Six Sigma Black Belt Project on

reduction in breakage in biscuits

LOCATION : NOIDA
PROCESS : BISCUITS MAKING
BLACK BELT : Dipti Nayak
 Understanding VOC

Customer Customer Comments Customer (CTQ's)

“During the last 8 months, rejection in biscuits due to Reduce rejection from 4.25% to 2.0 % in
breakage is around 297 kg against the 7064 kg of biscuits sweet biscuits
produced on monthly basis. Average of 8 months rejection
Champion – General Manager breakage is around 4.25%, rejection should be around 2.0%
of total biscuits processed in kgs”

Operations Manager “ Due to high rejection rate of 4.25%, target of productivity Increase the productivity from 95.75% to
is not met which is set at 98% 98%
 d
Project Charter Define

Project Leader: Dipti Nayak Team Members

Stakeholders Business Leader

Champion Vice President

Sponsor General Manager

Business Case:
ABC Limited is one the leading biscuit manufacturing company with its MBB Pranay Kumar
location worldwide. Its Manufacturing facility in Noida, India is
manufacturing sweet and namkeen biscuits for market leading brands,
Noida unit is facing lot of rejection due to breakage in biscuits due to
which the scrap /rejection has increased and needs immediate attention
as it is causing loss to the A-G biscuits. Currently it is operating under BB Dipti Nayak
4.25% rejection against the estimated target of 2.0%. Sweet biscuits
monthly production is around 7064 kg and around 297 Kg are rejected to
due breakage.

Operations Manager, supervisor, Key operators,

Team Member
quality executives, Engineering manager

Problem Statement: Goal Statement:

During the period from Mar’2012 to Oct’2012 the average rejection rate To reduce rejection due to breakage in sweet biscuits at Noida location from
is 4.25% of the sweet biscuits manufactured 4.25% to 2.0% by Mar’2013

Timelines/Milestones
Start Date End Date
/Phases
Project In Scope: Start date: 05th Nov 2012 -
1. Sweet biscuits manufactured in Noida location DEFINE 05th Nov 2012 th
19 Nov, 2012
Project Out of Scope:
MEASURE 20th Nov, 2012 30th Nov, 2012
1. Everything beyond as mentioned above is out of scope for this project
ANALYZE 03rd Dec,2012 29th Dec, 2012
IMPROVE 02nd Jan, 2013 31st Jan,2013
CONTROL 01st Feb,2013 30th Apr, 2013
 d
Define
ARMI and Communication Plan
Key Stakeholders ARMI Worksheet
Define Measure Analyze Improve Control

GM Finance, Engg I I I I I
Sponsor - VP I I I I I
Champion - GM I&A I&A I&A I&A I&A
MBB A&I A&I A&I A&I A&I
Operation Manager I&M I&M I&M I&M I&M
Ashish R&M R&M R&M R&M R&M
Team Members M M M M M

Communication Plan

Information Or Activity Target Audience Information Channel Who When

Project Status Sponsor, Champion, BB, E-mails Dipti Nayak BI-Weekly
Members

Tollgate Review BB,MBB, Champion E-mails or Meetings Dipti Nayak As per Project Plan

Project Deliverables or Activities Members, MBB Emails, Meetings Dipti Nayak Weekly

A – Approval of team decisions I.e., sponsor, business leader, MBB.

R – Resource to the team, one whose expertise, skills, may be needed on an ad-hoc basis.
M – Member of team – whose expertise will be needed on a regular basis.
I – Interested party, one who will need to be kept informed on direction, findings.
 d
Define
SIPOC
Suppliers Inputs Process Output Customers

Amul, Gopal,
Ghee Packed Biscuits PARLE
Madhusudan
Kalkaji, Bikaji,
Maida
Panwar
Mawana Sugar
Packaging
Multi solutions
Materials

Dough Baking cooling

START Mixing Fermentation
making

END packing
 m
MEASURE
CTQ Tree

%breakage per Month

(Project Y Metric)

