Six Sigma Foundations Continuous Improvement Training Six Sigma Simplicity Key Learning Points Data Needs to be Verifiable Reliable Correct Stakeholders Believe Data AGENDA What is Measurement Systems Analysis ID: 593898
Download Presentation The PPT/PDF document "Measurement Systems Analysis" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Measurement Systems Analysis
Six Sigma FoundationsContinuous Improvement Training
Six Sigma SimplicitySlide2
Key Learning Points
Data Needs to be:Verifiable
Reliable
Correct
Stakeholders Believe Data Slide3
AGENDA
What is Measurement Systems Analysis
?
Reproducibility and Repeatability
SummarySlide4
What is measurement systems analysis
Whatever it takes to ensure that measurements are reliable and credibleAnticipate ‘doubting Thomas’ and share the pre-checks on your measurementsSlide5
Measurement Systems Analysis– Terms
Repeatability
Chance that same
person (operator)
and process will give consistent measurement
Reproducibility
Chance that different
person (operator)
will give consistent measurement
Attribute
Qualitative e.g. Pass/Fail, Hot/Cold, Employee name, work order number
Variable
Quantitative Numbers, % Pass, Cost, Cycle TimeSlide6
MSA - Who should create it
Process owner/team guided by Black or Green Belt Quality group
May be part of Quality Management System (QMS)Slide7
Exercise
12 months ago you agreed to the installation of a $1Million MRP software package with the aim of improving on time delivery
Now a sister plant is looking at introducing a different package to solve the same problem
The reason that they give for using a different package is because they say that your attempt to improve delivery failed
They quote the data shown on the next graph.Slide8
MRP system ‘improvement’Slide9
MRP System An Answer
Comparing the data from before and after there is no relationship between the month of the year and the on time deliverySlide10
Before the MRP system there was more spread than afterwards
After the system introduction the average on time delivery fellBut difference is small 1.2/88 (delta/sigma) and we only had monthly figures!
Because of our measurement system we would be unwise to assume that the observed sample average (before and after) is an accurate prediction of future on time deliverySlide11
Sample Rates
It is unsafe to say whether the project worked or not because the
project
to introduce the system was poorly set up!!
In future the Green belt should check
The source of the data
The speed and size of sampling
The rate of changes you introduce elsewhere which may have a
domino effect
The delta sigma of improvement you want to seeSlide12
Noise
Was on-time delivery the right thing to try and change?If ‘
it depends’
on many factors such as:
order clauses, customers, capacity, changing volumes of orders, seasonal delivery demands and backlog of orders…this is called
noise
In future, check …
That the metric is as free of noise as possible
That there is not a less noisy measure - such as % of orders delivered within 8 weeks, or RTY through a bottleneck (such as test)Slide13
Linearity
A measurement is linear if there is a straight line relationship between the observed or measured value and the actual valueSlide14
Impact of Linearity
Is 95% on time delivery only 5% better than 90%?If lead time is 50% do sales go up by 50%
Is it necessary to train 100% of employees or just 50%
Do all the engineers need a computer?
In each case we will be more in control if we fully understand the relationship between
Y and X
and how the magnitude of X or Y effects our decisionsSlide15
Repeatability and Reliability
T
o determine the accuracy of our measurement system and the limits on the usefulness of the measurement deviceSlide16
The necessity of training farmhands for first-class farms in the fatherly handling of farm livestock is foremost in the eyes of farm owners. Since the forefathers of the farm owners trained the farmhands for first-class farms in the fatherly handling of farm livestock, the farm owners feel they should carry on with the family tradition of training farmhands of first class farmers in the fatherly handling of farm livestock because they believe it is the basis of good fundamental farm management.
Task: You have 60 seconds to document the number of times the 6th letter of the alphabet appears in the following text.
Repeatability Exercise
Number:
____Slide17
You receive a complaint from a salesman that the on time delivery of your plant is low. The last 20 units shipped to a customer were late.
The plant quotes 100% on time. Every departmental manager is 100% on time according to their schedule
How can this be?
Repeatability - DiscussionSlide18
R&R - Why do we need it
To identify how much the chosen measuring system limits our ability to improve
Because measurement systems are often significantly flawed
To guarantee the truth of improvements madeSlide19
How to do R&R
Select a minimum of 30 parts/documents from the process and at least 2 operators
Half should have defects some of which are marginally defective, half defect free
Each operator examines each piece at least twice in a random order (ensure understanding of type of data being collected - Variable or Attribute)
Plot measurement against operator, run, part and standard
If R&R are not acceptable adjust process and repeatSlide20
Attribute ExampleSlide21
Attribute Analysis
Agrees with himself 14/14, 11/14, 14/14
Agrees with standard 26/28, 21/28, 22/28
Overall agree with standard = 26+21+22/(3*28)Slide22
Presenting Results
Operator 2 was inconsistent
No operator was perfectSlide23
Presenting Results
Overall accuracy of measurement 82%What is it about some parts that we can not measure them correctly?Slide24
Using computer software
Some of the initial conclusions that you can make are limited
If you want more precise information the team should seek advice from a specialistSlide25
Attribute R&R - Graphical outputSlide26
3
2
1
1
1
1
0
0
9
0
8
0
7
0
6
0
5
0
4
Oper
ID
By Operator
B
u
f
a
l
o
,
N
Y
P
l
a
n
t
1
5
m
m
S
x
S
g
m
a
B
B
0
1
0
1
/
1
9
9
8
G
a
g
e
#
0
2
0
3
7
1
M
s
c
:
T
o
l
e
a
n
c
e
:
R
e
p
o
r
t
e
d
b
y
:
D
a
e
o
f
u
d
y
:
G
a
g
e
n
a
m
e
:
Gage R & R (ANOVA) for Measure
1
0
9
8
7
6
5
4
3
2
1
1
1
1
0
0
9
0
8
0
7
0
6
0
5
0
4
Part ID
By Part
B
u
f
f
a
l
o
,
N
Y
P
l
a
n
t
1
5
m
m
S
i
x
S
i
g
m
a
B
B
0
1
/
0
1
/
1
9
9
8
G
a
g
e
#
0
2
0
3
7
1
M
i
s
c
:
T
o
l
e
r
a
n
c
e
:
R
e
p
o
r
t
e
d
b
y
:
D
a
t
e
o
f
s
t
u
d
y
:
G
a
g
e
n
a
m
e
:
Gage R & R (ANOVA) for Measure
Variable R&R - Graphical Analysis
In this example 3 operators measured twice the diameter of 10 turned components. What would the ideal graph look like for each? Slide27
R&R Class Exercise
Now its your turn
Grab a bag of
m&m’sSlide28
Measurement Systems Analysis
Six Sigma FoundationsContinuous Improvement Training