vertical line from the point of intersection
3. Calculate the total & Calculate cumulative % 8. Identify left side of vertical line as “vital few” and
right side as “useful many”
4. Select the scale in such a way that total defect in Y axis of
left side is exactly matching with 100% on right side Y axis 9. Action to be initiated on vital few.
5. Plot the Pareto Chart as shown in example : Bar chart for 10. Pareto of vital few’s may be further analyzed up to
numbers and line chart for Cumulative % next level to see the vitals ones (3 level Pareto)
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
23
[Link] points on Pareto Diagram
1. Include necessary information like title, quantity, period of
data collection
2. Pareto can be prepared at different levels
A) Pareto by Phenomenon
Quality : Defects, Failures, Complaints, repairs, warranty
Cost: Amount of loss, expense
Delivery: Delay in delivery, Stock shortage, Delay in payments
Safety: Accidents, Mistakes, Breakdowns
b) Pareto by Causes :
Operator : Shift, Group, Age, Experience, Skill level
Machine: Make, Model, Instrument, Tools,
Raw material : Manufacturer, Lot, Type
Method: Angle, Speed, Position, temperature

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
[Link] points on Pareto Diagram

3. Check various classification

Construct many kinds of Pareto diagrams to view from various
angle
Eg: Rejection cost product wise, Rejection cost defect wise, rejection
cost, rejection cost machine wise, Defect wise for selected product,
Operator wise for selected defect, Machine wise for selected defect
4. Show the cost where possible, it is better to put cost in the
Y-axis (Since cost is an important parameter of measurement)
5. Simple solution if known, than same can be implemented
even though it is relatively of lesser importance.
6. Pareto by causes is very important to find out the causes, so
don’t end up with making Pareto only at first level
7. Others are relatively large, look for adding some more
classification

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
Pareto - Stratification
Linking or stratifying Pareto charts can also focus on
the primary causes of defects as shown below.
Field Service Customer Complaints Shipping Complaints
10
12 10 60
0
0
11 0 55 90%
Number of Complaints

0
10 50 80%

Cumulative Percent
0

Number of Complaints
Cumulative Percent
45 75
90 75
40 65%
80
70 35
60 50 30 50
50 25 40%
40 20
30 25 15 25
20 10
10 5
0 0 0 0
Shipping

Installation

Delivery

Clerical

Misc.

Crushed

Wrong
Code

Quantity
Wrong

Moisture

Others
A two-level
Pareto

This is to prioritize. One may not

try to find the exact 80%
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
26
Pareto Diagram

Exercise 1
Number Exercise: Prepare Pareto and
Type of of
Defect Defects
what is the conclusion?
Strain 104
Scratch 42
Pin Holes 20
mark 3
Crack 10
Stain 6
Gap 4
Color
variation 1
Dent 5
Dimensional
variation 2
%C
variation 3
Total 200

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
27
Exercise 2 – Pareto Diagram
Strain
Scratch
O Pinhole
# Crack
Others

How will apply Pareto

in this situation?-
Discuss in team

Prepare Pareto to
based on your
discussion and
interpret the results

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
28
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
4. Cause Effect Diagram

The analysis of “why?” for “what?” is cause and

effect diagram

Pareto chart

Cause-and-effect
diagram

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
30
 C&E Diagram Structure

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
In 1953, Kaoru Ishikawa, Professor of the University of Tokyo, used the Cause
& effect diagram for the first time.
A cause & effect diagram is also called a fish bone diagram since it looks like
the skeleton of a fish.

The EFFECT or PROBLEM is stated on the right side of the diagram and
the major INFLUENCES or CAUSES are listed to the left.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
32
There are two steps of making cause & effect diagrams:

1. Identify all the causes in one cause & effect diagram.

2. Take all the identified causes & classify them systematically

in another cause & effect diagram.

• Write all causes.