<2.0% breakage per

Month
(Target)

Reduce %age
rejection CTQs
2.0 % breakage per Month
(Upper Specification Limit)

Any Month % breakage

greater than 2.0%
(Defect Definition)

Project Y = (breakage in Kgs/per month) * 100

Production of biscuits in Kgs/per month
 m
Data Collection Plan MEASURE

Specification Limit
Project-Y Operational Definition Defect Def Performance Std Opportunity
LSL USL

Daily breakage
% breakage is the percentage calculated % breakage in a
%age breakage in a data compiled
based upon the breakage in Kgs (say n) to Month greater Less than 2.0% - 2.0%
month on Monthly
Production in kgs in a Month (say N) than 2.0%
basis

Plan to collect Data Plan to sample

Project-Y Data Type Unit of Measure Formula to be used What Database or Is this an If new, When will When is the
Container will be existing planned start
the database be
used to record this database or date for data
ready for use?
data? new? collection?

Excel is being used

(n/N)*100
to record the data
n is breakage in Kgs
%age breakage in a Percentage and this data is Mar’2012 to
Continuous per Month, N is the yes - -
month breakage in Kgs located in H:Drive Oct’2012
production in Kgs
/Six Sigma project
per month
file
 m
MEASURE

Validate Measurement System – Gage R & R

Purpose : The purpose of this study is to repeatability and reproducibility of the measurement system.

Scope : The Scope of this measurement system is limited to the recoding of breakages in Kg.

Procedure : 1) 10 Data samples were collected randomly of 10 days

2) These 10 days breakage were checked by 3 operators on the same weighing machine.
3) After data collection these results were analyzed in Minitab software
 m
MEASURE

Validate Measurement System – R & R KPI Data Type

%breakage
Continuous
in Kgs

Gage R&R (Xbar/R) for Data

Reported by :
G age name: Tolerance:
Date of study : M isc:

Components of Variation Data by P/N

100 % Contribution 16
% Study Var
Percent

12
50

8
0
Gage R&R Repeat Reprod Part-to-Part 1 2 3 4 5 6 7 8 9 10
P/N
R Chart by Operator
Data by Operator
Operator 1 Operator 2 Operator 3
Sample Range

UCL=0.01158 16
0.010
_
0.005 R=0.0045 12
0.000 LCL=0
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 8
P/N Operator 1 Operator 2 Operator 3
Operator
Xbar Chart by Operator
Operator 1 Operator 2 Operator 3 P/N * Operator Interaction
16
Sample Mean

16 Operator
_ Operator 1

Average
12 LCL=11.58
UCL=11.59
X=11.59 Operator 2
12 Operator 3

8
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 8
P/N 1 2 3 4 5 6 7 8 9 10
P/N

Inference: Graphical summary on R & R suggests that there is no significant variation between
appraisers.
 m
Validate Measurement System – R & R MEASURE

Gage R&R Study - XBar/R Method KPI Data Type

%breakage
%Contribution Continuous
in Kgs
Source VarComp (of VarComp)
Total Gage R&R 0.00001 0.00
Repeatability 0.00001 0.00
Reproducibility 0.00000 0.00
Part-To-Part 8.03352 100.00
Total Variation 8.03353 100.00

Study Var %Study Var

Source StdDev (SD) (6 * SD) (%SV)
Total Gage R&R 0.00268 0.0161 0.09
Repeatability 0.00266 0.0159 0.09
Reproducibility 0.00034 0.0020 0.01
Part-To-Part 2.83435 17.0061 100.00
Total Variation 2.83435 17.0061 100.00

Number of Distinct Categories = 1491

Inference : As gage R & R is less than 10% of the study variation, we concludes that data can be
trusted and good to go.
 m
MEASURE

