Project Name (DMAIC)

Courier Service Efficiency Improvement at BCE

Date: 15/06/2020

Purpose:
To improve the delivery time of
courier for BCE
Your Name Here- Rajesh Kamath
Black Belt –Dr Tansen Choudhari
Master Black Belt
Team Member- Nil
 Six Sigma Project Summary
Project Title: Courier
CourierService
Serviceefficiency
efficiencyimprovement
improvement

Rajesh Project Function Finance

Project leader: RajeshKamath
Kamath Finance
Date: /17/6/2020 Project Number: 01

Why do this Project? Team Involved:

To reduce courier delivery time Amol Mane

from average 38 hrs to 30 hrs Rukmangat Devkate

Defect Definition? Project Timelines:

Late delivery time of courier 15/6/2020 -20/6/2020
Wrong courier
Misplaced courier
Wrong weighment of envelopes
Project CTQ’s: Risk Plan: Project Benefits:

Delivery better than competitor.

Courier delivery time
Reduced process cost
Delivery accuracy
Process cost -Customer
receipt to delivery
Weighment accuracy
Rough order of magnitude of savings : $
 Define

Why
Why do
do this
this project?
project?

Describe the issue, the issues it is causing and your goals.

Problem, Goal and Statement:

What is the problem? What is the measurable goal?

Courier delivery time is higher than competetors.

Delivery time is the measurable goal
Business Case:
Why are you doing the project? What are the benefits?
Reduce courier delivery time to stay in business
Potential for additional business
Fewer customer grievances wrt courier
Process cost reduction
 Define

Who
Whoisisthe
theCustomer?
Customer? External customer

What
Whatare
areCustomer
CustomerCTQs?
CTQs? • Identify Customer and prioritize Customer needs using
QFDs, Surveys etc…and translate into CTQs
• Delivery time of courier
• Process cost
• Weighment accuracy
•No misplaced courier
•No wrong courier
 Measure

CTQ Flowdown

BCE QFD
 Measure

Defect
DefectDefinition
Definition

What is being measured:

The courier delivery time –Continous data
Envelope weight –Continous data
Defect:
Defect Courier delivery > 30 days
Unit: Zonal/Region/Corporate
Opportunity: Activity /Transport/Weighment
 COPIS

BCE SIPOC
 BCE COURIER FLOW VSM

T1

3 4/5
1 T2
2 T3 T4
Regional Corporate
Zonal Office Area office
office process office process
Process Process
T5

Customer Zonal Office Area office Region

process office process
Process
T6
T7 6
T8
8 7

5.33
11.98 13.20 3.03 4.04

T1/T7 0.087 T2/T6 0.175 T3/T5 0.294 T4 0.389 T8 0.081

5
1 2 and 8 3 and 7 4 and 6

VA/NVA =1.026/37.58 =0.027

Existing Analyze
ExistingProcess
ProcessCapability
Capability

Baseline
Improve
C1 – Total delivery time

Due to outlier
process is not
reflected as
Normal.
Hence needs to
be re plotted
with outlier
removed which
may be a
special cause of
error.
With outlier removed the delivery
time reflects to be normal
distribution with mean as 37.59
 Gage R&R Study - ANOVA Method

Gage R&R for C2

Gage name: Weighing Machine
Date of study: 19-06-2020
Reported by: Amol Mane
Tolerance:
Misc:

Two-Way ANOVA Table With Interaction

Source DF SS MS F P
C1 2 1735703 867851 35705.9 0.000
C3 2 53 26 1.1 0.420
C1 * C3 4 97 24 0.4 0.817
Repeatability 9 575 64
Total 17 1736428
α to remove interaction term = 0.05
 Gage R&R Study - ANOVA Method

Two-Way ANOVA Table Without Interaction

Source DF SS MS F P
C1 2 1735703 867851 16783.2 0.000
C3 2 53 26 0.5 0.612
Repeatability 13 672 52
Total 17 1736428

Gage R&R

Variance Components
%Contribution
Source VarComp (of VarComp)
Total Gage R&R 52 0.04
Repeatability 52 0.04
Reproducibility 0 0.00
C3 0 0.00
Part-To-Part 144633 99.96
Total Variation 144685 100.00
 Gage R&R Study - ANOVA Method

