Six Sigma Project on Increase in Productivity through Waste Reduction

12/23/20141 12/23/20141

DMAIC

12/23/20142

Customer Sample Comments

Key Output Characteristics

Important to Customer

(CTQ's)

VP • The objective is to increase the overall

productivity & minimise the waste without

compromising quality

• Improvement in productivity

• Reduce waste

• Quality

Process owner • We have to reduce the total no of breakage

and maintain the breakage below 3% .

• Improvement in Productivity

• Reduce % of breakage.

DefineMap the Project

12/23/20143

Project Charter Define

Project Leader: Soumen Jana Team Members

Business Case: “X” has been one of the leading biscuit brands in

India for the last 8 decades. “X” is reputed to enjoy the most sales

among all biscuit brands in the world. Its biscuits offer a fantastic

combination of quality, taste, and nutrition. “X” biscuits are

available even in faraway villages. At present the company has a 40

percent share of the Indian market for biscuits and is a multi-million

dollar organization.

Stakeholders VP

Sponsor GM

SME Sen Manager

Team Member SME,QA, 2 Associates, Financial controller, GB, BB

Problem Statement: For the period of June 10 to Jan 11 the total productivity was reduced as

%breakage is more than 3.

Goal Statement:To improve process productivity by decreasing the total no of breakage.

%Breakage should be decreased to below 3.

Project In Scope: 1. Production, Quality, vendors and store

Project Out of Scope:

Timelines/Milestones/Phases

Start Date End Date

Start date: 1h Feb 2011 -

DEFINE 15th Feb 2011 10th March

MEASURE 11th March 15th April

ANALYZE 17th April 30th May

IMPROVE 5th May 20th June

CONTROL 25th June 15th Aug 2011

12/23/20144

Key Stakeholders Define Measure Analyze Improve Control

VP I

GM A R A M A

Senior Manager M M M M M

BB M M M M M

GB M M M M M

M1,M2 -- --- M M M

Financial controller I I I I M

Message Audience Media Who When

Communication Plan

Define

ARMI Worksheet

ARMI & Communication Plan

12/23/20145

Customer Output Process Input Supplier

COPIS Define

Purchase Raw

Materials

Quality

Inspection

Issue RMs to

production

Mixing of RMs

in predefined

condition

Dough Making

Designing &

cutting

Vendors Order placed Raw Material Store and Quality

Warehouse RM sampleApproved RMs Production

Warehouse and

QualityApproved RM

Start Process Production

Production

operatorRaw material

charging as

per sequence

Homogeneous

pasteProduction

Production Paste Proper Dough Production

Production Proper Dough Proper shape Production

12/23/20146

Process Map Define

Baking

Empty the

baking trey

Divide in 6

groups

FG Store

Production Raw shaped

biscuitBaked biscuit Production

Production Approved

Biscuits

Packed biscuit FG store

12/23/20147

Y Operational

Definition

Defect

Definition

Performance

Standard

Specification

Limit

Opportunity

%Breakage It is defined as the

percentage of total

no of biscuits

broken against total

production

%Breakage greater

than 3

%Breakage <=3 USL = 3

LSL = 0

Daily

Y Data Type UnitDecimal to

be Used

Data Base

Container

Existing or

new data

base

If New when

data base

would be

ready.

Plan start

date for

DCP

%Breakage Discrete No Upto Two Excel Existing NA 11th March

11

Data Collection Plan

Equipment

Used for

measurement

Equipment

Calibration

Information

Responsibility

Any

Training

need

Operator

Information

Measure

12/23/20148

Measurement System Analysis

Rule Rule Description Acceptable Result

AR&R % of

Tolerance<10%

B

% Contribution

(R&R Std

deviation)

Smaller than

Part-to-part

variance

CNumber of distinct

categories>4

Overall Gage result

Minitab Descriptive Stats

Insert Minitab session window descriptive stats.

For discrete data use the attribute gage study

exhibit from Minitab

Measure

12/23/20149

Measure

Median 4.25

Mean 4.430

Mode 4.969

Std Dev 0.924

Cp 0.541

Cpk -0.516

Z value -1.547

Z Score of the process is really poor, there is immediate need to improve the process capability

Process capability Analysis

12/23/201410

Data Stability test

Take away: Patterns suggest that the variation observed is due to "special causes“. further investigation needs to be done to ascertain the

causes of and mixtures.

P value for mixture is less than 0.05.

means data is not stable.

Analyze

12/23/201411

Normality Test

Normality: P value = 0.038

Shape: Non-Normal

Measure of central tendency :data

is non-normal measure of central

tendency will be Median = 4.249

Analyze

12/23/201412

I-MR test of % BreakageAnalyze

The red points on the I-MR chart shows that currently the Breakage is out of control in the process and

requires an urgent attention.