Graphical summary - Y
Summary for C1
A nderson-Darling N ormality Test
A -S quared 0.80
P -V alue 0.038
M ean 4.2488
S tDev 0.9243
V ariance 0.8543
S kew ness 0.091904
Kurtosis 0.612410
N 208
M inimum 1.8074
1st Q uartile 3.7242
M edian 4.2457
3rd Q uartile 4.8933
2 3 4 5 6 7 M aximum 7.4933
95% C onfidence Interv al for M ean
4.1225 4.3752
95% C onfidence Interv al for M edian
4.1513 4.4089
95% C onfidence Interv al for S tDev
9 5 % C onfidence Inter vals
0.8432 1.0228
Mean

Median

4.10 4.15 4.20 4.25 4.30 4.35 4.40

Inference : The summary above represents that data (our project Y) in non-normal in nature as p value
is 0.038 which is less than 0.05. the measure of central tendency will be median
 m
MEASURE
Stability analysis

Run Chart of Total

20.0

17.5

15.0
Total

12.5

10.0

7.5

5.0
1 20 40 60 80 100 120 140 160 180 200
Observation
Number of runs about median: 89 Number of runs up or down: 126
Expected number of runs: 105.0 Expected number of runs: 138.3
Longest run about median: 7 Longest run up or down: 5
A pprox P-Value for C lustering: 0.013 A pprox P-Value for Trends: 0.021
A pprox P-Value for Mixtures: 0.987 A pprox P-Value for O scillation: 0.979

Inference : Stability analysis done by run chart, p-value for mixture, oscillation are more than 0.5,
suggests that there is data is stable from mixture and trend point of view, analysis needs to be
carried out for existence of trends and clustering. Data is not stable as p-value for clustering and
trends are below 0.05
 m
MEASURE

Process Sigma Level

TOTAL OPPUTUNITIES 56515

TOTAL FAILED 2377

(TOTAL FAILED) X 106)/TOTAL OPPURTUNITIES

DPMO
42059

CURRENT SIGMA LEVEL 3.22

Inference : Current process sigma level is 3.22

 a
Cause and Effect diagram from breakage Analyze

Cause-and-Effect Diagram
Measurements Material Personnel

Operator
WA P (maida) % V endor for M aida qualification

M aida % V endor for G hree

G hee % N /Y(7th to 10th ) 297 Kgs of

biscuits
suffers
break ge out
of 7064
O v en H eat up time
produced
M ixing Time
O v en Ty pe
Baking Temperature

Baking Time

Env ironment Methods Machines

Inference : brainstorming session has identified 11 contributors (Xs) that can affect the performance of
Y, these 11 contributing Xs will be analyzed using proper tools to identify & conclude whether they
have significant impact on Y.
 a
Data collection plan for Xs that contributes to printing defects Analyze
How it is
Typ Plan to Test to be
Measure Operation definition measured/
e sample performed
collected
Formula( n*100)/N
(Kgs of biscuits broken (n) x 100) / No. Mar’12 to
% Breakage Y of Kgs of biscuit produced (N))
Data for n, N is
Oct’12
Run chart
collected individually

Amount of time the contents are Mar’12 to Regression

Baking Time X Baked in baking machine
Timer
Oct’12 Analysis
Baking Mar’12 to Regression
X Temperature at which Baking is done Temperature meter
Oct’12
Temperature Analysis
There are three Ovens Oven A, Oven Mar’12 to Mood’s
Oven Type X B, Oven C in which the Baking is done
Oven Identification
Oct’12 Median Test
Mar’12 to
Amount of Time required to mix the
Oct’12
Regression
Mixing Time X contents
Timer
Analysis

Pre heating time before the contents Mar’12 to

Oven Heat-up Oct’12
Regression
X are placed for baking at certain Timer
Time temperature Analysis
 a
Data collection plan for Xs that contributes to Stitching defects Analyze
Measure Type Operation definition How it is Plan to Test to be
measured/ sample performed
collected
Water absorption percentage, is the Mo = weight w/o moisture
WAP Mar’12 to Mood’s
X total amount of moisture in the W= weight w/moisture
Oct’12
Maida(%) contents % = (W-Mo)*100/W Median Test