• Don't classify them.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
33
4a. Brainstorming

It is used to come up with creative ideas from team (idea may be cause
or solution)
It is structured way to identify the causes, think about relationship
which helps to solve the problem
Effective brainstorming
1) Write on what you are doing brainstorming
2) Ask people to think individually
3) Collect as many idea’s as possible (Focus on quantity of ideas)
4) All ideas are welcome , none are excluded in the beginning
5) Absolutely no discussion on the ideas
6) Do not criticize or judge any idea
7) Set the time limit
8) Each one calls out idea and facilitator to record them on board or
flipchart one by one exactly in the same way
9) People can build new ideas based on others ideas

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
34
 Steps to prepare C&E Diagram

Step 1
List all the causes that have been suggested by team members
as a part of brain storming.
Example

Poor operator skill

Improper
clamping on
jig or fixture
Wrong inspection method Wrong spindle speed
Operator
fatigue
1
Wrong Wrong
Insufficie inspection feed
nt training instrument
Dimensional
Variation
No inbound
inspection of
Machine Too much raw material Raw material
vibration tightening of job dimension too
close to final
dimension
Wrong setting of job on locator Improper material storage Improper / warped shape of raw material

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
35
 Step 2
Connect the sub causes to the main causes. The main
causes should then be connected to the effect.

Step 3
Assign an importance to each factor, & mark the particularly
important factors which seem to have a significant effect.

Step 4
Draw the diagram & continually look for improvement.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
36
Example of Cause Effect Diagram

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
37
Steps for C&E diagram

1. Decide the quality characteristic

2. Find as many causes as possible which are considered
to affect the quality characteristics
3. Sort-out the relation among the causes & make a C&E
diagram by connecting those elements with the quality
characteristic by cause & effect relations
4. Mark important factor which seems to have significant
effect based on observations made & judgment
5. Add necessary information such as title, name of the
product, process or group, list of participants, date

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
38
Some Hints for Cause effect diagram

• On preparation:
– List all factors with involvement of all concerned.
(Conduct brainstorming)
– State the characteristic (effect) very clearly.
– No. of CE diagrams= no. of characteristics you wish to
control.
– Both characteristics and factors should be measurable
– Select factors where action is possible.

• On usage
– Assign importance of each factor based on data.
– Try to improve the CE diagram continually while using it.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
39
Cause & Effect Relationship

Cause 1 Effect 1

Cause 2 Effect 2

Effect 1
Cause 1
Effect 2

Cause 1
Effect 1
Cause 2

C & E’s may be related in different ways.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
40
Benefits of Cause and Effect Diagrams:
• Can be applied to product & service related functions
• Complement the brainstorming process
• Provide concise information
• Promote teamwork
• Clarify understanding
• Identify potential problem areas
• Categorize causes of variation
• Makes improvement easier

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life
41
4b. Cause Analysis Table

 It is very simple and effective tools for problem solving

 It is logical and asks key questions
 It enhances the understanding of the current situation
 It helps identify the area for data collection
 It helps is identifying the actions
 Each cause identified during brainstorming/ causes that are
important as per relations diagram can be further analyzed
 It helps in determining the area where process specification
needs to be established
 It helps in identifying the area where check points to be
added.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 42
CAUSE ANALYSIS TABLE
Causes IS THRE IF YES WHAT IS IS IT IS THERE ACTION PLAN R TAR ST
(description) ANY WHAT IS THE CHECKED? ANY E GET AT
SPEC? SPEC? BASIS? HOW? DIFEREN S DAT US
FREQUENCY CE? P E
?
NO Define Spec

YES NO Establish Basis

YES YES NO Start Checking

YES YES YES YES WHY -WHY

YES YES YES NO This is not cause

Why – Why analysis

Process Failure Mode Causes
No No Description WHY WHY WHY WHY WHY ACTION PLAN

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 43
Cause Analysis Table Instructions
Causes Is there any If YES what is What is the Is it checked? Is there Action Plan Resp. Targe Status
N Description specification? the basis of How? any t Date
o (YES/NO) specification? Spec? Frequency? Difference

How to Fill Cause Analysis Table?