X

=3

12/23/201413

Proposed tests according to problem and factor’s data typeAnalyze

S No. Potential Cause Operational Definition Data Type Test of be performed

1%Breakage from

MachinesPercentage of broken biscuit for each machine against total biscuit produced

Cont Correlation & Regression

2 Baking time Time required for baking of biscuits produced in a day ContCorrelation & Regression

3 Baking temp. Temp. required for baking of biscuit produced in a day Cont Correlation & Regression

4 Oven Used for baking of biscuits Discrete 1-sample sign.

5 WAP(maida%) ContCorrelation & Regression

6 Oval heat up time Cont Correlation & Regression

7 Vendor of Maida Discrete 1-sample sign.

8 Moisture %(maida) Moisture present in maida in percentage ContCorrelation & Regression

9 Moisture (Ghee%)Moisture present in ghee in percentage

ContCorrelation & Regression

10 Vendor of Ghee Discrete 1-sample sign.

12/23/201414

Regression Analysis: % Breakage versus %A

The regression equation is

% Breakage = 3.57 + 0.817 %A

Predictor Coef SE Coef T P

%A 0.8170 0.1039 7.87 0.000

S = 0.812521 R-Sq = 23.1% R-Sq(adj) =

22.7%

AnalyzeRegression Test between %Breakage & %Breakage from each machine

Regression Analysis: % Breakage

versus %B


% Breakage = 4.04 + 0.305 %B


%B 0.3046 0.1446 2.11 0.036

S = 0.916705 R-Sq = 2.1% R-

Sq(adj) = 1.6%

Regression Analysis: % Breakage versus

%C


% Breakage = 3.73 + 0.855 %C


%C 0.8550 0.1332 6.42 0.000

S = 0.845816 R-Sq = 16.7% R-Sq(adj) =

16.3%


versus %D


% Breakage = 4.36 - 0.328 %D

Predictor Coef SE Coef T

P

%D -0.3283 0.3136 -1.05

0.296


versus %E


% Breakage = 3.75 + 0.908 %E


%E 0.9079 0.1925 4.72 0.000

S = 0.880239 R-Sq = 9.7% R-Sq(adj)

= 9.3%


% Breakage = 3.47 + 0.641 %F


%F 0.64115 0.08125 7.89 0.000

S = 0.811917 R-Sq = 23.2% R-Sq(adj)

= 22.8%

The Regression test shows that since the P-Value is < 0.05, %Breakage from Machine A,B,C,E,F (X) has impact

on Total % Breakage (Y).

12/23/201415


Baking Time


% Breakage = 2.70 + 0.0933 Baking Time


Baking Time 0.09331 0.04421 2.11 0.036

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

1.6%

Regression Test between %Breakage &

%Breakage from each machineRegression Test between %Breakage & Baking time

Analyze

The Regression test shows that since the P-Value is < 0.05, Baking time has impact on % Breakage (Y).

12/23/201416

Regression Analysis: % Breakage versus Baking Temp


% Breakage = 13.6 - 0.0443 Baking Temp


Baking Temp -0.04434 0.02049 -2.16 0.032

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

Regression Test between %Breakage & Baking temp Analyze

The Regression test shows that since the P-Value is < 0.05, Baking temp has a –ve impact on % Breakage (Y).

12/23/201417

Mood Median Test: Ovens vs %Breakage

Mood median test for C2

Chi-Square = 9.93 DF = 2 P = 0.007

Subscripts N<= N> Median Q3-Q1

% Breakage_Oval C 16 32 4.475 0.837

% Breakage_Ovan A 56 55 4.238 1.343

% Breakage_Ovan B 32 17 4.028 1.008

Individual 95.0% CIs

Subscripts ---------+---------+---------+-------

% Breakage_Oval C (-------*-------)

% Breakage_Ovan A (----------*--------)

% Breakage_Ovan B (---------*-------)

---------+---------+---------+-------

4.00 4.25 4.50

Overall median = 4.246

Regression Test between %Breakage & OvenAnalyze

The Moods median test shows that since the P-Value is < 0.05, there is a difference between three ovens data.

Sign Test for Median: % Breakage_Oval C

Sign test of median = 3.000 versus not = 3.000

N Below Equal Above P Median

% Breakage_Oval C 48 2 0 46 0.0000 4.475

Sign Test for Median: % Breakage_Ovan A



% Breakage_Ovan A 111 12 0 99 0.0000 4.238

Sign Test for Median: % Breakage_Ovan B



% Breakage_Ovan B 49 5 0 44 0.0000 4.028

The 1-sample sign test shows that since the P-Value is < 0.05, there is an impact of ovens on %Breakage.

12/23/201418

Regression Analysis: % Breakage versus Oval Heatup

Time


% Breakage = 4.44 - 0.0113 Oval Heatup Time


Oval Heatup Time -0.011260 0.008782 -1.28 0.201

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

Regression Test between %Breakage & Oval heat up timeAnalyze

The Regression test shows that since the P-Value is > 0.05, oval heat up time has no impact on % Breakage (Y).

