7 QC TOOLS
Let’s do an exercise!
You have to cut down your
house expenditure by 20% per month How will you do it ?
Objective
OBJECTIVE: After the completion of this training programme: You will be able to choose the appropriate QC tool to solve problems in your work area.
Types
7QC tools Check Sheet
Pareto Diagram Cause & Effect diagram
7 QC Tools
Graph &Control charts Histogram
Stratification Scatter Diagram
Check sheet
Check sheet
Check sheet
What is a check sheet?
Why is a check sheet necessary?
Check sheet
Example:
Check sheet
Check sheet Check sheets are forms used for
•checking results of work • verifying and collecting data
Check sheet
Types of Check Sheet Discrete value such as no. Of recording errors, no. of Item sold & Rejections etc.
Indiscrete value such as height, weight, length, time & temp., Etc.
Measured Data
Point Scale Data
1 Point, 2 Point … etc.
Counted Data
Check Sheet Ordered Data 1st, 2nd Order … Very Good, Good, No Good … - Type
Primary Data
YES / NO or / X - Type
Check sheet
Examples for various types of check sheet Ordered data check sheet Example:
Quality of painting in paint shop can be classified as good, very good, bad etc.
Check sheet
Example: CHECK SHEET FOR PAINT QUALITY (Max 100) Sl No
Time of check
Quality
1
10.00
Good
2
11.00
Very Good
3
12.00
Good
4
13.00
Good
5
14.00
Very Good
6
15.00
Bad
7
16.00
Good
8
17.00
Good
Check sheet
Point scale data check sheet Example: Star ratings of motorcycles
Check sheet
Example: CHECK SHEET FOR STAR RATINGS (Two Wheelers) Sl No
Model
1
TVS VICTOR
2
HERO HONDA SPLENDOR
3
LML FREEDOM
4
BAJAJ BOXER
5
KINETIC CHALLENGER
6
YAMAHA CRUX
Rating
Check sheet
Measured data check sheet Example
Data of bush diameters (Cylindrical grinding) in 3rd Shift Sl. No. Diameter Range No. of components 25.010 – 25.015 1 02 25.015 – 25.020 2 05 25.020 – 25.025 3 08 25.025 – 25.030 4 10 25.030 –25.035 5 19 25.035 – 25.040 6 29 25.040 – 25.045 7 33 25.045 – 25.050 8 36 25.050 – 25.055 9 40 25.055 – 25.060 10 48 25.060 – 25.065 11 57 25.065 – 25.070 12 42 25.070 – 25.075 13 29 25.075 – 25.080 14 15 25.080 – 25.085 15 11 25.085 – 25.090 16 09 25.090 – 25.095 17 05 25.095 – 25.100 18 02 Total 400
Check sheet
Counted data check sheet Example Sales of TVS two wheelers Sl. No. Year
No. of Units
1
1994 – 95
287500
2
1995 – 96
410800
3
1996 – 97
513400
4
1997 – 98
568700
5
1998 – 99
603000
6
1999 – 00
668700
7
2000 – 01
863600
8
2001 – 02
866240
9
2002 – 03
1062000
Check sheet
Example: Primary data check sheet A check sheet with yes or no type data Quality Control Department
Sl. No.
Employee Name
TEIAN award
1
Saravanan
Yes
2
Srinivasan
No
3
Krishnan
Yes
4
Venkatraman
No
5
Subramanian
Yes
6
Padmanabhan
Yes
7
Velu
Yes
8
Senthil
Yes
Check sheet
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
Check sheet
Example of check sheet Defect check sheet Month ,day Component
4/1
2
1 2 3 4 5 6 7 8 9 10
No. of defects
3
4
Check sheet
Example of check sheet Data collection sheet
Check sheet
A1
A2
C1 E1 D1 B1 D2 D1 B2 D2
C2 E2
E1
C1 E2
E1
C2 E2
E1
E2
Check sheet
Make a check list of all the expenses in your home & the amount you spend on these expenses
Check sheet
Sl.No
Expense
Amount
1
House Rent
1500
2
Electricity & Water Bill
250
3
Cable TV Bill
150
4
News paper bill
100
5
Milk
400
6
Maid servant
150
7
Groceries
2000
8
Entertainment & Lifestyle
200
9
Travel
200
10
Educational
1000
11
Hospital
200
12
Loan repayment
1000
13
Clothes
200
14
Petrol
300
15
Others
300
Pareto
Pareto diagram
Pareto
Do you remember this? (14th March 2001 - Eden gardens )
Pareto
Let’s look at the second innings score board: India SS Das
hit wicket b Gillespie
39
S Ramesh
c ME Waugh b Warne
30
VVS Laxman
c Ponting b McGrath
281
SR Tendulkarc Gilchrist b Gillespie
10
SC Ganguly
c Gilchrist b McGrath
48
R Dravid
run out
180
N R Mongia
b McGrath
4
Zaheer Khan
not out
23
Harbhajan Singh
not out
8
Total
657
Pareto
Do you remember this? (14th March 2001 - Eden gardens )
Pareto
Who got the wickets?
