 

1.5505“. 1..

.A u ,
5LA1‘:
: in r.

a

0:. t
1.5.4....

14

In. .

3......

513'! 111‘
132:. a v ,1!
4...! r: 99.;

L. 3:...

an:
. J:
..

.
5.4:
.12

x
.t :7 3.19.
1.11 i.
i...

>.r. .

.37}331

I!!!
I

.s In...

t
5

if?! a 9.“.
“W a! ._ ‘11..
53...)... 6
no.1: «I. a

3..

,
333...»:
.' .‘o. is]:
‘ .. s.

..

a...

.
3.3.3.
:...n;a:
a...
an.

.
19.: 59'
0...!) i.
A. a}! V

a:
at...
a:

n.

s
97

NA... '1

ﬁi. ;

. ; m1.
.: 2 :21!) 3.3.55... 2.9

 

3.1.5,
5.4.3.23...
1.! x ,

1.2 a.
{it}.
.1. .9 |.

~

. {hunt}!
I

H.
1

51th.”? :2

3- I...

Hi... ..

u. 5....

3:33,... an
.1.

nahmsﬂmwm

.1114
A

-

E.

3 B...“

«$1.35) 1:
E. his.

.5... H. a
. . a... z
a. l.
1 J35
.25...

.1,

It i
1' 1...,
3.5:

2.: 1..
i2.
tag:
.25
2...,"
:32,
1.8
unit-Incl,
,
3;...

than”
i 3-

{Ram
. ., f...
iﬁéhn

...

,4
."'l;.

.n\-

 

 

"\

400¥

This is to certify that the
thesis entitled

The Effect of Conveyor Speed, Packaging Materials, and
Product on the Readability of Radio Frequency Identiﬁcation
Transponders

presented by

Jonathan Ryan Thomas Falls

has been accepted towards fulﬁllment
of the requirements for the

 

 

 

Master degree in Packaging
V) Major ProfessofS'Signature

Agust 24, 2006

 

Date

MSU is an Afﬁrmative Action/Equal Oppodunity Institution

_ _.-.—-—-.--.--.-.---—-u-u-n--o-‘-------4--u-.-»-n-n---.--.-o-o---o----n-.-.-.-.-u-o-a-.-—

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DaleDue indd-p.1

THE EFFECT OF CONVEYOR SPEED, PACKAGING MATERIALS, AND
PRODUCT ON THE READABILITY OF RADIO FREQUENCY IDENTIFICATION
TRANSPONDERS
By

Jonathan Ryan Thomas Falls

A THESIS

Submitted to
Michigan State University
In partial fulﬁllment of the requirements
For the degree of

MASTER OF SCIENCE
School of Packaging

2006

ABSTRACT
THE EFFECT OF CONVEYOR SPEED, PACKAGING MATERIALS, AND
PRODUCT ON THE READABILITY OF RADIO FREQUENCY IDENTIFICATION
TRANSPONDERS
By

Jonathan Ryan Thomas Falls

Since the ﬁrst supplier mandates put forth by Wal-Mart and the Department of
Defense, the use of (915 MHz) radio frequency identiﬁcation (RFID) has been
implemented into supply chains with mixed results. When working optimally, RFID can
provide valuable information regarding inventory data and Shipment locations. However,
tag readability issues exist due to a variety of reasons: product and package interference,
RFID equipment set-up locations, and even frequency allocations, depending on the
country of use.

The purpose of this research is to determine if conveyor speed, packaging
materials, and product have an affect on the readability of RFID transponders. The
variables for this testing were conveyor speed (300 feet per minute (fpm), 600 fpm),
package type (case of chips in plastic tubs, case of chips in metalized spiral wound
ﬁberboard containers (MSWFC)), package shape (case of metal cans, case of metal
bottles, and case of metal tins), product type (case of bottled ketchup, case of bottled
motor oil) and tag generation (Alien Gen 1, Alien Gen 2).

The research found that conveyor speed, package type, package shape, and
product type all had a signiﬁcant effect on the average amount of tag reads per trial, Tag
type was found to have a signiﬁcant effect when testing the product effect, and package

shape effect, but did not have a signiﬁcant effect when testing the package type effect.

