Six sigma-black-belt-project-sample

Road- Map of Project
Define
•Identify Customer and project CTQ
•Develop Team Charter
•Develop process Map
•Get approval for Project
Measure
•Select CTQ Characteristic
•Data Collection Plan (DCP)
•Conduct MSA
•Conduct Basic Data Analysis ( To see frequency of Defect)
Analyze
•Identify Potential X’s through statistical tools.
•Conduct Statistical tests to see is relationship exists between X and Y
Improve
•Screen for Vital X’s effecting Y
•Discover Variable Relationship
•Establish Operating Tolerance
Control
•Define Process capability
•Establish Control plan
Define

Define Project Charter
Business Case
Jeddah Cable company manufactures wires and cable products used for construction , energy ,signal process
,transportation, communication and oil and gas sectors, present directly in 10 distribution locations from Malaysia to U.K
with direct offices in every major city in the GCC. From the months of august and September 2012 , there is an
increase in the shorted % of drums , the shorted % for every month should be 1.3% but from past 2 months its more
than 2% which creates lot of scrap and also re-work which is a business loss for the company.
Problem Statement
Shorted Reports for the months of August and September 2012 have been reviewed, the shorted % during these 2
months is 2% which is very high the shorted % should be 1.3% if this is not achieved ,there is more of scrap and re-
work which again is a waste of time and money .Moreover the customers are also dissatisfied because of the delayed
delivery dates. This is creating business loss and needs to be addressed immediately.
Goal Statement
The main aim of this project is to decrease the shorted from 2% to 1.3% to avoid business loss and re-work time
with increased performance.

Project Charter…
In scope
Quality Department, Production Department.
Define
Out scope
New Hires From the month of October
Project Plan
Project Team
Sponsor – Hilal
Champion Meer vajithulla
MBB – Dheerendra S Negi
Mentor – Dheerendra S Negi
BB – Shaik Noor Ahmed
Target Date Actual Date
Start Date 1 Oct’12 1 oct’12
Define 7 Oct ‘12 7 Oct ‘12
Measure 14 Oct ‘12 14 Oct ‘12
Analyze 21 Oct ‘12 21 Oct ‘12
Improve 28 Oct ‘12 28 Oct ‘12
Control 5 nov’12 5 nov’12

Define Voice Of Customer (VOC)
Customer VOC CTQ
Quality manager Shorted % should be less
than 1.3% for all the months
so that quality target at the
end of the year can be
achieved.
Shorted % should be less
than 1.3% every month.
Production manager To avoid re-work and
wastage of scrap the shorted
% should be maintained
Shorted % should be less
than 1.3% every month
Customer service manager To avoid delayed delivery of
the cables and to increase
the customer satisfaction we
should avoid shorted cables.
. Shorted % should be less
than 1.3% every month

Define ARMI
Key
Stakeholders
ARMI Worksheet
Define Measure Analyze Improve Control
Hilal I I I I I
Kareem I I I I I
Sharif I & A I & A I & A I & A I & A
ramana I & M I & M I & M I & M I & M
Shaik Noor
Ahmed
R & M R & M R & M R & M R & M
Communication Plan
Information Or Activity Target Audience Information Channel Who When
Project Status Leadership E-mails Hilal BI-Weekly
Tollgate Review BB,LBB,MBB & Champion E-mails or Meetings Hilal As per Project Plan
Project Deliverables or Activities Project Team Emails, Meetings Hilal Weekly
A – Approval of teams decision outside their charter / authorities, i.e. Sponsor
R – Resource of the team, the one whose expertise may be needed on a irregular / ad hoc basis
M – Member of the team, with the authority & periphery of the charter
I – Interested party, one who needs to kept informed on the findings & directions, if in the later
phase found to be as a worthy outcome
d
Define

CTQ TreeDefine
Maintaining
shorted % as
1.3% every
month
CTQs
Improving the quality of the
cable by avoiding shorted .
(Project Y)
Decreasing the scrap during re-
work
Meeting the targets during the
year end
Avoiding re-work and machine
hours
(Project Y)
d
Define

SIPOC
Process
 Raw material
received to the
machine
 operator loads the
machine with the raw
material
 operator starts the
machine
 process is done
and the insulation is
done on the cable
 after the quantity of
Xlpe is consumed
completely the
operator slows down
the machine
 Operator stops the
machine
Outputs
o received material
o material is loaded
o Machine started
o Processing is done
o Machine is slowed
down
o Machine is stopped
Customers
Finished cable is
ready
Suppliers
•Sabic.
Inputs
 Xlpe
Labour
Machiene
Row material
d
DefineDefine