O
M
R
W
Zaheer Khan
8
4
30
0
V Prasad
3
1
7
0
Harbhajan Singh
30.3
8
73
6
V Raju
15
3
58
1
S Tendulkar
11
3
31
3
S Ganguly
1
0
2
0
Pareto
Who got the maximum runs? • Laxman & Dravid – 461 / 657 runs. • 22% of the 9 batsmen who batted got 70% of the runs!
Who got the maximum wickets? • Harbhajan & Tendulkar – 9 / 11 wickets. • 30% of the 6 bowlers who bowled got 80% of the wickets!
This illustrates the Pareto principle
Pareto
Pareto • Vilfredo Pareto was an Italian engineer in the 19th Century who studied the number of people in various income classes & declared ‘’20% of the people own 80% of the country’s wealth;
80% of the people own 20% of the country’s wealth”
Pareto
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 trivial many.
Pareto
“Get to the biggest problems first” ‘Solve the vital few’
Pareto 200
84 79
150
75
73
55.5 100 75
50
66 33 45
50
25 20
25
15
12
10
8
6
5
4
4
2
2
1
Steps
1. Collect data
2. Arrange data in the descending order 3. Calculate the relative % for individual data 4. Calculate the cumulative % for individual data 5. Draw a graph with scales on both axis 6. Draw bar chart based on data 7. Using cumulative % data, draw cumulative curve 8. Identify the VITAL FEW (thumb rule > 70%)
Quality
Factory production
Manufacturing Planning
Stores
Others
Information Systems
Dept
Research & Development
Service
Personnel
Finance
Production Engineering
Plant Maintenance
Materials
0 Marketing
0
In %
65.5
125 In nos
Creating a Pareto Diagram
100 100
99.5
98.5
97.5
95.5
93.5
91
88
175
Pareto
Same problem, but different approach…
You have to cut down your
house expenditure by 20% / month How will you do it ? Paret o
Pareto
Now take your check sheet. Arrange these expenses & amounts in an order, with the highest expense being the first & lowest expense being the last
Pareto Sl.No
Expense
Amount
1
Groceries
2000
2
House rent
1500
3
Educational
1000
4
Loan repayment
1000
5
Milk
400
6
Petrol
300
7
Others
300
8
Electricity & water bill
250
9
Hospital
200
10
Travel
200
11
Entertainment & lifestyle
200
12
Clothes
200
13
Maid Servant
150
14
Cable TV bill
150 7850
Pareto
Calculate the percentage contribution of each of these expenses.
Percentage can be calculated by the formula
Individual expense Total expense
X 100
Pareto Sl.No
Department
Nos.
Relative %
1
Groceries
2000
25.47
2
House rent
1500
19.10
3
Educational
1000
12.74
4
Loan repayment
1000
12.74
5
Milk
400
5.10
6
Petrol
300
3.83
7
Others
300
3.83
8
Electricity & water bill
250
3.19
9
Hospital
200
2.54
10
Travel
200
2.54
11
Entertainment & lifestyle
200
2.54
12
Clothes
200
2.54
13
Maid Servant
150
1.92
14
Cable TV bill
150
1.92
7850
100
Pareto Sl.No
Department
Nos.
Relative %
Cumulative %
1
Groceries
2000
25.47
25.47
2
House rent
1500
19.10
44.57
3
Educational
1000
12.74
57.31
4
Loan repayment
1000
12.74
70.05
5
Milk
400
5.10
75.15
6
Petrol
300
3.83
78.98
7
Others
300
3.83
82.81
8
Electricity & water bill
250
3.19
86.00
9
Hospital
200
2.54
88.54
10
Travel
200
2.54
91.08
11
Entertainment & lifestyle
200
2.54
93.62
12
Clothes
200
2.54
96.16
13
Maid Servant
150
1.92
98.08
14
Cable TV bill
150
1.92
100
7850
100
100
Pareto Sl.No
Department
Nos.