1 Write causes no as C1, C2 - just to give numbering for the cause

2 Write the causes as per you have identified during brain storming d (Cause & Effect Diagram)

Write YES or NO , at " is there any spec?" column. Eg : If the causes is related to operator
3
skill, write YES only of operator qualification criteria covers this skill
Write the basis only if it is justifiable like DOE, IS, Theoretical calculation, or which is an
4 accepted method, or consequence is clear if not followed. Basis should be accepted by all
the people
Write the method of checking and frequency of checking, instrument used, calibration related
5
details etc.

Note : 1. Mention cause very clearly, so that there is no ambiguity. Eg don’t write operator error,
malfunctioning, or improper method, write very clearly what type of operator error, what type of
malfunctioning or what way it is improper
2 Write the actual spec or give reference of SOP containing spec

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 44
EXAMPLE -1
CAUSE ANAYSIS TABLE
Pro Failure Causes Is there If YES what What is the Is it checked? How? Is there Action Plan Resp. Target Status
cess Mode N Description any spec? is the spec? basis of Spec? Frequency? any Date
No o (Y/N) Difference
10 Leakage in C1 Not properly Define the specification of Mr. Nawle 15.10.2 Taken trials-
adhesive tightened tightening torque- Take trails to 005 to be added
bottle NO define the specification in
specification
sheet
C2 Bottle has YES No hole in the If there is hole in Add bottle checking during Mr. Phule 15.10.0 Not Started
hole bottle bottle it will No incoming inspection or in the 5
create leakage suppliers FMEA & Control plan
20 Less C3 Mixing carried YES 30 Min R & D result & Yes, by clock, time Conduct WHY WHY
stability of out for less Minimum past experience seen by the operator & YES
the adhesive time stop time also seen by Analysis (RAT)
operator & noted in Operator need to be trained about Mr. 15.10.0 Completed on
production book for importance of mixing time Prasad 5 10.10.05
every batch Make error proofing- once time is Mr. 25.11.0 P.O. Raised
set mixing machine will Jeevan- 5 for auto
Action Decided based on
automatically switch only after HOD switch
WHY WHY analysis
completing the required cycle time. maint
C4 Operator YES As per SOP R & D results As per SOP No No action is required
adding more
or less
chemicals

ROOT CAUSE ANAYSIS TABLE

Pro Failure Causes
cess Mode N Description WHY WHY WHY WHY WHY ACTION PLAN
No o
20 Less C3 Mixing carried Operator Operator over confident- as Operator not Operator need to be trained about
stability of out for less time stops mixing last time it did not produce knowing the importance of mixing time
the at less time defect due to less time importance of this
adhesive activity
Operator stops mixing by miss Operator mistake Make error proofing- once time is set
calculation in time mixing machine will automatically
switch only after completing the
required cycle time.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 45
 Example-2
CAUSE ANAYSIS TABLE
Pro Failure Causes Is there any If YES what is the What is the Is it checked? How? Is there Action Plan
cess Mode N Description specification? specification? basis of Spec? Frequency? any
No o (YES/NO) Difference
Composit YES Less than No Conduct DOE to establish the
error more 50Micron right spec for composite error
than spec C1 Feed & Speed YES 1200 +/-20RPM, No Take trials to establish the
0.16mm/Rev basis
C2 Tool wear out YES 250 nos Trials taken - Checked manually YES Counter should automatically
report available stop the machine after 250 nos-
Add mechanism to machine
C3 Centre worn out No Define centre wear out, spec &
add it in weekly maintenance.
check sheet
C4 Dead centre pressure variation No Decide the method to control
during setting dead centre pressure control &
implement
C5 Mis alignment of machine Yes Below 10Micron- No Establish the basis for 10micron
centre defined in PM using CAD
check sheet
C6 Operator not skilled to set lead No Skill matrix should address
and profile error setting of lead & profile, Add in
operator training manual
C7 Lead Error Yes Max 27micron JIS Every setting & every 2 Yes Blank harness to be checked
hours- lead profile error and passed at incoming Q.A.
checking machine only if it is within spec
Cutter changing frequency need
to be established. Cutter
sharpness checking to be
added in weekly P.M. Check
Feed control should be locked
once set & feed should be
maintained as per spec
Train the operator to tight the
lock nut & add locknut
tightness in setting check sheet