Vendor for Supplier Name on Mar’12 to Mood’s

X Different suppliers which supply Maida
packet Oct’12
Maida Median Test
%age of Maida in total contents, Mar’12 to Mood’s
% Maida X weight of Maida = M
M*100/W
Oct’12 Median Test

%age of Ghee in total contents, weight G*100/W Mar’12 to Mood’s

% Ghee X of Ghee = G Oct’12 Median Test

Vendor for Supplier Name on Mar’12 to Mood’s

X Different suppliers which supply Ghee
packet Oct’12
Ghee Median Test

Mann-
Operator 10th Pass (Yes) Mar’12 to
X Qualification of operators
10th Failed (No) Oct’12
Whitney
qualification
Test
 a
Hypothesis Testing : Analyze

Regression Analysis: Project Y versus Baking Time

The regression equation is

Project Y = 2.70 + 0.0933 Baking Time

Predictor Coef SE Coef T P

Constant 2.6962 0.7383 3.65 0.000
Baking Time 0.09331 0.04421 2.11 0.036

S = 0.916668 R-Sq = 2.1% R-Sq(adj) = 1.6%

Analysis of Variance

Source DF SS MS F P
Regression 1 3.7436 3.7436 4.46 0.036
Residual Error 206 173.0977 0.8403
Total 207 176.8413

Inference :P-Value observed is 0.036 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. Baking Time does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Regression Analysis: Project Y versus Baking Temper

The regression equation is

Project Y = 13.6 - 0.0443 Baking Temper

Predictor Coef SE Coef T P

Constant 13.646 4.343 3.14 0.002
Baking Temper -0.04434 0.02049 -2.16 0.032

S = 0.916172 R-Sq = 2.2% R-Sq(adj) = 1.7%

Analysis of Variance

Source DF SS MS F P
Regression 1 3.9306 3.9306 4.68 0.032
Residual Error 206 172.9106 0.8394
Total 207 176.8413

Inference :P-Value observed is 0.032 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. Baking Temperature does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus OVEN

Mood median test for Project Y

Chi-Square = 9.93 DF = 2 P = 0.007

Individual 95.0% CIs

OVEN N<= N> Median Q3-Q1 ---------+---------+---------+-------
OVEN A 56 55 4.238 1.343 (----------*--------)
OVEN B 32 17 4.028 1.008 (---------*-------)
OVEN C 16 32 4.475 0.837 (-------*-------)
---------+---------+---------+-------
4.00 4.25 4.50

Overall median = 4.246

Inference :P-Value observed is 0.007 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. Oven Type does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Regression Analysis: Project Y versus Mixing TIme

The regression equation is

Project Y = 7.09 - 0.338 Mixing TIme

Predictor Coef SE Coef T P

Constant 7.0874 0.5901 12.01 0.000
Mixing TIme -0.33816 0.06992 -4.84 0.000

S = 0.878025 R-Sq = 10.2% R-Sq(adj) = 9.8%

Analysis of Variance

Source DF SS MS F P
Regression 1 18.030 18.030 23.39 0.000
Residual Error 206 158.811 0.771
Total 207 176.841

Inference :P-Value observed is 0.000 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. Mixing Time does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Regression Analysis: Project Y versus OVEN Heatup Time

The regression equation is

Project Y = 4.44 - 0.0113 OVEN Heatup Time

Predictor Coef SE Coef T P

Constant 4.4407 0.1628 27.28 0.000
OVEN Heatup Time -0.011260 0.008782 -1.28 0.201

S = 0.922852 R-Sq = 0.8% R-Sq(adj) = 0.3%

Analysis of Variance

Source DF SS MS F P
Regression 1 1.4003 1.4003 1.64 0.201
Residual Error 206 175.4410 0.8517
Total 207 176.8413