Relative %
Cumulative %
1
Groceries
2000
25.47
25.47
2
House rent
1500
19.10
44.57
3
Educational
1000
12.74
57.31
4
Loan repayment
1000
12.74
70.05
5
Milk
400
5.10
75.15
6
Petrol
300
3.83
78.98
7
Others
300
3.83
82.81
8
Electricity & water bill
250
3.19
86.00
9
Hospital
200
2.54
88.54
10
Travel
200
2.54
91.08
11
Entertainment & lifestyle
200
2.54
93.62
12
Clothes
200
2.54
96.16
13
Maid Servant
150
1.92
98.08
14
Cable TV bill
150
1.92
100
7850
100
100
VITAL FEW
TRIVIAL MANY
Pareto
100
100
7500
VITAL FEW 75
57.31
5000 44.57
2500
50
25.47
Trivial Many 0
25
0 Groceries
House Rent
Educational Expenses
Loan Repayment
Others
Cumulative(%)
Amount
70.05
Pareto
Why pareto ? • To Clearly prioritise the magnitude of the problem. • To identify the vital few and trivial many problems. • To find 80/20 rule which states that 80% of the problems are created by 20% of the causes.
Pareto 84 79
150
75
73 65.5
125
In nos
100 100
99.5
98.5
97.5
95.5
93.5
91
88
175
55.5
100 75
50
66 33 45
50
25 20
25
15
12
10
8
6
5
4
4
2
2
1
1. The most important problem 2. The rate of each problem to the whole 3. The degree of improvement action 4. The comparison of improvement level 5. Before & after remedial action taken
Quality
Factory production
Manufacturing Planning
Stores
Others
Information Systems
Dept
Research & Development
Service
Personnel
Finance
Production Engineering
Plant Maintenance
Materials
0 Marketing
0
In %
Pareto diagram is used to find out … 200
Relation between Pareto diagram & Cause & effect diagram
The biggest problem has been found. What next?
A combination of Pareto diagram & cause and effect diagram is an ideal way to arrive at the main problem & its causes. Take the biggest problem from the pareto diagram & put it on the right side in the cause & effect diagram. Derive the causes for the same.
Cause & effect diagram Cause & Effect diagram derived from pareto
Pareto chart
Cause-and-effect diagram
Cause & effect diagram
Cause & Effect diagram
Cause & effect diagram
Why Cause & Effect ? • To identify and systematically list the different causes that can be attributed to a problem (or an effect)
• To identify the reasons why a process goes out of control
• To decide which causes to investigate for process improvement.
Cause & effect diagram
What is Effect ? EFFECT = A Result or an outcome EFFECT is What happens
Effect – Tyre puncture
Cause & effect diagram
What is cause ? CAUSE = Reason or Factor contributing to the EFFECT CAUSE is WHY it happens
Cause & effect diagram
The analysis of “why?” for “what?” is
cause and effect diagram
Cause & effect diagram 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.
Cause & effect diagram
The EFFECT or PROBLEM is stated on the right side of the diagram and the major INFLUENCES or CAUSES are listed to the left.
Effect
Cause & effect diagram
There are two steps of making cause & effect diagrams:
2
1. Identify all the causes in one cause & effect diagram. 2. Take all the identified causes & classify them systematically in another cause & effect diagram.
Cause & effect diagram
Cause & effect diagram for identifying the causes
Cause & effect diagram
In this type of cause & effect diagram: • Write all causes. • Don't classify them.
Cause & effect diagram
How to obtain most number of causes?
Brainstorming!
Cause & effect diagram
Remember
DURING BRAINSTORMING: • All the ideas/causes should be noted. • Don’t label any ideas “good” or “bad”. • Encourage free flow of ideas.
Cause & effect diagram
HOW TO DO IT?
1 Effect
Causes
Cause & effect diagram Example Poor operator skill
Wrong inspection method Improper clamping on jig or fixture
Insufficie nt training
Machine vibration
Wrong setting of job on locator
Wrong inspection instrument
Too much tightening of job
1
Wrong spindle speed
No inbound inspection of raw material
Improper material storage
Operator fatigue Wrong feed
Raw material dimension too close to final dimension Improper / warped shape of raw material
Dimensional Variation
Cause & effect diagram
Cause & effect diagram for systematically listing causes
Use these steps to make a successful Cause & effect diagram for systematically listing causes Step 1
List all the causes that have been suggested by team members as a part of brain storming.
Cause & effect diagram 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.