Train operator to check lead

error

Refer next page for Root Cause Analysis Table

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 46
 Example -2

WHY - WHY ANALYSIS

Pro Failure Causes
cess Mode N Description WHY WHY WHY WHY WHY ACTION PLAN
No o
Composite C2 Tool wear out Operator not Counting is difficult & Operator gives less Counter should automatically stop the
Error More changing re- tool re-sharpening will importance to tool machine after 250 nos- Add
Than Spec sharpening the reduce production sharpening mechanism to machine
tool after 250 frequency
C7 Lead error more than Cutter worn out Blank hardness more Blank harness to be checked and
27microns after 2 hours than spec passed at incoming Q.A. only if it is
within spec
Cutter life is over Cutter changing frequency need to be
established. Cutter sharpness
checking to be added in weekly P.M.
Check
More feed Feed control should be locked once
set & feed should be maintained as
per spec
Setting disturbed Lock nut not tightened Operator is not Train the operator to tight the lock nut
during operation. properly aware of & add locknut tightness in setting
importance of check sheet
locknut tightening
Operator checking Operator not trained to Train operator to check lead error
lead error wrongly check lead error

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 47
Why-Why Analysis
Why-Why analysis can be done along with Cause and Effect diagram. It
can also done during Cause Analysis Table. It is also used for taking
corrective action at system level
 Root Cause – Most Basic Reason a Problem Has or Could Occur

1. Ask “Why” 3 to 5 times approximately

Symptom 1
Why did this “why” Symptom 2
happen?
“why” Symptom 3
“why” Symptom 4
And more “why’s” Probable Root Cause

2. Get to something “Actionable.”

 Something can be done that will, if fixed, prevent problem from existing or
recurring.
 You or your department can do something about the probable root cause. (Do
you have control over the probable root cause?)

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 48
Why-Why… (continued)

3. Check the logic in reverse direction.

If there are possibilities of other causes – write them below the sub
cause, to create a tree diagram

4. Check the logic from probable root cause to problem/top event.

 If the probable root cause is eliminated or corrected, would it prevent the
problem from existing or occurring?
 When the probable root cause occurs, does the problem occur?

5. Don’t write solution in place of cause.

No procedure and no training are usually potential solutions, not a potential
root cause.

6. When to stop why-why analysis?.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 49
Example: Why-Why
Find out
Winding arm
Why Why Why bent
Uneven Inconsistent
? ? ?
winding edges slitting width
Blade holder Investigate

Why loosened
Fail quality
? inspection

Poor securing Investigate

Setup time is Slitting width is
mechanism in
too long difficult to set Why
Why Why blade holder
? ? ?

Shopfloor Why Investigate

Delivery cycle Why Why Cannot find Why No proper storage
Slitting cycle controller did ?
time exceed 4 ? ? subsidiary ? place in shopfloor
time is long not prepare
weeks materials for sub. material
sub. materials

Why Wait for raw Why

? material from ?
warehouse

Packer at Contractor did Warehouse

Packing Investigate
warehouse not pull FG to contractor did not
staging time
cannot find warehouse on- plan for sufficient
Why too long Why Why Why
finished goods time personnel
? ? ? ?

Exercise 4 & 5
CAT & Why-Why

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 50
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 51
5. Histogram
A Histogram can be used:
• To display large amounts of data values in a relatively
simple chart form.
• To tell relative frequency of occurrence.
• To easily see the distribution of the data.
• To see if there is variation in the data.
• To make future predictions based on the data.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 52
6. Overview of Histogram

• The characteristics of the frequency distribution are shown

more clearly when results are plotted in form of block
diagram
• The horizontal axis is divided into segments corresponding
to ranges of the group
• On each segment a rectangle is constructed whose height
is proportional to the frequency in the group
• Higher bar represents more data values
• Lower bar represents less data values.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 53
Histogram
How to make a histogram?
Let us make a histogram using an example.