Inference :P Value observed is 0.201, which is greater than 0.05, which suggests that we do not have
sufficient evidence to reject null hypothesis. Oven heat up time does not significantly impacts
breakage
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus WAP(maida) %

Mood median test for Project Y

Chi-Square = 2.26 DF = 5 P = 0.813

Individual 95.0% CIs

WAP(maida) % N<= N> Median Q3-Q1 ----+---------+---------+---------+--
7.0 11 6 4.17 0.87 (-------*----------------)
7.1 17 18 4.34 1.41 (----------------*--------)
7.2 27 25 4.21 1.12 (---------*-------)
7.3 24 27 4.30 1.48 (----------------*--------)
7.4 9 8 4.17 1.07 (-----------*---------------------)
7.5 16 20 4.47 1.14 (------------*-----)
----+---------+---------+---------+--
3.90 4.20 4.50 4.80

Overall median = 4.25

Inference :P Value observed is 0.813, which is greater than 0.05, which suggests that we do not have
sufficient evidence to reject null hypothesis. WAP (Maida)% does not significantly impacts breakage
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus Moisture(Ghee) %

Mood median test for Project Y

Chi-Square = 14.22 DF = 3 P = 0.003

Individual 95.0% CIs

Moisture(Ghee) % N<= N> Median Q3-Q1 --+---------+---------+---------+--
--
11 15 2 3.48 1.25 (-----------*----)
12 52 48 4.21 1.34 (-*--)
13 9 10 4.26 1.26 (----*----------)
14 27 44 4.47 0.99 (--*-)
15 1 0 3.51 *
--+---------+---------+---------+--
--
2.80 3.50 4.20 4.90

Overall median = 4.25

Inference :P-Value observed is 0.003 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. Ghee (%age) does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus Vendor for Maida

Mood median test for Project Y

Chi-Square = 1.61 DF = 2 P = 0.446

Vendor Individual 95.0% CIs

for Maida N<= N> Median Q3-Q1 -------+---------+---------+---------
Bikajee 55 57 4.270 1.143 (--------*-------)
Kalkaji 26 19 4.097 1.298 (----------*------------)
Panwar 23 28 4.413 1.156 (-------------*------)
-------+---------+---------+---------
4.00 4.25 4.50

Overall median = 4.246

Inference :P Value observed is 0.446, which is greater than 0.05, which suggests that we do not have
sufficient evidence to reject null hypothesis. Vendor of Maida does not significantly impacts
breakage
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus Ghee%

Mood median test for Project Y

Chi-Square = 14.22 DF = 3 P = 0.003

Individual 95.0% CIs

Moisture(Ghee) % N<= N> Median Q3-Q1 --+---------+---------+---------+---
-
11 15 2 3.48 1.25 (-----------*----)
12 52 48 4.21 1.34 (-*--)
13 9 10 4.26 1.26 (----*----------)
14 27 44 4.47 0.99 (--*-)
15 1 0 3.51 *
--+---------+---------+---------+---
-
2.80 3.50 4.20 4.90

Overall median = 4.25

Inference : P-Value observed is 0.003 which is less than 0.05, which suggests that we have sufficient
evidence to reject null-hypothesis. % of Ghee does significantly impacts breakage.
 a
Hypothesis Testing : Analyze

Mood Median Test: Project Y versus Vendor for Ghee

Mood median test for Project Y

Chi-Square = 3.82 DF = 2 P = 0.148

Vendor for Individual 95.0% CIs

Ghree N<= N> Median Q3-Q1 --------+---------+---------+--------
Amul 46 32 4.086 1.089 (----------*-------)
Gopal 26 32 4.402 1.186 (-----------*----------)
Madhusudan 32 39 4.400 1.156 (---------*-------)
MD 0 1 4.554 *
--------+---------+---------+--------
4.00 4.25 4.50

Overall median = 4.246

Inference : P Value observed is 0.148, which is greater than 0.05, which suggests that we do not have
sufficient evidence to reject null hypothesis. Vendor for Ghee does not significantly impacts
breakage
 a
Hypothesis Testing : Analyze