Cause & effect diagram
2 Effect
Causes
Cause & effect diagram
MAN
2
MACHINE
Fatigue
Imbalance
HEALTH
STABILITY
Concentration
Illness
Training
Vibration Clamping
SPIRIT Attentiveness
SKILL
JIGS & FIXTURES
Experience
Location Degree of tightening
Inspection
SETTING
QUALITY Storage
Shape FORM Dimension
MATERIAL
Placement on locator Instrument INSPECTION
Feed WORKING
Method
METHOD
Spindle speed
Dimensional Variation
Cause & effect diagram
Thorough investigation of causes
ASK WHY? 5 TIMES 4th Why 3rd Why
5th
C Why
Policies
Procedure
1st Why
D
2nd
People
Why
Plant
Why defects?
Graph&control charts
Graph & Control charts
Graph&control charts
Graph What is Graph? When there are more than 2 interrelated data sets,you write the datasets on a graph so as to clearly define the relationships.
Why Graph? The details of the data should be • Correctly understood • In their entirety • With just a one look
Graph&control charts
Graph Types of Graphs & Its application Bar graphs, line graphs, pie charts, and band graphs. There are also other specialty graphs such as radar charts, Z charts and area graphs
Applications
To understand relative sizes of numbers
To understand trends over time
To understand percent-ages of totals
Graph&control charts
Pie chart - Composition of sales turnover 2001-02 Export Sales 1%
Vehicle sales 92%
Spare parts sales 6%
Other income 1%
Graph&control charts
Bar chart - Sales performance Plan for 2002-03
Rs Million x 100 300
262
250 184 162
200 133
150
104 83
100 50
194
62
14
17
19
27
91-92
92-93
93-94
41
0 90-91
94-95
95-96
96-97
97-98
98-99
99-00
00-01
'01-02
'02-03
Graph&control charts
Line chart - Hourly output of Workmen 50 45 40
ACTUAL
40 38
45
47
41
44 42
Feb
Mar
42
44 42
43 41
Apr
May
Jun
PLAN
35 30 25 20 15 10 5 0 Jan
Graph&control charts
Line chart Cumulative Percentage 100
100
80 70.05 60
59.26 44.57
40 25.47 20
0 1
2
3
4
5
Histogram
Histogram
Histogram
• In quality control, we try to discover facts by collecting data & then take necessary action based on those facts. • The data is not collected as an end in itself, but as a means of finding out the facts behind the data.
Data
FACTS
Histogram
Example • Consider a sampling inspection. • We take a sample from a lot, carry out measurements on it. • We decide whether we should accept the whole lot or not. • The sample tells us the whether the lot is OK or NOT OK.
The totality of items under consideration is called the population.
Histogram
• The data obtained from a sample helps us to take a decision on the population. • The larger the sample size is, the more information we get about the population. • But an increase of sample size also means an increase in the amount of data
Histogram
• It becomes difficult to understand the population from these data even when they are arranged into tables. • In such a case we need a method which will enable us to understand the population at a glance. A histogram answers our needs.
Histogram
What is histogram ? Histogram is a bar chart…. 25
20
15
The sizes of the vertical bars reflects the number of data
The horizontal axis shows the values of the characteristics
10
5
0
2.50
2.51
2.52
2.53
2.54
2.55
Histogram
How to make a histogram? Let us make a histogram using an example. Example: To investigate the distribution of the diameters of steel shafts produced in the grinding process, the diameters of 90 shafts are measured as shown in the table.
