THE SIX SIGMA APPROACH TO REDUCE THE PRODUCT DAMAGES IN THE WAREHOUSE Assist.Prof.Dr.Nihal ERGİNEL Prof.Dr.Nimetullah BURNAK 2003.

THE SIX SIGMA APPROACH TO REDUCE THE PRODUCT

DAMAGES IN THE WAREHOUSE

Assist.Prof.Dr.Nihal ERGİNEL

Prof.Dr.Nimetullah BURNAK

2003

Introduction

The most common complaint of end customers about the defects is the product damages.

This damage can occur to the warehouse in which the products are packaged, stored, and shipped.

Six sigma methodology is applied to eliminate the complaints about the product damages.

Steps

We set process map, and cause – effect diagram to find the effective factors.

Their priorities are assigned via Chi-Square test and cause – effect matrix.

It is decided that the product damages can occur in any one of the two phases.

1. The first phase is the packaging for which packaging materials and packaging method are considered.

2. The second phase is the transportation for which handling equipments are analyzed.

DOE is conducted separately for each of the phases, and the results are discussed.

Process Map

Warehouse Truck Shopping Center Truck Customer

forklift attachment wet floor dropped product wet floor transporting withoutpressure package

nail stroke in shopwindow nailforklift attachment stroke to the door, pressure angel traffic accident traffic accident elevator, banister, etc.

hand vechile stroke

wood support stroke

type of board gutter Critical parameter

sytrophor density Process parameter

wet/dry board Noise

sytrophor thickness Instructions

distance betweenboard and product

number of producttrasported with forklift

Instructions

Cause and Effect Diagram

damage on the product

forklifts in the warehouse

packaging equipments

truck

attachment pressure

attachments angel

sytrophor density

sytrophor tickness

board type

nail in the truck

traffic accident

wet board due to the wet floor of truck

packaging method

with/ without wood support

distance between board and product

number and the wideness of the ribbon

shopping center warehouse

number product transported with forklift

hand vechile stroke

wood support stroke

dropped product

stroke in shopwindow

customer

transported witout package

srofe to the door

stroke to the elevator

stroke to the banister

Chi-Square test

Table-1: Chi-Square Test: front; two sides; non-damage vs. model

Expected counts are printed below observed countsFront Two side Non-damage Total

A 224106,90

8061,97

4545145586,13

45755

B 5645,20

1726,20

1927319274,60

19346

C 105131,34

5376,13

5605856008,53

56216

D 48149,56

10186,70

6386763779,74

64016

Total 433 251 184649 185333Chi-Sq =128,277 + 5,248 + 0,401 + 2,581 + 3,231 + 0,000 + 5,282 + 7,030 + 0,044 + 68,968 + 2,359 + 0,119 = 223,539DF = 6, P-Value = 0,000

It is clear that Model A has significantly higher number of damages on front than the other models. Model-D has significantly lower number of damages on front than the other models. This main difference between the packaging types of model A and D is the distance between the product and the board. The mentioned distance is larger in model-D than in model-A.

Cause-Effect Matrix

Rating of Importance to Customer10

1 2

Process Inputs Dam

age

on

the

Pro

duct

Total1 Sytrophor density 9 92 Sytrophor thickness 9 93 Board gutter type 8 84 Distance between product and board 8 85 Forklift attachment pressure 7 76 Number product transported with forklift 7 77 Forklift attachment pressure angel 6 68 Door, elevator, banister etc. stroke 6 69 Hand vehicle stroke 6 6

10 Wood support stroke 6 611 Forklift stroke 6 612 Transport without package 5 513 Stroke in shopwindow 5 514 Package without wood support 5 515 Truck nail 5 516 Dropped product 5 517 Truck wet floor 4 419 Wet/dry board 3 320 Ribbon number and wideness 2 224 Truck traffic accident 1 1

Total 113

Cause and Effect Matrix

DOE-1on the packaging materials and methods

The response is the area of the damage measured in mm2 by using caliper. Gage R&R study was conducted before DOE-1.

Factors Level (low) Level (high)

Sytrophor density SD1 SD2

Sytrophor thickness ST1 ST2

The distance between board and product

D1 D2

Board type Type-1 Type-2

Anova Table for DOE-1

Analysis of Variance for mm2 Source DF SS MS F P syt.dens 1 298,43 298,43 23,36 0,001 distance 1 120,78 120,78 9,45 0,011 board 1 30,25 30,25 2,37 0,152 distance*board 1 85,93 85,93 6,73 0,025 Error 11 140,55 12,78 Total 15 675,94

The residual analysis is performed, and it was seen that residuals were normal. The plot of residuals versus time shows that there is no any pattern. Also, there is no indication of inequality of the variances by the plot of residuals versus fitted values.

Main effects

syt.dens. distance

17

19

21

23

25

mm

2

Main Effects Plot - Data Means for mm2

Interaction effect

1 -1

1-1

25

20

15

board

distance

Me

an

Interaction Plot - Data Means for mm2

This interaction effect shows that response is robust for board type-2.

Pie chart

distance (121; 17,9%)

board ( 30; 4,5%)

syt.dens (298; 44,2%)

Error (141; 20,8%)

distance*boa ( 86; 12,7%)

Pie Chart of Source•The sytrophor density can explain the 44%,• the distance between board and the product can explain 18%, • the interaction between distance and board type can explain 13%,

of the damaged area.

DOE-2 on the set up values of forklift which is transportation equipment

Factors Level (low) Level (high)

Attachment pressure PR1 PR2

Attachment angel AG1 AG2

The number of product transported with forklift

TP1 TP2

The model of the product Model-1 Model-2

The response is the severity of the damage.

The response :

Severity

Defination

1 No damage

2 Little degree damage

3 Middle degree damage

4 Big degree damage

5 Very big damage

The attribute Gage R&R was made.

Anova for DOE-2

There is no any interaction. Attachment pressure and angel are affected on the damage severity.

Analysis of Variance for response Source DF SS MS F P pressure 1 3,1250 3,1250 5,95 0,059 angel 1 6,1250 6,1250 11,67 0,019 Error 5 2,6250 0,5250 Total 7 11,8750

Main Effects

angelpressure

3,1

2,7

2,3

1,9

1,5

resp

onse

Main Effects Plot - Data Means for response

Residual analysis

The residual analysis is performed, and it was seen that residuals were normal.

The plot of residuals versus time shows that there is no any pattern.

Also, there is no indication of inequality of the variances by the plot of residuals versus fitted values.

Pie Chart

pressure (3; 26,3%)Error (3; 22,1%)

angel (6; 51,6%)

Pie Chart of Source

The attachment pressure and angel are the significantly main effects with 26% and 52%, respectively

Results

By using the results from DOE-1 the packaging method and the material used were redesigned.

The values of attachment pressure and the attachment angel should be controlled periodically according to DOE-2.

These improvements were implemented, and a significant reduction was observed at the ratio of damaged product.

Further Research

Since these researches just in warehouse, it may be extended to other points of the process map, and the ratio of damaged product can be reduced.

Thank you.