Data Collection Plan m
MEASURE
What question do you want to answer?
What is the current shorted % that we have
KPI Operational Definition Defect Def
Performance
Std
Specification Limit
Opportunity
LSL USL
shorted % of every
month should be
1.3%
Cable is shorted whenever
ther e is high voltage test
done ,this should not happen
the total shorted
% should not be
more than 1.3%
<1.3% NA 1.3 Monthly %
How would you gauge the current AHT?
Process wise & Team wise current AHT of Reach out customer
care
KPI
Data
Type
Data Items
Needed
Unit
Plan to collect Data
Plan to
sample
What Database
or Container
will be used to
record this
data?
Is this an
existing
database
or new?
If new,
When will
the
database
be ready
for use?
When is
the
planned
start date
for data
collection?
Total number of
drums produced
Continuo
us
Total drums
produced
No of
drums
Excel sheet Existing NA 14-Nov-12
Record the
total
number of
shorted
drums.
Define

Validate Measurement
System Gage R&R ANOVA Method
Gage R&R
%Contribution
Source VarComp (of VarComp)
Total Gage R&R 5.6 0.01
Repeatability 3.9 0.01
Reproducibility 1.6 0.00
Operater_1 0.3 0.00
Operater_1*Agent Name_1 1.3 0.00
Part-To-Part 42843.7 99.99
Total Variation 42849.3 100.00
Process tolerance = 320
Study Var %Study Var %Tolerance
Source StdDev (SD) (6 * SD) (%SV) (SV/Toler)
Total Gage R&R 2.361 14.17 1.14 4.43
Repeatability 1.981 11.89 0.96 3.71
Reproducibility 1.285 7.71 0.62 2.41
Operater_1 0.577 3.46 0.28 1.08
Operater_1*Agent Name_1 1.147 6.88 0.55 2.15
Part-To-Part 206.987 1241.92 99.99 388.10
Total Variation 207.001 1242.00 100.00 388.13
Number of Distinct Categories = 123
 The above test was conducted with 2 using 3 samples.
 Gage failed even 2nd time as not meeting criteria of below mentioned:-
 % Contribution of total Gage R&R is less than Part to Part
 % Tolerance of Gage R&R to that of the total variation is less than 10%
 No. of distinct categories is more than 4
Measure

Measure: Normality
11
 Normality: P value >
0.005
 Shape: Normal
 Measure of central
tendency :Data is
normal therefore we will
consider it as Mean-
83%
 Since the P value for Anderson Darling test is > 0.05, therefore this indicates that the data is Normal
 The Minimum Internal Quality has been reported as 56% & Maximum has been reported as 100%. The
Standard deviation is 7, indicating that there is more variation in the process
 The aim of the project is to improve the avg. daily Quality from 84% to 98% or above
Measure

Measure Process Stability
 According to P Value, we can see, there is no Cluster, Mixture, Oscillation & Trend.
 This data is stable.
 The outlier have been note down & would be investigating & to find out for any special reason

Measure Process Capability
 Based on Capability Sixpack:-
CPK (-0.59) it shows, our process is not doing as per the client expectation
 Process is not capable to perform as per the client expectation, therefore
required a significant improvement.

Cause & Effect Diagram (Ishikawa)
Shorted of drums happen mainly due to measurements,
material,personnel,environment ,methods and machines
Analyze

S.
N.
Project Potential
Causes
Operational Definition Data
Type
Test to be Performed
1
Shorted
%
Training
duration
Operators and Q.C inspectors should be trained on the
shorted drums.
Continuou
s
Mamm whitney
2 Shorted
% packing The drums should be solidly packed to avoid shorted
Continuou
s
Correlation Regression
3 Shorted
% traverse
A cable should be in a good traverse so that it does not
comes in contact with the drum flange
Continuou
s Mann-whitney
4 Shorted
%
Machine
loose
tension
Machine tension should be high as lots of shorted
happen due to cable is in the take up process
Continuou
s
Mann –whitney
5
Shorted
% Xlpe
Xlpe should be heated before it goes for the assembly
process
Continous Correlation Regression
6 Shorted
% Conductor
s overlap
The conductors during assembly process should not
overlap amongst others to avoid shorted
continous Correlation regression
Analyze