Histogram
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
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)
Histogram
Diameter after grinding Sample
Maximum Minimum
Value of Value of
Results of Measurement
Number
the Line the line
1 - 10
2.51
2.517
2.522
2.522
2.51
2.511
2.519
2.532
2.543
2.525
2.543
2.51
11 - 20
2.527
2.536
2.506
2.541
2.512
2.515
2.521
2.536
2.529
2.524
2.541
2.506
21 - 30
2.529
2.523
2.523
2.523
2.519
2.528
2.543
2.538
2.518
2.534
2.543
2.518
31 - 40
2.52
2.514
2.512
2.534
2.526
2.53
2.532
2.526
2.523
2.52
2.534
2.512
41 - 50
2.535
2.523
2.526
2.525
2.532
2.522
2.502
2.53
2.522
2.514
2.535
2.502
51 - 60
2.533
2.51
2.542
2.524
2.53
2.521
2.522
2.535
2.54
2.528
2.542
2.51
61 - 70
2.525
2.515
2.52
2.519
2.526
2.527
2.522
2.542
2.54
2.528
2.542
2.515
71 - 80
2.531
2.545
2.524
2.522
2.52
2.519
2.519
2.529
2.522
2.513
2.545
2.513
81 - 90
2.518
2.527
2.511
2.519
2.531
2.527
2.529
2.528
2.519
2.521
2.531
2.511
The
The
Largest Smallest Value
Value
2.545
2.502
Histogram
R = largest value – smallest value = 2.545 – 2.502
= 0.043
Histogram
CLASS INTERVAL Class
1 2 3 4 5 6 7 8 9
2.5005 – 2.5055 2.5055 – 2.5105 2.5105 – 2.5155 2.5155 – 2.5205 2.5205 – 2.5255 2.5255 – 2.5305 2.5305 – 2.5355 2.5355 – 2.5405 2.5405 – 2.5455 Total
Midpoint Frequency marks (tally)
2.503 2.508 2.513 2.518 2.523 2.528 2.533 2.538 2.543
/ //// //// //// //// //// //// //// ////
//// //// //// //// //// //// // //// //// //// //// /
Frequency
1 4 9 14 22 19 10 5 6
90
Histogram
INTERVAL BREADTH Class
1 2 3 4 5 6 7 8 9
2.5005 – 2.5055 2.5055 – 2.5105 2.5105 – 2.5155 2.5155 – 2.5205 2.5205 – 2.5255 2.5255 – 2.5305 2.5305 – 2.5355 2.5355 – 2.5405 2.5405 – 2.5455 Total
Midpoint Frequency marks (tally)
2.503 2.508 2.513 2.518 2.523 2.528 2.533 2.538 2.543
/ //// //// //// //// //// //// //// ////
//// //// //// //// //// //// // //// //// //// //// /
Frequency
1 4 9 14 22 19 10 5 6
90
Histogram
Step 2 Determine the class interval & interval breadth. The class interval is calculated by the formula Class interval = √ n where n is total number of observations Here, n = 90 Therefore, √ n = 9.48. Rounding to nearest integer, The class interval is 9. R Interval breadth = √n = 0.005
=
0.043 9
Histogram
Step 3 Prepare a frequency table form Prepare a form as shown below on which class, mid – point, frequency marks, frequency etc can be recorded Class
Total
Midpoint Frequency marks (tally)
Frequency
Histogram
Step 4 Determine the class boundaries • Each boundary should include the smallest & the largest of values. Steps
• Determine the lower boundary of the first class. • Add the interval breadth to obtain the class boundary
Histogram
The lower boundaries of the first class can be either 2.5000 or 2.5005 (it has to be less than the smallest value 2.502). Therefore, 2.5005 + interval breadth 2.5005 + 0.005 = 2.5055 Therefore first class boundary 2.005 – 2.5055
Histogram
The second class boundary 2.5055 –2.5105 Since the class interval is 9, the last class boundary will be the 9th. Note that this has to contain the largest recorded value. Therefore, 9th class boundary 2.5405 – 2.5455
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. Mid – point of each class =
Sum of the upper & lower boundaries of each class 2
Histogram
Step 6 Obtain the frequencies Read the observed values one by one & record the frequencies Falling in each class using tally marks, in groups of five as follows: Frequency
Frequency notation
1
/
2
//
3
///
4
////
5
////
Histogram
Class
1 2 3 4 5 6 7 8 9
2.5005 – 2.5055 2.5055 – 2.5105 2.5105 – 2.5155 2.5155 – 2.5205 2.5205 – 2.5255 2.5255 – 2.5305 2.5305 – 2.5355 2.5355 – 2.5405 2.5405 – 2.5455 Total
Midpoint Frequency marks (tally)
2.503 2.508 2.513 2.518 2.523 2.528 2.533 2.538 2.543
/ //// //// //// //// //// //// //// ////
//// //// //// //// //// //// // //// //// //// //// /
Frequency
1 4 9 14 22 19 10 5 6
90
Histogram
How to draw a histogram
Step 1 • Mark the horizontal axis with a scale. • The scale need not be on the base of the class interval.
• A unit of measurement of data can be used. •In the current example we can take 0.01mm of diameter = 10mm on the histogram scale. •Leave a space equal to the class interval on the horizontal axis on each side of the scale.
Histogram
Step 2 Mark the left – hand vertical axis with a frequency scale.
Step 3 Draw the bar chart as per the data in the frequency table.
Step 4 Draw a line on the histogram to represent the mean, & also draw a line representing the specification limit, if any.
Histogram
HISTOGRAM No. of occurances
25
X = 2.5247
20 15 10 5 0
2.50
2.51
2.52
2.53
Diameter in mm
2.54
2.55
Histogram
Application To analyze processes and discover items to be improved To research process capability To control the process (in a time series)
To verify effects of an improvement
Histogram
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
Histogram
Uses of 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.