COPYright by
Jonathan Ryan Thomas Falls
2006

 

 

 

ACKNOWLEDGEMENTS

I would like to thank my committee: Dr. Robb Clarke and Dr. Laura Bix of the
School of Packaging, and Dr. Frederick Rodammer of the Eli Broad College of Business.
I would especially like to thank Dr. Clarke for introducing me to the ﬁeld of RF ID and
allowing me to gain my master’s degree under his guidance.

I would like to thank Mr. Frank Reitano for Ms time and effort in helping me
conduct this research. I would also like to thank Xuewen Huang for his statistical
consultations and guidance. I also appreciate the editing and proof reading efforts of
Renate Snider and Stuart Wilson, my NSC 840 instructors.

To my RFID friends who have helped me along the way: TJ, Nav, Tom, Cortney,
Don, Bryan, Reed, and Craig, you have all contributed to my thesis in one way or
another, and I thank you for that.

To Nina, thank you for always being supportive, understanding, and loving
throughout the thesis writing process; you unfailingly provided me with motivation to
ﬁnish what I needed to do.

Finally, I’d deeply grateful to my parents, Tom and Judy Falls, for supporting me
in my journey through graduate school. Thanks go to Dad for ﬁnding a thesis test
location for my experimental work and in helping me complete my research. Thanks go

to Mom for telling Dad to shower when we got home.

iv

 

 

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. ix
LIST OF FIGURES .......................................................................................................... xv
CHAPTER 1 — INTRODUCTION ..................................................................................... 1
CHAPTER 2 — LITERATURE REVIEW .......................................................................... 6
The Electronic Product Code .......................................................................................... 6
Generations ..................................................................................................................... 7
RFID Tag Classiﬁcations .............................................................................................. 10
RFID in the United States ............................................................................................. 11
The Wal-Mart Effect ..................................................................................................... 12
What’s the Business Value of Reading or Not Reading a Tag? ................................... 14
Meeting the Mandate .................................................................................................... 15
The Department of Defense .......................................................................................... 18
The Food and Drug Administration .............................................................................. 18
European Adoption: ...................................................................................................... 20
European Retailers: ....................................................................................................... 21
RF ID in Asia: ................................................................................................................ 23
RF ID Read Locations: .................................................................................................. 26
Types of Conveyors ...................................................................................................... 27
Conveyor Use and Research ......................................................................................... 28
Previous Testing and Research Gap .............................................................................. 29
CHAPTER 3 - METHODOLOGY .................................................................................. 32
Materials ........................................................................................................................ 32
Product Effect ........................................................................................................... 32
Package Effect ........................................................................................................... 35
Package Shape Effect ................................................................................................ 37
Equipment and Software: .............................................................................................. 39
Procedure for Determining Optimal Tag Location ....................................................... 43
Procedure for Testing RFID Tagged Case Loads on Conveyor ................................... 46
Statistical Analysis ........................................................................................................ 49
CHAPTER 4 — RESULTS ................................................................................................ 50
Hotspot Test Results ..................................................................................................... 50
Product Effect Results from Hotspot Testing ........................................................... 51
Package Effect Results From Hotspot Testing ......................................................... 54
Package Shape Effect Results From Hotspot Testing ............................................... 57
Statistical Analysis of Conveyor Effect ........................................................................ 62

Terminology .................................................................................................................. 63

Product Effect Results: .................................................................................................. 64
Ketchup ..................................................................................................................... 64
Motor Oil ................................................................................................................... 65

Statistical Analysis of the Product Effect Test .............................................................. 66
Ketchup-Ketchup Interactions .................................................................................. 78
Ketchup Oil Interactions ........................................................................................... 81
Oil-Oil Interactions ................................................................................................... 87

Package Effect Results .................................................................................................. 90
Chips in Plastic Tubs ................................................................................................. 90
Chips in MSWFC ...................................................................................................... 90

Statistical Analysis of the Package Effect Test ............................................................. 91
Metal-Metal Interactions ......................................................................................... 103
Metal-Plastic Interactions ....................................................................................... 106
Plastic-Plastic Interactions ...................................................................................... 112

Package Shape Effect Results ..................................................................................... 115
Aluminum Cans: ..................................................................................................... 115
Aluminum Bottles: .................................................................................................. 115
Aluminum Tins: ...................................................................................................... 116

Statistical Analysis of the Package Shape Effect Test ................................................ 117
Bottle-Bottle Interaction ......................................................................................... 135
Can-Can Interaction ................................................................................................ 138
Tin-Tin Interaction .................................................................................................. 141
Bottle-Can Interactions ........................................................................................... 144
Bottle-Tin Interactions ............................................................................................ 149
Can-Tin Interactions ............................................................................................... 154