Gage Evaluation
Study Var %Study Var
Source StdDev (SD) (6 × SD) (%SV)
Total Gage R&R 7.191 43.15 1.89
Repeatability 7.191 43.15 1.89
Reproducibility 0.000 0.00 0.00
C3 0.000 0.00 0.00
Part-To-Part 380.307 2281.84 99.98
Total Variation 380.375 2282.25 100.00

Number of Distinct Categories = 74

As guage R & R < 10
% measurment
system is acceptable
Weighing Machine
Reported by: Amol Mane
Gage name: Weighing Machine Tolerance:
Date of study: 19-06-2020 Misc:

Components of Variation C2 by C1
100 % Contribution 800
% Study Var
Percent

50 400

0
0 1 2 3
Gage R&R Repeat Reprod Part-to-Part
C1
R Chart by C3
prasad Satish Shyam C2 by C3
UCL=32.67 800
Sample Range

30

15 _
R=10 400

0 LCL=0
1 2 3 1 2 3 1 2 3
0
C1 prasad Satish Shyam
C3
Xbar Chart by C3
800
prasad Satish Shyam
C1 * C3 Interaction
800
Sample Mean

C3
prasad
400 __
Average

Satish
UCL=342.7
X=323.9
LCL=305.1 Shyam
400

0
1 2 3 1 2 3 1 2 3
0
C1 1 2 3
C1
 Analyze

Fishbone
Fishbone Indirect
Indirect5s
5s
Analysis/7W
Analysis/7W
Measurements Materials Men & Women

Possible factor (X)

Vital Y (i.e
VTW, Cycle, Cost,
Vital X’s identified as: Defects etc..)
• xx
• xx
• xx Vital X

Identify your Vital Xs- may also

Mother Methods Machines want to use Paretos, Boxplots etc…
Nature
 Improve

New
NewProcess
ProcessMap
Map// Kaizen
Kaizen
Future
FutureVSM
VSM Closures
Closures
P r e d e fin e d
P ro c e s s
If B (w h e n y o u d o
n o t w a n t to
lis t a ll o f it )
chnage D ocum ent I n s p e c t io n o r
S ta rt of control
P ro c e s s S te p
D e c is io n
S T o ra g e

If A P ro c e s s S te p

Chnage I n s p e c t io n o r
End fo c o n tro l P ro c e s s S te p Yes P ro c e s s S te p P ro c e s s S te p
D e c is io n

No

Map out a new process

 BCE COURIER FLOW VSM –AFTER
IMPROVEMENT
T1

3 4/5
1 T2
2 T3 T4
Regional Corporate
Zonal Office Area office
office process office process
Process Process
T5

Customer Zonal Office Area office Region

process office process
Process
T6
T7 6
T8
8 7

3.974
7.56 8.8 2.68 1.96

T1/T7 0.087 T2/T6 0.175 T3/T5 0.294 T4 0.389 T8 0.081

5
1 2 and 8 3 and 7 4 and 6

VA/NVA =1.026/24.97=0.041
 Improve
Baseline: DPMO & XX s
New
NewProcess
ProcessCapability
Capability Improvement: DPMO & XX s
Xx% Improvement

Baseline
Improve
After Improvement Baseline

Avg delivery time improved from 37.5 days to

26 days
 Old
OldVs
VsNew
NewComparison
Comparison Improve

Boxplots of Baseline and IMp

(means are indicated by solid circles)

200

100

Two sample T for Baseline vs IMp

0

to
N Mean lot
StDev SE Mean
-100
Baseline 104 l
23.6it
36.7y P 3.6
i
bab6.11 0.64
IMp 91 o
4.77
r
d P ge
Baseline IMp

n
, a ha n
e stCI
V T afor
95%
c l c mu Baseline - mu IMp: ( 11.6, 26.09)
O T-Test i
t mu Baseline = mu IMp (vs not =): T = 5.16 P = 0.0000 DF = 109
s t , H statis
Normal Probability Plot for Stacked By factor -te
l e T s h ow Homogeneity of Variance Test for Stacked
p
s am
o Baseline 95% Confidence Intervals for Sigmas Factor Levels

99
Tw Improve
1

F-Test
95 2
90 Test Statistic: 36.018
0 5 10 15 20 25 30 35 40 45
80
P-Value : 0.000
Percent

70
60
50
40
30
20
10 1
5 Levene's Test
1 Test Statistic: 31.140
2
P-Value : 0.000

-100 0 100 200 -100 0 100 200

Data Stacked
 Control

Controls
Controls

• What are your controls?

• How are you going to make sure the issue you fixed stays fixed?