Example:
To investigate the distribution of the thickness of the
spectacle lens produced, the thickness of 90 lens
are measured as shown in the table.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 54
 Histogram
Thickness of Lens
Sample
Results of Measurement
Number
1 - 10 2.51 2.517 2.522 2.522 2.51 2.511 2.519 2.532 2.543 2.525
11 - 20 2.527 2.536 2.506 2.541 2.512 2.515 2.521 2.536 2.529 2.524
21 - 30 2.529 2.523 2.523 2.523 2.519 2.528 2.543 2.538 2.518 2.534
31 - 40 2.52 2.514 2.512 2.534 2.526 2.53 2.532 2.526 2.523 2.52
41 - 50 2.535 2.523 2.526 2.525 2.532 2.522 2.502 2.53 2.522 2.514
51 - 60 2.533 2.51 2.542 2.524 2.53 2.521 2.522 2.535 2.54 2.528
61 - 70 2.525 2.515 2.52 2.519 2.526 2.527 2.522 2.542 2.54 2.528
71 - 80 2.531 2.545 2.524 2.522 2.52 2.519 2.519 2.529 2.522 2.513
81 - 90 2.518 2.527 2.511 2.519 2.531 2.527 2.529 2.528 2.519 2.521

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 55
Histogram
Step 1
Calculate the range (R)
Obtain the largest & smallest of observed values & calculate R.
R = (the largest observed value) – (the smallest observed value)
2.545-2.502 = 0.043
Diameter after grinding
Sample
Results of Measurement
Number
1 - 10 2.51 2.517 2.522 2.522 2.51 2.511 2.519 2.532 2.543 2.525
11 - 20 2.527 2.536 2.506 2.541 2.512 2.515 2.521 2.536 2.529 2.524
21 - 30 2.529 2.523 2.523 2.523 2.519 2.528 2.543 2.538 2.518 2.534
31 - 40 2.52 2.514 2.512 2.534 2.526 2.53 2.532 2.526 2.523 2.52
41 - 50 2.535 2.523 2.526 2.525 2.532 2.522 2.502 2.53 2.522 2.514
51 - 60 2.533 2.51 2.542 2.524 2.53 2.521 2.522 2.535 2.54 2.528
61 - 70 2.525 2.515 2.52 2.519 2.526 2.527 2.522 2.542 2.54 2.528
71 - 80 2.531 2.545 2.524 2.522 2.52 2.519 2.519 2.529 2.522 2.513
81 - 90 2.518 2.527 2.511 2.519 2.531 2.527 2.529 2.528 2.519 2.521

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 56
Histogram
Step 2
Determine the no. of bins & class interval
Class interval is determined so that range will include the maximum and
and minimum of values
The no. of bins is calculated by the formula
No. of bins = √ n where n is total number of observations
Here, n = 90
Therefore, √ n = 9.48.
Rounding to nearest integer,
The no. of bins is 9. This is just a thumb
R 0.043 rule – does not work
Class Interval =
all the time
√n 9
= 0.005 (approx)

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 57
Histogram
Divide the range by 1,2,5, or 0.1,0.2,0.5 , or 10,20,50, [Link] the values
are so as to obtain from 5 to 20 class intervals of equal width.
Where there are two possibilities, use narrower interval ( more classes )
for below 100 readings and for 100 & above readings, use wider intervals
( less classes)
Step 3
Prepare a frequency table form
Prepare a form as shown below on which class, mid – point, frequency
marks, frequency.

sn Class Midpoint Frequency marks Frequency

(tally)

Total

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 58
Histogram
Step 4
Determine the class boundaries
Determine the boundaries of each bin so that they include the smallest & the
largest of values, and write these down on the frequency table.
First, determine the lower boundary of the first bin & add the class interval to
obtain the class boundary
The lower boundaries of the first bin can be either 2.5000 or 2.5005 (it has to
be less than the smallest value 2.502).
Therefore, 2.5005 + class interval
2.5005 + 0.005 = 2.5055
Therefore first class boundary : 2.5005 ~ 2.5055
The second class boundary : 2.5055 ~ 2.5105
Note that this has to contain the largest recorded value. Therefore,
9th class boundary : 2.5405 ~ 2.5455