Histogram
Types of distribution • The shape of the distribution gives a more clear 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 Normal distribution is commonly used type.Here the values are symmetric about the center and area gives the value of probability
Histogram - Interpretation
(ii) Positively skewed Values are more concentrated in one side nearer to origin of x line. Here most of the values lies in the lower part of the values of histogram (iii) Negatively skewed Values are more concentrated in one side far from the origin Values lies in the higher part of the values of histogram
Histogram - Interpretation
(iv) Bi modal distribution In this type of distribution, there are two peak values of frequency
(v) Multi modal distribution There are number of peak values of frequency
Stratification
Stratification
Stratification
Stratification
Stratification Stratification is the act of fine tuning the data in order to make sure of the significance of the assured factors, to the grass root level.
Stratification
Case study Problem : More No. of Accidents
Let us stratify the the data regarding the accidents
Stratification
STATISTICS
REPORTABLE ACCIDENTS
:
08
NON-REPORTABLE ACCIDENTS
:
33
NEAR MISS INCIDENTS
:
21
LOST TIME INJURIES
:
41
MANDAYS LOST
:
187
Rep.acct. for
– Operator not reporting back to duty more than 48hrs
Non-reportable acct. beyond less than 48 hrs
– Operator disablement extending the day of shift but
Lost time injury
– Reportable + Non-reportable
Stratification
ANALYSIS –REPORTABLE ACCIDENT
Total no of reportable accident :
8
Stratification ACCORDING TO CATEGORY
Contract Labour (1) 13%
Regular Employee (4) 49%
Total no.of Reportable accidents : 8
Temp.workman (3) 38%
Stratification
ACCORDING TO PHENOMENON
Adjusting/Cleaning/Loading/Unloading while M/C running
6
Wrong handling of material handling equipment
2
Hit against object
0
Hit by objects/Fallen objects
0
others
0
Fall from Height
0
Fall from Two wheeler
0
Contact with chemical
0
wrong assembly
0
0
1
2
3
4
No of Accidents
Total no.of Reportable accidents : 8
5
6
7
Stratification
ACCORDING TO BODY PARTS INJURED
Leg (2) 25%
Hand (1) 13%
Total no.of Reportable accidents : 8
Finger (5) 62%
Stratification
ACCORDING TO PLANT
Others (1) 13%
Plant-1 (0) 0% Plant-2 (3) 37%
Sp. Wh (1) 13%
R & D (1) 13% Plant-3 (2) 24%
Total no.of Reportable accidents : 8
Stratification
ACCORDING TO UNIT 1
No.of accidents
PLANT-1
0
0
0
0
0
0 FAB
ENG UNIT
Total no.of Reportable accidents : 8
PAINT
VECH UNIT
STORES
Stratification
ACCORDING TO UNIT
No.of accidents
2
PLANT- 2
1
1
1
1
0
0
0
0 FAB
ENG UNIT
Total no.of Reportable accidents : 8
PAINT
VECH UNIT
STORES
PLATING
Stratification
ACCORDING TO UNIT 3
PLANT- 3
No.of accidents
2 2
1
0
0
0
0 M/C shop Gear Shop Total no.of Reportable accidents : 8
HT/Plating
Stores
Stratification
ACCORDING TO PLANT
No.of accidents
2
Other areas
1
1
1
1
0 R&D Spares Ware house Total no.of Reportable accidents : 8
Civil
Stratification
ACCORDING TO SHIFT 5 4 No.of accidents
4 3 2
2
2 1 0
0
0 I II Total no.of Reportable accidents : 8
III
GEN
OT
Total no.of Reportable accidents : 8 A
RC
H
0 M
RY
RY
UA
NU A
BE R
BE R
1
FE BR A
JA
EM
EM
0
DE C
NO V
TO BE R
0
O C
BE R
UG US T
0
SE PT EM
A
0 0
JU LY
Y
0
JU NE
A
PR IL
M
A
No.of accidents
Stratification
ACCORDING TO MONTH
4 3
3 2
2 1 1
1 0
Stratification
ACCORDING TO FAULT
4
No.of accidents
3
3
3 2 2 1 0 OPERATORS FAULT
SUPERVISORY FAULT
Total no.of Reportable accidents : 8
SYSTEM AND ENVIRONMENT FAULT
Stratification
The data has been stratified 1. According to employee category
2. According to phenomenon 3. According to body parts injured 4. According to plant a. According to unit 5. According to shift 6. According to month 7. According to fault
Stratification
Why Stratification? Stratification helps us to Zero in on to the relevant areas for investigation. Rather than investigating all the data that is present, we can narrow down the field of investigation by stratification.