Control Procedure Metric Owner

 Control

Savings
Savings

• What are the savings

• Make sure you can show where savings come from
 Appendix
Appendix––Helpful
HelpfulToolkits
Toolkits

VOC-CTQ Translation

Customer
Customer Voice of the Customer Key Issue(s) Requirement
(CTQ's)

Clarify Data Develop Data Collection Plan

Segmentation Collection Plan
Data When
Measure Type Operational Definition Factors Who From To Quantity
Define Process Map

COPIS

Customer Output Process Input Supplier

•End user •Application •Application

•GCF
•PO form form
•End user
•OM •List •List
•Others
•Others •Others •Others

Y
Start Step Step Selection Step

Step End Step

To process map, please enter your actual work step to supersede words
To process map, please enter your actual work step to supersede words
marked in red and add steps according to what you need. The attached file
marked in red and add steps according to what you need. The attached file
will give you more info about process map, please click it.
will give you more info about process map, please click it.

More info process map

Product Report
Report 7: Product Performance

Characteristic Defs Units Opps TotOpps DPU DPO PPM ZShift ZBench

1 1 3670 1 3670 0.000 0.000272 272 1.500 4.958 Process Capability ： Z ＝ XXσ
2 2 6422 1 6422 0.000 0.000311 311 1.500 4.921

Use Minitab Six Sigma Product report to do this

Use Minitab Six Sigma Product report to do this
3 7 5994 1 5994 0.001 0.001168 1168 1.500 4.544
analysis. We can know the current process
analysis. We can know the current process
capability. Please click the attached file to help Guide for using Minitab
capability. Please click the attached file to help to do this analy
4 7 4665 1 4665 0.002 0.001501 1501 1.500 4.468
you finish this work.
you finish this work.

Total 17 20751 0.000819 819 1.500 4.649

E ffective C om municatio n

GE S ourcing Flexibility
H R S G S tandardization

S up plier Flexibility
Low H R S G C o st
How's to Satisfy Customer Wants

S ingle S u pplier

S hort C ycle
QFD

S S S S S S S
E E E E E E E
Customer Wants RATE

HRSG Competitiveness - product, price, value 3 1 9 1 1 3 1 9

On Time Delivery of HRSG 9 1 1 3 3 3 9 9
Supplier Responsiveness/Communication 3 9 9 1 9 3 3
Product Quality 9 9 3 3 3 3 3 3
Sum 24 20 13 16 8 18 16
Weighted Sum 144 72 99 72 48 90 108
 Pareto
By Pareto we are easily to know which defect is our main type of
By Pareto we are easily to know which defect is our main type of
variance defects or which reason is main reason. Then we can focus
variance defects or which reason is main reason. Then we can focus
on it/them to improve it. Please click the attached file to learn more
Pareto Chart for Defect Types on it/them to improve it. Please click the attached file to learn more
info.
info.

100
150
80

Percent
100 60
Count

40 Pareto tool
50
20

0 0 Please click the attached file to help you do FMEA

Please click the attached file to help you do FMEA
t s
me un s
Defect Na A mo Te
l
Ad
dre

Count 75 56 24 12
Percent 44.9 33.5 14.4 7.2
Cum % 44.9 78.4 92.8 100.0

FMEA Tool FMEA Tool

FMEA

Pilot Plan
Occurrence

Detection
Severity

Potential Responsibility

RPN
Item/ Process Potential Failure Potential Effect(s) of Cause(s) of Current Recommended and Target Date
Step Mode Failure Failure Controls Actions Completion
 Individual difference test

For discrete data ( Data types are Defects, Units, Opportunities.)

For discrete data ( Data types are Defects, Units, Opportunities.)

Chi-Square test

P – value > 0.05

No significantly statistical difference among 4 operators.

About P – value:
About P – value:
What we only need to know is when P – value > 0.05, it tell us all
What we only need to know is when P – value > 0.05, it tell us all
operators have the same level. Otherwise, when P – value <0.05, we
operators have the same level. Otherwise, when P – value <0.05, we
know these operators have different level on accuracy.
know these operators have different level on accuracy.

About Chi-Square test, please click the attached file to learn how to do
About Chi-Square test, please click the attached file to learn how to do
it.
it.