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 59
Histogram
Step 5
Calculate the mid – point of the class
Using the following equation, calculate the mid-point of class, & write
this down on the frequency table.
Sum of the upper & lower boundaries of
Mid – point of each class = each class
2

Step 6
Prepare the fill up the frequency table with tally mark and count the
frequency

Step 7
Draw the Bar Graph with X axis as mid point of interval and Y axis as
frequency. Draw the smooth curve of histogram

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 60
Histogram
Class Midpoint Frequency marks Frequency
(tally)

1 2.5005 – 2.5055 2.503 / 1

2 2.5055 – 2.5105 2.508 //// 4
3 2.5105 – 2.5155 2.513 //// //// 9
4 2.5155 – 2.5205 2.518 //// //// //// 14
5 2.5205 – 2.5255 2.523 //// //// //// //// // 22
6 2.5255 – 2.5305 2.528 //// //// //// //// 19
7 2.5305 – 2.5355 2.533 //// //// 10
8 2.5355 – 2.5405 2.538 //// 5
9 2.5405 – 2.5455 2.543 //// / 6

Total 90
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 61
Uses of Histogram

A Histogram can be used:

• To display large amount of data values in a relatively
simple chart form.
• To tell relative frequency of occurrence.
• To easily see the distribution of the data.
• To see if there is variation in the data.
• To make future predictions based on the data.

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 62
Interpretation of Histogram

 Look at the specifications – it indicates the

acceptance and rejection or non-conformance
 Look at the peak – it represents the setting
 Look at the width – it represents variability
 Look at the shape – it represents the
predictability

What are your interpretations on these histograms?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 63
Histogram
Types of distribution
• The shape of the distribution gives a more elusive concept
than mean or standard deviation
• From the distribution we can deduce the peak value of
frequency and symmetry of the data range

(i) Normal distribution

(ii) Positively Skewed

(iii) Twin Peak (Bimodal)

(iv) Comb type

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 64
Histogram with Stratification

Height of 10000 people

African Japanese Indian

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 65
Interpretation of Histogram

 Look at the specifications – it indicates the

acceptance and rejection or non-conformance
 Look at the peak – it represents the setting
 Look at the width – it represents variability
 Look at the shape – it represents the
predictability

What are your interpretations on these histograms?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 66
Property of Normal Distribution
00.4
68.2%

0.3
95.4%
0.2

99.73%
0.0.1

0 .0

-- -4 -3 --2 --1 0 1 2 3 4
Output s

Since +/- 3 sigma takes care of 99.73% of the values of the process, it is
considered as process variation (Voice of the process)
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 67
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 68
6. Scatter diagram

In actual practice, it is often essential to study the relation of two

corresponding variables.
For example, how does the speed of driving a car affect its fuel
efficiency?

To study the relation of two variables such as the speed of the car & the
fuel efficiency we can use what is called a Scatter diagram.

The two variables we will deal with are:

a) A quality – characteristic & a factor affecting it , (Y & X)
b) Two related quality characteristics, or (Y & Y)
c) Two factors relating to a single quality characteristic. (X1 & X2 which
are related to Y)

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 69
Scatter Diagram - Example
Collect paired data (x,y) for which you want to study the relations &
arrange the data in a table. It is desirable to have at least 30 pairs of data.
Air pressure Percent
Day
(kgf/cm2) Defective
1 8.6 0.889
2 8.9 0.884
3 8.8 0.874 0.93
4 8.8 0.891
5 8.4 0.874 0.92
6 8.7 0.886
7 9.2 0.911 0.91
8 8.6 0.912