Hypothesis TestAnalyze
Shorted % Vs Supervisor
Mood median test for Collection
Chi-Square = 6.78 DF = 5 P = 0.238
Team Individual 95.0% CIs
Leader N<= N> Median Q3-Q1 -----+---------+---------+----
-----+-
Anish 27 30 8884 3762 (----*----------)
Biky 22 36 9434 3897 (-----*------)
Gaurav 15 17 9009 4486 (--------------*--------------)
Mohit 35 24 8263 4812 (-----*--------)
Nitin 27 20 8614 3761 (----------*------)
Pankaj 26 25 8597 3125 (------*--------)
-----+---------+---------+---------+-
8000 9000 10000 11000
Overall median = 8839
Greater than
0.05
Above Hypothesis test shows shorted and Team leaders are independent of each other so we
conclude Shorted Has No Impact on the process

Shorted Quality Score versus Trainer
Mood Median Test: Quality Score versus Trainer
Mood median test for Quality Score
Chi-Square = 219.72 DF = 7 P = 0.000
Individual 95.0% CIs
Trainer N<= N> Median Q3-Q1 ---+---------+---------+----
-----+---
Gagan 275 181 0.7900 0.0700 *
Kishore 153 291 0.8400 0.0900 *
Lalit 243 154 0.7900 0.0700 *
Praveen 242 147 0.7900 0.0700 *
Ruhani 151 288 0.8400 0.0900 (------------*
Sudesh 152 292 0.8400 0.0900 *
Varun 218 146 0.7900 0.0700 *
Vinay 257 181 0.7900 0.0700 *
---+---------+---------+---------+---
0.795 0.810 0.825 0.840
Overall median = 0.7900
Greater than
0.05
Above Hypothesis test shows shorted and Shifts are independent of each other so we conclude
Shift Has No Impact on the collection of the process

Shorted % Vs Experience [YC XC]
Regression Analysis: Quality Score versus
Experience
The regression equation is
Quality Score = 0.7099 + 0.005302 Experience
S = 0.0533682 R-Sq = 25.6% R-Sq(adj) =
25.6%
Analysis of Variance
Source DF SS MS F P
Regression 1 3.3028 3.30277 1159.61
0.000
Error 3369 9.5955 0.00285
Total 3370 12.8982

Age & Quality Score of shorted
Correlations: Quality Score, Age
Pearson correlation of Quality Score and Age = -0.029
P-Value = 0.094
The age and quality scores are plotted and there is no impact.

Analyze Box Plot for Quality
After plotting the box plot we found out that there is huge variation in shorted cables.

shorted Vs Utilization
General Linear Model: Collection versus Shifts
Factor Type Levels
Utilization fixed 246
Source DF Seq SS Adj SS Adj MS F P
Utilization 245 2394882512 2394882512 9775031 1.13
0.292
Error 58 501128672 501128672 8640150
Total 303 2896011184
S = 2939.41 R-Sq = 82.70% R-Sq(adj) = 9.60%
Greater than
0.05
Above Hypothesis test shows shorted and Utilization are independent of each other so we
conclude Utilization Has No Impact on the shorted of the process

shorted Vs Accounts Worked
General Linear Model: Collection versus Work
Factor Type Levels
Work fixed 23
Analysis of Variance for Collection, using Adjusted SS for Tests
Source DF Seq SS Adj SS Adj MS F P
Work 229 2420959614 2420959614 10571876 1.65 0.32
Error 74 475051570 475051570 6419616
Total 303 2896011184
S = 2533.70 R-Sq = 13.60% R-Sq(adj) = 32.83%
Greater than
0.05
Above Hypothesis test shows shorted and Accounts worked are not dependent of each other so
we conclude Accounts worked do not have significant Impact on the shorted of the process

shorted Vs Average Payment Size (APS)
Correlations: Collection, Average Payment Size
Pearson correlation of Collection and Average Payment Size = -0.003
P-Value = 0.961
Regression Analysis: Collection versus Average Payment Size
The regression equation is
Collection = 9194 - 0.14 Average Payment Size
Predictor Coef SE Coef T P
Constant 9193.5 502.5 18.30 0.000
Average Payment Size -0.144 2.937 -0.05 0.961
S = 3096.67 R-Sq = 0.0% R-Sq(adj) = 0.0%
Analysis of Variance
Source DF SS MS F P
Regression 1 23115 23115 0.00 0.961
Residual Error 302 2895988068 9589364
Total 303 2896011184
Above Hypothesis test shows shorted and APS are independent of each other so we conclude
APS do not have significant Impact on the shorted of the process
P- Value for
both correlation
and
Regression is
greater than
0.05