Scatter diagram
Scatter diagram
Scatter diagram
Scatter diagram
It is often essential to study the relation of two corresponding variables. For example, how does the speed of driving a two wheeler affect its fuel efficiency?
Scatter diagram
To study the relation of two variables such as the speed of the two wheeler & the fuel efficiency we can use what is called a Scatter diagram.
Scatter diagram
A scatter diagram is a type of a Graph. The X & Y axes contain the the two variables.
Speed of the two wheeler
Scatter diagram
Based on the data available, dots are marked on the graph & the distribution of the dots is observed.
.. .. ... .. Speed of the two wheeler
How to read scatter diagrams You can grasp the correlation between pairs of data just by looking at the shape of a scatter diagram.
Scatter diagram
5 examples are given
350 300 250 200 150 100 50 0
35 30 25 20 15 10 5 0 0
5
10
15
20
Positive correlation
0
100
200
300
400
Negative correlation
Scatter diagram
500
40
400
30
300
20
200
10
100
0
0
0
5
10
15
20
Positive correlation may be present
0
100
200
300
400
Negative correlation may be present
Scatter diagram
700 600 500 400 300 200 100 0 0
100
200
300
No correlation
400
Scatter diagram
The two variables we will deal with are: a) A quality – characteristic & a factor affecting it,
b) Two related quality characteristics, or c) Two factors relating to a single quality characteristic.
Let’s consider the steps in making a scatter diagram
Scatter diagram
Step 1 Collect paired data (x,y) between which you want to study the relations & arrange the data in a table. It is desirable to have at least 30 pairs of data.
Scatter diagram
Step 2 • Find the maximum & minimum values for both x & y. • Decide the scales of horizontal & vertical axes. • Make lengths of both axes become equal. • Keep the number of unit graduations between 3 to 10.
=
Scatter diagram
Step 3 Plot the data on the section paper. Step 4 Enter all necessary items. Make sure that the following items are included. a) Title of the diagram b) Time interval c) Number of pairs of data d) Title & units of each axis
Scatter diagram
Let us do an exercise! To what extent does riding speed affect braking distance? Let’s draw a Scatter diagram & find out.
Scatter diagram Data of vehicle riding speed & braking distance
Sl No.
Riding speed (km/hr)
Braking distance (feet)
1
5
2
2
10
5
3
15
8
4
20
10
5
25
13
6
30
19
7
35
23
8
40
26
9
45
30
10
50
33
11
55
37
12
60
41
13
65
47
14
70
53
15
75
59
Scatter diagram Data of vehicle riding speed & braking distance
Sl No.
Riding speed (km/hr)
Braking distance (feet)
16
80
65
17
85
70
18
90
80
19
95
86
20
100
92
21
105
101
22
110
108
23
115
117
24
120
124
25
125
131
26
130
144
27
135
151
28
140
160
29
145
167
30
150
175
Scatter diagram
Step 1
As seen in the table, we have 30 pairs of data.
Step 2
In this example, let riding speed be indicated by X (horizontal axis), & braking distance by Y (vertical axis).
Scatter diagram
We mark off • The horizontal axis in 50 (km/hr) intervals, from 0 to 200 (km/hr)
• The vertical axis in 20 (feet) intervals, from 0 to 200(feet)
Step 3 Plot the data.
Scatter diagram
200 180 160 140 120 100 80 60 40 20 0 0
50
100
150
200
Scatter diagram
Step 4 Enter the • number of samples (n = 30), • horizontal axis (Speed [km / hr]) • vertical axis (distance in feet )
• title of diagram .