Mann-Whitney Test and CI: No, Yes

N Median
No 96 3.9150
Yes 112 4.4698

Point estimate for ETA1-ETA2 is -0.5641

95.0 Percent CI for ETA1-ETA2 is (-0.8246,-0.3031)
W = 8136.0
Test of ETA1 = ETA2 vs ETA1 not = ETA2 is significant at
0.0000
The test is significant at 0.0000 (adjusted for ties)

Inference :: P Value observed is 0.5641, which is greater than 0.05, which suggests that we do not have
sufficient evidence to reject null hypothesis. Operator qualification does not significantly impacts
breakage
 a
Summary on Findings on Xs contributing to Printing Defects Analyze

Test
Measurable Type Data Type P-Value Inference
Performed

% Breakage Y Continuous Run chart N/A N/A

Regression Statistically, X has significant impact on Y, as

Baking Time X Continuous 0.036
P Value is less than 0.05
Analysis
Baking Regression Statistically, X has significant impact on Y, as
X Continuous 0.032
Temperature Analysis P Value is less than 0.05

Mood’s Statistically, X has significant impact on Y, as

Oven Type X Discrete 0.007
P Value is less than 0.05
Median Test

Regression Statistically, X has significant impact on Y, as

Mixing Time X Continuous 0.000
P Value is less than 0.05
Analysis
Oven Heat-up Regression Statistically, X has no significant on Y, as P
X Continuous 0.201
Time Analysis value is greater than 0.05
 a
Summary on Findings on Xs contributing to Stitching Defects Analyze

Typ Data Test

Measurable P-Value Inference
e Type Performed
Mood’s
WAP Maida(%) Y Discrete N/A N/A
Median Test
Vendor for Mood’s Statistically, X has no significant on Y, as P
X Discrete 0.446
Maida Median Test value is greater than 0.05

Mood’s Statistically, X has no significant on Y, as P

% Maida X Discrete 0.813
value is greater than 0.05
Median Test

Mood’s Statistically, X has significant impact on Y, as

% Ghee X Discrete 0.003
P Value is less than 0.05
Median Test

Vendor for Mood’s Statistically, X has no significant on Y, as P

X Discrete 0.148
Ghee Median Test value is greater than 0.05

Operator Mann- Statistically, X has no significant on Y, as P

X Discrete 0.564
qualification Whitney Test value is greater than 0.05
 i
Box – plot on Baking Time, Baking Temperature Improve
Boxplot of Project Y Boxplot of Project Y
8 8

7 7

6 6
Project Y

Project Y
5 5

4 4

3 3

2 2

A1 A2 A3 A4 A5
A B C D
Baking Temperature Category
Baking Time category

Boxplot of Project Y Baking Time and temperature data have

been categorized to facilitate to draw them
A B C D
in box plot .
A1 A2 A3
8
Box plot of Baking time and Temperature
6 does not produce any concrete information
on specific contributors.
4
Box plot of time versus temperature shows
Project Y

2 that with Baking time category A and

A4 A5 A B C D Temperature category A3 there is some
8
scope of improvement
6

A B C D
Baking Time category
Panel variable: Baking Temperature Category
 Box Plot on Oven Type and Mixing Time i
Improve
Boxplot of Project Y_1 Boxplot of Project Y_1
8 8

7 7

6 6
Project Y_1

Project Y_1
5 5

4 4

3 3

2 2

OVEN A OVEN B OVEN C 7 8 9 10

OVEN Mixing Time

Boxplot of Project Y_1 Box Plot for Oven type and Mixing time has been
7 8 9 10 Plotted
OVEN A OVEN B
8 With oven type it is hard to conclude which Oven plays role
In reduced breakage, but with Mixing time, “10” time is
6
playing role in breakage
4
Project Y_1

Box plot for Mixing time and Oven type shows that If mixing
2 time is “10” and Oven Type is “C” combination of these two
OVEN C contributes to reduction in breakage
8