% Defective
9 9.2 0.895 0.9
10 8.7 0.896
11 8.4 0.894 Plot Scatter Diagram 0.89
12 8.2 0.864
13 9.2 0.922
14 8.7 0.909 0.88
15 9.4 0.905
16 8.7 0.892 0.87
17 8.5 0.877
18 9.2 0.885 0.86
19 8.5 0.866
20 8.3 0.896
21 8.7 0.896
0.85
22 9.3 0.928 8 8.5 9 9.5
23 8.9 0.886
24 8.9 0.908
25
26
8.3
8.7
0.881
0.882
Air Pressure
27 8.9 0.904
28 8.7 0.912
29 9.1 0.925
30 8.7 0.872
What is your conclusion?
Can you interpret?
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 70
Scatter diagram
Data of blowing air pressure & percent defective of plastic tank
Air pressure Percent
Day
(kgf/cm2) Defective 1. Collect at least 30 pairs of data of x & y (without varying
1 8.6 0.889
2 8.9 0.884
other possible factors)
3 8.8 0.874
4 8.8 0.891 2. Identify maximum and minimum
5 8.4 0.874
6 8.7 0.886 3. Select appropriate scale Scatter plot of Air pressure & % defective
7 9.2 0.911
8
9
8.6
9.2
0.912
0.895
4. Plot the graph
0.924
10 8.7 0.896
11 8.4 0.894
12 8.2 0.864
0.914
13 9.2 0.922
14 8.7 0.909
15 9.4 0.905 0.904

% Defective
16 8.7 0.892
17 8.5 0.877
18 9.2 0.885 0.894
19 8.5 0.866
20 8.3 0.896
21 8.7 0.896 0.884
22 9.3 0.928
23 8.9 0.886
24 8.9 0.908 0.874
25 8.3 0.881
26 8.7 0.882
27 8.9 0.904 0.864
28 8.7 0.912 8.2 8.4 8.6 8.8 9 9.2 9.4
29 9.1 0.925 Air pressure (Kg/cm 2)
30 8.7 0.872

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 71
Scatter diagram & Correlation Coefficient

How to read scatter diagrams?

40 400 800

30 300 600

20 200 400

10 100 200

0 0 0
0 5 10 15 20 0 100 200 300 400 0 100 200 300 400

Positive correlation Negative correlation No correlation

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 72
Scatter Plot
Scatter
Plots-
Various
Possibilities

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 73
Scatter Plot

The same data is plotted on different scale. Does scale

make a difference? What should be the appropriate scale?
Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 74
Scatter diagram
CORRELATION COEFFICIENT( r ):
•The quantitative measure of correlation between two variables
• “r” will range from -1 to +1
• -1 indicates very strong –ve correlation
• +1 indicates very strong +ve correlation
• Score of 0 indicates no correlation

POSITIVE NEGATIVE

r= S(xy) STRONG r 0.8 r  -0.8

MODERATE 0.5  r < 0.8 -0.5  r >-0.8
√(S(xx) * S(yy) WEAK r < 0.5 r >-0.5

Rule of Thumb

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 75
Scatter diagram
2 2
Sr. No x y
1 8.6 0.889
x y xy
73.96 0.790321 7.6454
S(xx)=∑xi2 – (∑xi )2
2 8.9 0.884
3 8.8 0.874
79.21 0.781456 7.8676
77.44 0.763876 7.6912
n
4 8.8 0.891 77.44 0.793881 7.8408
5 8.4 0.874 70.56 0.763876 7.3416 S(xx)= 2312.02 - (263.2)2 /30= 2.88
6 8.7 0.886 75.69 0.784996 7.7082
7 9.2 0.911 84.64 0.829921 8.3812
8
9
8.6 0.912
9.2 0.895
73.96 0.831744 7.8432 S(yy) = ∑yi2 – (∑yi )2
84.64 0.801025 8.234
10
11
8.7 0.896
8.4 0.894
75.69 0.802816 7.7952 n
70.56 0.799236 7.5096
12 8.2 0.864 67.24 0.746496 7.0848
13 9.2 0.922 84.64 0.850084 8.4824 S(yy)= 23.97833- (26.816)2/30 = 0.0084
14 8.7 0.909 75.69 0.826281 7.9083
15 9.4 0.905 88.36 0.819025 8.507
16 8.7 0.892
17 8.5 0.877
75.69 0.795664 7.7604
72.25 0.769129 7.4545 S(xy)= ∑ xiyi – (∑xi ) * (∑yi )
18 9.2 0.885 84.64 0.783225 8.142
19 8.5 0.866 72.25 0.749956 7.361 n
20 8.3 0.896 68.89 0.802816 7.4368
21 8.7 0.896 75.69 0.802816 7.7952
22 9.3 0.928 86.49 0.861184 8.6304 S(xx)= 235.357 - [(263.2)*(26.816)]/30=0.0913
23 8.9 0.886 79.21 0.784996 7.8854
24 8.9 0.908 79.21 0.824464 8.0812
25 8.3 0.881 68.89 0.776161 7.3123 r= S(xy)
26 8.7 0.882 75.69 0.777924 7.6734
27
28
8.9 0.904
8.7 0.912
79.21 0.817216 8.0456 √(S(xx) * S(yy)
75.69 0.831744 7.9344
29 9.1 0.925 82.81 0.855625 8.4175
30 8.7 0.872 75.69 0.760384 7.5864
Total 263.2 26.816 2312.02 23.978338 235.357 r= 0.0913/ √ 2.88*0.0084= 0.59
∑xi ∑yi ∑xi2 ∑yi2 ∑xy