shorted Vs Gender
Chi-Square = 1.44 DF = 1 P = 0.230
Gender N<= N> Median Q3-Q1 -------+---------+---------+-------
--
Female 23 31 9240 3995 (------*--------------------)
Male 129 121 8686 3863 (-----*----)
-------+---------+---------+---------
8800 9600 10400
Greater than
0.05
Above Hypothesis test shows shorted and Gender are independent of each other so we
conclude that gender do not have significant Impact on the shorted of the process

shorted Vs Marital Status
Chi-Square = 0.37 DF = 1 P = 0.543
Marital Individual 95.0% CIs
Status N<= N> Median Q3-Q1 ----+---------+---------+---------
+--
Married 48 53 8912 3338 (----------------*----------------)
Unmarried 104 99 8686 4173 (----------*----------)
----+---------+---------+---------+--
8400 8800 9200 9600
Greater than
0.05
Above Hypothesis test shows shorted and Marital Statusare independent of each other so we
conclude that gender do not have significant Impact on the shorted of the process

shorted Vs Education
Chi-Square = 0.03 DF = 2 P = 0.987
Education N<= N> Median Q3-Q1 -------+---------+---------+--
-------
graduate 114 113 8820 3846 (-------*------)
postgraduate 15 15 8656 3308 (----------*--------------------)
Undegraduate 23 24 8866 3722 (-----------*----------------)
-------+---------+---------+---------
8400 9100 9800
Greater than
0.05
Above Hypothesis test shows shorted and Education are independent of each other so we
conclude that Education do not have significant Impact on the shorted of the process

shorted Vs Experience
Chi-Square = 3.26 DF = 3 P = 0.353
Experience N<= N> Median Q3-Q1 --------+---------+---------+--
------
0 to 1 43 43 8827 3847 (--------*--------)
1 to 2 46 56 9222 4649 (---------------*------------)
2 to 3 49 36 8259 3654 (-----*----------)
3 and above 14 17 9165 3537 (--------------*---------------)
--------+---------+---------+--------
8400 9100 9800
Greater than
0.05
Above Hypothesis test shows shorted and Experience are independent of each other so we
conclude that Experience do not have significant Impact on the shorted of the process

shorted Vs Experience Type
Chi-Square = 129.87 DF = 2 P = 0.000
Previous Individual 95.0% CIs
experience N<= N> Median Q3-Q1 -+---------+---------+--------
-+-----
C 53 85 9275 3473 (*-)
CS 13 67 11174 3222 (--*---)
obs 86 0 6125 2419 (-*-)
-+---------+---------+---------+-----
6000 8000 10000 12000
Smaller than
0.05
Above Hypothesis test shows shorted and Experience Type are dependent of each other so we
conclude that Experience type do have significant Impact on the shorted of the process

shorted Vs Tenure
Mood Median Test: Collection versus Tenure
Chi-Square = 52.39 DF = 2 P = 0.000
Tenure N<= N> Median Q3-Q1 ------+---------+---------+----
-----+
0 to 1 years 111 48 7747 3493 (--*---)
2-3 years 25 61 9675 3369 (----*--------)
3 yrs + 16 43 10248 3908 (-----*-------)
------+---------+---------+---------+
8000 9000 10000 11000
Smaller than
0.05
Above Hypothesis test shows shorted and Tenure are dependent of each other so we conclude
that Tenure do have significant Impact on the shorted of the process

Box Plot Demonstration of shorted
and trainer
Analyze
After plotting boxplot for trainers we found out that sudesh,ruhani,and kishore have huge
variation compared to others.

S No. Vital Cause Data Type Test performed P- Value
1 supervisors discrete Moods median test 0.238
2 Age of operator discrete
Pearson
correalation
0094
3 experience discrete Moods Median 0.006
4 education discrete Moods median 0.987
4 Marital status discrete Moods median 0
Results of validation of potential X’s
impacting AHT
Analyze

Improve Vital X’s effecting our Y (collection)
Low Cost
High Cost
HighImpact
LowImpact
•Tenure
•Experience Background
•Team Leader
•Process Knowledge
•Utilization
•Accounts Worked
•RPC
•APS
•Gender
•Marital Status
•Education
Training of operators- Training
of operators can be provided as it does
not cost anything , the managers and
engineers can easily train them .
•Shift
•Experience
Machine issues-
machines can be costly
because it involves lots of
money to change them.