Scatter diagram
Distance in feet
RIDING SPEED VS BRAKING DISTANCE 200 180 160 140 120 100 80 60 40 20 0
n=30
0
50
100 Speed in km/hr
150
200
Scatter diagram
Product:
Plastic tanks
Production method: Blow Moulding Problem: Defective tanks( Thin walls) Suspected root cause: Variation in Compressed air pressure
Scatter diagram Data of blowing air pressure & percent defective of plastic tank Date Air pressure Percent (kgf/cm2) Defective Oct-01 8.6 0.889 2 8.9 0.884 3 8.8 0.874 4 8.8 0.891 5 8.4 0.874 6 8.7 0.886 7 9.2 0.911 8 8.6 0.912 9 9.2 0.895 10 8.7 0.896 11 8.4 0.894 12 8.2 0.864 13 9.2 0.922 14 8.7 0.909 15 9.4 0.905 16 8.7 0.892 17 8.5 0.877 18 9.2 0.885 19 8.5 0.866 20 8.3 0.896 21 8.7 0.896 22 9.3 0.928 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 28 8.7 0.912 29 9.1 0.925 30 8.7 0.872
Scatter diagram
Step 1 As seen in the table, we have 30 pairs of data. Step 2
In this example, let blowing air pressure be indicated by X (horizontal axis), & percent defective by Y (vertical axis). Then,
The maximum value of X: Xmax = 9.4 (kgf/cm2) The minimum value of X : Xmin = 8.2 (kgf/cm2) The maximum value of Y: Ymax = 0.928 (%) The minimum value of Y : Ymin = 0.864 (%)
Scatter diagram
We mark off
the horizontal axis in 0.5(kgf/cm2) intervals, from 8.0 to 9.5 (kgf/cm2) and the vertical axis in0.01(%) intervals, from 0.85 to 0.93(%)
Step 3 Plot the data.
Scatter diagram
0.93 0.92 0.91 0.9 0.89 0.88 0.87 0.86 0.85 8
8.5
9
9.5
Scatter diagram Step 4 Enter the time interval of the sample number of samples (n = 30),
n=30
0.92 0.91 0.9
horizontal axis
(blowing air pressure
(Oct 1 – Oct 30)
0.93
[kgf/cm2]),
0.89 0.88
vertical axis (percent defective [%])
0.87
title of diagram .
0.86 0.85 8
8.5
9
9.5
Blowing air pressure(kgf/cm2)
Scatter diagram of Blowing Air Pressure Vs Percentage of Tanks defective
Scatter diagram
Why Scatter Diagram? As illustrated in the figure if two characteristics are correlated, & their correlation can be represented as a line or a curve: The scope of action can be determined easily.
In the diagram, Action on Sector A1 will get more significant results than on Sector A3
Scatter diagram
Let us do an exercise ! Let us say you are not satisfied with the marks that your son/daughter is scoring. So, you want him/her to study more. But does studying more really help? How to find out?
Scatter diagram
Percentage of marks
Try a scatter diagram !
90 80 70 60 50 0.5 1.0 1.5 2.0 2.5
No. of hours of study/day
There is another tool which we use as a part of all the 7 QC tools
Which is it?
Brain storming
BRAIN STORMING !
Brain storming
BRAIN STORMING Brain storming is a technique to obtain creative ideas from a group of persons in a shortest possible time on an effect.
WHY
To identify the problem - to identify the causes To find solution
- to prevent problem
Brain storming WHY BRAIN STORMING?
TO IDENTIFY THE PROBLEM
TO IDENTIFY CRITICAL CAUSES TO FIND THE SOLUTION TO PREVENT THE PROBLEM
Brain storming
BRAIN STORMING Brain storming can be conducted in two ways 1. Structured
•
Every person in a group must give an idea as their turn arises.
•
Forces even shy people to participate.
•
Creates a certain amount of pressure to contribute.
Brain storming 2. Unstructured • Group members simply give ideas as they come to mind. • Creates more relaxed atmosphere • Risks domination. Thumb rule : 5 – 15 minutes works well
Brain storming
BRAIN STORMING SESSION • Let all the members speak freely and give ideas
• Encourage wild ideas • “Quantity” rather than “Quality” ideas
• Suspend judgment on “Good” or “Bad” • Ride on another’s ideas • Never criticize other persons’ opinions
Brain storming • Never
prohibit a person from speaking • See the problem from different angles/facets • Write down all the viewpoints • List the cause/ideas
• Think of the countermeasures to eliminate the causes • Leader/facilitator needs to guide the members in generating ideas • Whenever necessary non – members can also be involved
CONCLUSION
Conclusion
Remember that the 7 QC Tools help in Problem Solving.
Problems can be solved through • Intuition / Experience • Statistical tools • Experimental research By using the QC story methodology for solving problems we are adapting a scientific approach.
Conclusion
CONCLUSION
Application of Tools QC story TOOLS Check sheet Pareto diagram Stratification Cause & effect diagram Histogram Scatter diagram Control chart,graphs DOE Test of significance Why, why analysis PM analysis
Gantt chart
Problem Observation Identification
Analysis Action
Check
Standardization
HD
Conclusion
Experience only
*
Experience & Scientific
Scientific
Medium risk
*
Less risk
* Investigative
High risk
Confirmative
Approach
Best guess