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 76
Correlation vs. Causation
 Regardless of method employed, it is important to keep in mind that a
strong mathematical (or graphical) relationship between two variables
does not confirm that one causes the other.

Two variables can be highly related to one another, but neither is caused
by the other.

 Validation of root cause is made only when two standards are met:
 There is a statistically significant relationship between the suspected root
cause and the effect.
 Knowledge of the process collaborates this causal relationship.

 One of these alone is not adequate validation.

 What is confounding, common cause & coincidence ?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 77
 7 QC tools

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 78
Fundamentals of Control charts
 How can we know whether variation is due to
common cause or special cause?
 How can we know when to adjust the process
and when not to adjust process?
 How can we know shape, spread and location is
consistent or not over period of time?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 79
Control chart Basic Chart Elements

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 80
80
Control Chart Steps
Let us look at practical example of control charts through
video

Some questions
1. How to decide subgroup size?
2. How long data need to be collected?
3. How to find out rational sub grouping?
4. Why to calculate within subgroup variation?
5. How can you say points above/below the control limits is
special cause?
6. What is needed before starting for control charts?

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 81
81
Understanding X Bar and R chart
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
X1 204 190 203 204 206 194 176 200 207 196 175 195 206 202 219 184 197 203 216 204 188 196 198 202 210
X2 201 203 207 192 204 201 213 212 204 226 204 208 216 195 198 193 197 179 218 206 201 201 195 201 203
X3 200 203 186 197 188 200 206 202 195 204 197 221 212 203 194 200 187 193 195 213 203 196 195 206 187
X4 171 184 233 207 195 199 215 203 200 186 205 216 189 213 204 189 204 179 221 215 186 216 164 188 195
X5 201 223 181 206 186 202 207 201 196 208 186 224 182 194 208 207 184 194 206 203 188 197 197 206 200
X Bar

•Within Subgroup variation

•Between Subgroup variation

X R

UCL (xbar) UCL (R)

CL (xbar) CL (R)
LCL (xbar) LCL (R)

X bar & R chart : UCLX = X + A2 R ; LCLX = X - A2 R

UCLR = D4 R ; LCLR = D3 R

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 82
82
Interpret Statistical Control
Special Cause Criteria

1 point more above UCL or Below LCL

9 points in a row on same side of centerline
6 points in a row, all increasing or all decreasing
14 points in a row, alternating up and down
2 out of 3 points > 2 standard deviations from centerline (same side)
 4 out of 5 points > 1 standard deviations from centerline (same side)
15 points in a row within 1 standard deviation of centerline (either side)
8. points in a row > 1 standard deviation from centerline (either side)

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 83
83
 Take support of MIQ while application of these
tools
 Discuss and have clarity on how you are going
to solve the problem and then start solving the
problem
 Follow the steps without any deviation

Learnt and not applied means NOT LEARNT

WISH YOU GREAT SUCCESS OF YOUR QC STORY PROJECT &

Mahindra Yellow Belt Certification

Build TQM competence | Application orientation | Spread TQM | Build team culture | Quality a way of life 84
84