Quality Function DeploymentImprove
Priority
Priorityweightage
Implementrecreationalactivitycalendar
Implementinhouseskillimprovement
trainingcalendar
Hiringteamtoscreenexperiencedprofilesfor
hiring
Implement>1yearexperienceincallingfor
telecomprocessintheJobdescription
Evaluatetheremunerationgridwith
leadershipinvolvingHR,Compensationand
Benefit&Recruitmentteam
Implemettherevisedgrid
Monitorweeklytypingspeedoftheassociates
hiremoreexperiencedengineers
Conductmonthlytypingassesment
IdentifyconsistentTypingtestfailures&
recommendprocessimprovementplan
RefertobenchpostnoresultinPIP
ConductsessionwithTraining,HR&
Recruitmentteam
Implementchangesbasedontheprocess
requirement&floatthenewapprovedversion
Implementchangestothehiringassesment
thresholdsbasedontheprocessrequirement&
floatthenewapprovedversion
Conductassesmentforexistingtrainersto
qualifyforgivingtraining
Implementchangestotheculture&pre
processtrainingassesmentthresholdsbasedon
theprocessrequirement
Changestotheculture&preprocesstraining
assesmentthresholdstobeapprovedfrom
Clientbeforeimplementation
LinktheHalfyearlyassementscoretothe
associateGoalform
Conductprocesstest
IdentifyscenariobasedtrainingneedfromAH
Tperspective
Conductprocesstrainingrefresherforbottom
performers
Evaluaterefreshertrainingmonthlyscores
Recommendallassementfailuresforprocess
improvementplan
RefertobenchpostnoresultinPIP
Linkthequaterlyassementscoretothe
associateGoalform
TotalPriorityWeightage
Totalpriorityscore
Introduce technical
training program for the
operators
4 4 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 8
Experienced people to be
hired by recruitment team
4 0 0 4 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 8
Offer good package to
experienced candidates as
per the market standards
4 0 0 0 0 4 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 8
maintenance team should
have experienced
engineers so that there
should not be any machine
break downs
5 0 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 5
evaluate the operator on
monthly basis
5 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 25 5

Failure Mode Effect AnalysisImprove
• Please find attached FMEA for
JEDDAH

BOM Comparison Post ImprovementImprove
Variable N N* Mean SE Mean StDev Minimum Q1 Median
Old BOM 304 0 3858.8 75.8 1322.3 342.9 2894.8 3729.5 4684.7
New BOM 304 0 4804.4 89.8 1565.9 422.3 3648.6 4692.6 5832.8
This Box-Plot
Demonstration clearly
shows Our new BOM
performance is
significantly better than
old BOM

Shorted Comparison Post
Improvement
Improve
Mann-Whitney Test and CI: Old Collection,
New Collection
N Median
Old Collection 304 8839
New Collection 304 12464
Point estimate for ETA1-ETA2 is -3994
95.0 Percent CI for ETA1-ETA2 is (-4853,-3191)
W = 69925.0
Test of ETA1 = ETA2 vs ETA1 < ETA2 is significant at 0.0000
Mann Whitney test has shown that there has been significant improvement in the median of the
new shorted
P-Value is
less than 0.05

shorted Comparison Post
Improvement
Improve
Descriptive Statistics: Old shorted , New shorted
Variable Mean TrMean StDev Minimum Q1 Median Q3 Maximum
Old shorted 9170 9114 3092 745 7301 8839 11130 19083
New shorted 16689 15250 12032 1183 9283 12464 19333 84275
This Box-Plot
Demonstration clearly
shows Our new shorted
performance is
significantly better than
old shorted

Control Chart of Training post
Improvement
Control
This Control chart shows our TOS process is under control with improvement i.e, we have sustained this
improvement now.

Control Chart of BOM post
Improvement
Improve
Above Control Chart shows our BOM process is under statistical control. That means we have
sustained our improvement in BOM

Shorted Comparison Post
Improvement
Improve
This Control chart clearly shows that our shorted process is working under control. That is we have
sustained the improvement that we achieved in our collection

Mean has been shifted from 682 to 519 sec. It means significant improvement has been done. 42
I
Improve

Shorted decreasedCONTROL
We have statistically proved that our shorted has
decrease and our process in under statistical Control

Six sigma-black-belt-project-sample

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Six sigma-black-belt-project-sample (20)

Recently uploaded (20)

Six sigma-black-belt-project-sample