Why we do this test? We want to know whether there is someone has a

Why we do this test? We want to know whether there is someone has a
different level with others. If have, we can search some specific ways to
different level with others. If have, we can search some specific ways to
help him improve accuracy. If no, low accuracy is popular problem,
help him improve accuracy. If no, low accuracy is popular problem,
we need to search other solutions.
we need to search other solutions.
Chi-Square tool
 Individual difference test

For continuous data ( Data types are percentage of every day…)

For continuous data ( Data types are percentage of every day…)

Normality test Mood’s median test

Normal Probability Plot

.999
.99
.95
Probability

.80
.50
.20
.05
.01
.001

0.85 0.90 0.95 1.00

accuracy
Average: 0.98173 Anderson-Darling Normality Test
StDev: 0.0401611 A-Squared: 7.464
N: 30 P-Value: 0.000
About P – value:
P – value < 0.05 About P – value:
What we only need to know is when P – value > 0.05, it tell us all operators have the same
What we only need to know is when P – value > 0.05, it tell us all operators have the same
These data are not normal distribution level on median of accuracy performance. Otherwise, when P – value <0.05, we know
level on median of accuracy performance. Otherwise, when P – value <0.05, we know
these operators have different level on median of accuracy performance.
About P – value in normality test: these operators have different level on median of accuracy performance.
About P – value in normality test:
What we only need to know is when P – value > 0.05, it tell us these
What we only need to know is when P – value > 0.05, it tell us these
percentage data are normal distribution. Otherwise, when P – value <
percentage data are normal distribution. Otherwise, when P – value <
0.05, they are not normal distribution P – value > 0.05
0.05, they are not normal distribution
No significantly statistical difference on median among 3 operators.
About Normality test, please click the attached file to learn how to do
About Normality test, please click the attached file to learn how to do
it.
it.
About Mood’s median test, please click the attached file to learn how to do it.
About Mood’s median test, please click the attached file to learn how to do it.

Normality Test tool

Mood's Median Test tool
 Individual difference test

For continuous data ( Data types are percentage of every day…)

For continuous data ( Data types are percentage of every day…)

Homogeneity test
Homogeneity of Variance Test for accuracy
P – value > 0.05
95% Confidence Intervals for Sigmas Factor Levels No significantly statistical difference on variance among 3 operators.
a

Bartlett's Test
About P – value:
Test Statistic: 43.297 About P – value:
P-Value : 0.000 What we only need to know is when P – value > 0.05, it tell us all operators have the same
What we only need to know is when P – value > 0.05, it tell us all operators have the same
level on variance of accuracy performance. Otherwise, when P – value <0.05, we know
level on variance of accuracy performance. Otherwise, when P – value <0.05, we know
these operators have different level on variance of accuracy performance.
b these operators have different level on variance of accuracy performance.

Levene's Test

Test Statistic: 2.408 About two P – value:

P-Value : 0.109
About two P – value:
The first is for normal data.
The first is for normal data.
c The Second is for non-normal data.
The Second is for non-normal data.
0.00 0.05 0.10

About homogeneity test, please click the attached file to learn how to do it.
About homogeneity test, please click the attached file to learn how to do it.

Homogeneity Test tool

Why we do these tests? We want to know whether there is someone
Why we do these tests? We want to know whether there is someone
has a different level with others. If have, we can search some
has a different level with others. If have, we can search some
specific ways to help him improve accuracy. If no, low accuracy is
specific ways to help him improve accuracy. If no, low accuracy is
popular problem, we need to search other solutions.
popular problem, we need to search other solutions.
Control Plan

Vital Few Xs Solutions Responsible Person Timeline Control plan is similar to pilot plan,
Control plan is similar to pilot plan,
remember what we will do, how to do
•XX •XX •XX •XX remember what we will do, how to do
it, who will go to do it and when it
it, who will go to do it and when it
should be finished.
•XX •XX •XX •XX should be finished.

•XX •XX •XX •XX

Control Chart

P Chart for Defect

Control chart help you monitor process. We can know how well our process
Control chart help you monitor process. We can know how well our process
goes. Please click the attached file to learn more info about P chart.
3.0SL=0.001148 goes. Please click the attached file to learn more info about P chart.
0.0010
Proportion

P=7.09E-04

0.0005

-3.0SL=2.70E-04

P Chart
0.0000
1 2 3 4 5
Sample Number

You need monitor several cycles of process operation to confirm whether your
You need monitor several cycles of process operation to confirm whether your
improvement is sustained, and then you can closed your GB project for
improvement is sustained, and then you can closed your GB project for
Accuracy.
Accuracy.