12/23/201419

Regression Test between %Breakage & Wap (maida%)Analyze


WAP(maida) %


% Breakage = 1.66 + 0.357 WAP(maida) %


WAP(maida) % 0.3570 0.4257 0.84 0.403

S = 0.924950 R-Sq = 0.3% R-Sq(adj) =

0.0%

The Regression test shows that since the P-Value is > 0.05, Wap (Maida%) has no impact on % Breakage (Y).

12/23/201420

Regression Test between %Breakage & Maida vendorAnalyze

Mood Median Test: Maida vendors


Chi-Square = 1.61 DF = 2 P = 0.446


% Breakage_Bikajee 55 57 4.270 1.143

% Breakage_Kalkaji 26 19 4.097 1.298

% Breakage_Panwar 23 28 4.413 1.156


Subscripts -------+---------+---------+---------

% Breakage_Bikajee (--------*-------)

% Breakage_Kalkaji (----------*------------)

% Breakage_Panwar (-------------*------)

-------+---------+---------+---------

4.00 4.25 4.50


The Moods median test shows that since the P-Value is >0.05, there is no difference between data of Maida

vendors

The 1-sample sign test shows that since the P-Value is < 0.05, there is an impact of Maida vendors on

%Breakage.

Sign Test for Median: % Breakage_Kalkaji



% Breakage_Kalkaji 45 5 0 40 0.0000 4.097

Sign Test for Median: % Breakage_Panwar



% Breakage_Panwar 51 4 0 47 0.0000 4.413

12/23/201421

Regression Test between %Breakage & Moisture % in MaidaAnalyze


Moisture(Maida) %


% Breakage = 4.35 - 0.0078 Moisture(Maida) %


Moisture(Maida) % -0.00784 0.06066 -0.13 0.897

S = 0.926490 R-Sq = 0.0% R-Sq(adj) = 0.0%

The Regression test shows that since the P-Value is > 0.05, Moisture% (Maida) has no impact on % Breakage

(Y).

12/23/201422

Regression Test between %Breakage & Moisture % in GheeAnalyze

Regression Analysis: % Breakage versus Moisture(Ghee) %


% Breakage = 1.15 + 0.243 Moisture(Ghee) %


Moisture(Ghee) % 0.24349 0.05938 4.10 0.000

S = 0.890878 R-Sq = 7.5% R-Sq(adj) = 7.1%

The Regression test shows that since the P-Value is < 0.05, Moisture% (Ghee) has an impact on % Breakage

(Y).

12/23/201423

Regression Test between %Breakage & Vendor for GheeAnalyze

Sign Test for Median: % Breakage_Amul



% Breakage_Amul 78 7 0 71 0.0000 4.086

Sign Test for Median: % Breakage_Gopal



% Breakage_Gopal 58 6 0 52 0.0000 4.402

Sign Test for Median: % Breakage_Madhusudan



% Breakage_Madhusudan 71 6 0 65 0.0000 4.400

Sign Test for Median: % Breakage_MD



% Breakage_MD 1 0 0 1 1.0000 4.554

Mood Median Test: C2 versus Subscripts


Chi-Square = 3.82 DF = 2 P = 0.148


% Breakage_Amul 46 32 4.086 1.089

% Breakage_Gopal 26 32 4.402 1.186

% Breakage_Madhusudan 32 39 4.400 1.156

% Breakage_MD 0 1 4.554 *


Subscripts --------+---------+---------+--------

% Breakage_Amul (----------*-------)

% Breakage_Gopal (-----------*----------)

% Breakage_Madhusudan (---------*-------)

% Breakage_MD

--------+---------+---------+--------

4.00 4.25 4.50


The Moods median test shows that since the P-

Value is >0.05, there is no difference between data

of Ghee vendors

The 1-sample sign test shows that since the P-Value is < 0.05, there is an impact of Ghee vendors on %Breakage.

12/23/201424

S No. Potential Cause Data Type Test of be performed Impact

1%Breakage from

MachinesCont Correlation & Regression Significant relationship of machines except D on Y

2 Baking time ContCorrelation & Regression Significant relationship on Y

3 Baking temp. Cont Correlation & RegressionSignificant relationship on Y

4 Oven Discrete 1-sample sign.Significant relationship on Y

5 Oval heat up time ContCorrelation & Regression No significant relationship on Y

6 WAP(maida%) ContCorrelation & Regression No significant relationship on Y

7 Vendor of Maida Discrete 1-sample sign.Significant relationship of Maida vendors on Y

8 Moisture (maida%) ContCorrelation & Regression No Significant relationship on Y

9 Moisture (Ghee%) ContCorrelation & Regression Significant relationship on Y

10 Vendor of Ghee Discrete 1-sample sign.Significant relationship of Ghee vendors except MD on Y

Summary of Statistical AnalysisSummary of Statistical AnalysisAnalyze

The factors highlighted in red are the vital X’s

Six Sigma Project on Increase in Productivity through Waste Reduction

News & Politics