2
7 8 9 10
Mixing Time
Panel variable: OVEN
 Box Plot on Ghee Percentage i
Improve

Boxplot of Project Y_1

8
Box plot on ghee percentage concludes
that breakage is low, when the ghee
7
percentage used is 11%

6
Project Y_1

11 12 13 14 15
% Ghee
 i
QFD-House of Quality (process requirements-Production Methods) Improve
 i
FMEA for Biscuit Manufacturing Process Improve
 i
Implementing Solutions and Time Lines – Pilot Lot Run Improve
 i
Evaluation of the Improvement in Pilot Lot Run Improve
I-MR Chart of %age breakage by C12
1 2
8 1
Individual V alue

4 U C L=3.800
_
2 X=2.048

LC L=0.295
0
1 23 45 67 89 111 133 155 177 199 221
O bser vation

1 2
1
4 1 1 1 1
M oving Range

2 U C L=2.154

1 __
M R=0.659
0 LC L=0
1 23 45 67 89 111 133 155 177 199 221
O bser vation

Process after improvement is marked in dotted circle, shows that improvement is there in the pilot lot. No
points is falling beyond control limits suggests that the process is in statistical control
 i
Hypothesis Test to validate statistically significant improvement Improve

Two-Sample T-Test and CI: Before, After

Two-sample T for Before vs After

N Mean StDev SE Mean

Before 208 4.249 0.924 0.064
After 15 2.048 0.492 0.13

Difference = mu (Before) - mu (After)

Estimate for difference: 2.201
95% CI for difference: (1.905, 2.497)
T-Test of difference = 0 (vs not =): T-Value = 15.48 P-Value = 0.000 DF =
21

Inference: Two sample proportion test was conducted to test whether there was statistically significant
improvement during pilot run, p value observed is 0.000 which is less than 0.05, thus we can
conclude that there is significant improvement.
 i
Hypothesis Test to validate whether the target is met or not Improve

One-Sample T: After

Test of mu = 2 vs not = 2

Variable N Mean StDev SE Mean 95% CI T P

After 15 2.048 0.492 0.127 (1.775, 2.320) 0.37 0.714

Inference: Hypothesis test was conducted to check whether pilot results proportion is meeting the target
proportion of 0.015, P-value observed is 0.714 which is greater than 0.05 which shows that there is
no significant difference in proportion of the pilot lot and the target, which means the target is met.
 c
Control chart for checking the sustainability over a period of time Control
I-MR Chart of %age breakage by C12
Post pilot monitoring
1 2 3 of the process
8 1
over one month
Individual V alue

4
U C L=3.322
_
2 X=2.034

LC L=0.746
0
1 26 51 76 101 126 151 176 201 226 251
O bser vation

Pilot lot monitoring

1 2 3
1
4 1 1 1 1
M oving Range

2
U C L=1.583
1 __
M R=0.484
0 LC L=0
1 26 51 76 101 126 151 176 201 226 251
O bser vation
Inference: Post pilot monitoring of process shows that process is still in statistical control.
 c
CONTROL PLAN Control
 c
IMPROVED PROCESS HANDOVERING – PROJECT CLOSURE Control
I-MR Chart of %age breakage by C12
1 2 3
8 1
Individual Value

4
U C L=3.322
_
2 X=2.034

LC L=0.746
0
1 26 51 76 101 126 151 176 201 226 251
O bse r v a tion

1 2 3
1
4 1 1 1 1
Moving Range

2
U C L=1.583
1 __
M R=0.484
0 LC L=0
1 26 51 76 101 126 151 176 201 226 251
O bse r v a tion

1. Process is being handed over to the process owner with the status of improvement
shown in the above control chart
2. Control chart should be in place for monitoring the process
3. Revised Process documents, auditing checklist, control plan handed over to operating
personnel with changes that were done during pilot run.
4. Out of control action plan for control chart developed and handed over to operations
manager