CHAPTER 5 — CONCLUSIONS AND RECOMMENDATIONS ................................ 160

Hypothesis 1. Conveyor speed will not have an inﬂuence on the average amount of tag

reads per trial for RFID transponders. ........................................................................ 160

The Effect of Conveyor Speed on the Product Effect Test, Ketchup and Motor Oil . 161
Two-Way Interactions ............................................................................................. 161

Product by Speed ................................................................................................ 161
Tag Type by Speed ............................................................................................. 162
Three-Way Interactions ........................................................................................... 163
Ketchup-Ketchup Interactions ............................................................................ 163
Ketchup-Motor Oil Interactions .......................................................................... 165
Motor Oil-Motor Oil Interactions ....................................................................... 167

The Effect of Conveyor Speed on the Package Effect Test, Chips in Plastic Tubs and

Chips in MSWFC: ....................................................................................................... 169
Two-Way Interactions ............................................................................................. 169

Product by Speed ................................................................................................ 169
Tag Type by Speed ............................................................................................. 170
Three-Way Interactions ........................................................................................... 171
MSWFC-MSWFC Interactions ........................................................................... 171

vi

MSWFC—Plastic Tub Interactions ....................................................................... 173

Plastic Tub-Plastic Tub Interactions ................................................................... 175
The Effect of Conveyor Speed on the Package Shape Effect Test, Metal Bottles, Metal
Cans, and Metal Tins .................................................................................................. 177
Two-Way Interactions ............................................................................................. 177
Product by Speed ................................................................................................ 177
Tag Type by Speed ............................................................................................. 178
Three-Way Interaction ............................................................................................ 179
Bottle-Bottle Interaction ..................................................................................... 179
Can-Can Interaction ............................................................................................ 179
Tin-Tin Interaction .............................................................................................. 181
Bottle-Can Interactions ....................................................................................... 1 83
Bottle-Tin Interactions and Can-Tin Interaction ................................................. 185
Hypothesis 2. Product will not have an inﬂuence on the average amount of tag reads
per trial for RFID transponders on packages moving on a conveyor. ........................ 185
The Product Effect Test, Ketchup and Motor Oil ....................................................... 186
Two-Way Interactions ............................................................................................. 186
Product by Speed ................................................................................................ 186
Product by Tag Generation ................................................................................. 186
Three-Way Interactions ........................................................................................... 187
Ketchup-Motor Oil Interactions .......................................................................... 187

Hypothesis 3. Packaging materials will not have an inﬂuence on the average amount
of tag reads per trial for RF ID transponders on packages while moving on a conveyor.

..................................................................................................................................... 189
Evaluation of Package Effect Test, Chips in Plastic Tubs and Chips in MSWFC ..... 189
Two-Way Interactions ............................................................................................. 189
Product by Speed ................................................................................................ 189
Product by Tag Type ........................................................................................... 190
Three-Way Interactions ........................................................................................... 191
MSWFC-Plastic Tub Interactions ....................................................................... 191
Hypothesis 4. Package shape will not have an inﬂuence on the average amount of tag
reads per trial for RF ID transponders on packages moving on a conveyor. ............... 193
Cans and Bottles ...................................................................................................... 193
Bottles and Tins ...................................................................................................... 194
Cans and Tins .......................................................................................................... 194
Evaluation of Package Shape Effect Test, Metal Bottles, Metal Cans and Metal Tins
..................................................................................................................................... 195
Two-Way Interactions ............................................................................................. 195
Product by Speed ................................................................................................ 195
Product by Tag Type ........................................................................................... 197
Three-Way Interaction ............................................................................................ 198
Bottle-Can Interactions ....................................................................................... 198
Bottle-Tin Interactions ........................................................................................ 200
Can-Tin Interactions ........................................................................................... 202
Hypothesis 5. Tag type will not have an inﬂuence on the average amount of tag reads
per trial for packages moving on a conveyor. ............................................................. 204

vii

The Effect of Tag Type in the Product Effect Test ..................................................... 205

Two Way Interactions ............................................................................................. 205
Product by Tag Type, Tag Type Only Variable .................................................. 205

Tag Type by Speed, Tag Type Only Variable .................................................... 206
Three-Way Interactions ........................................................................................... 207
Ketchup-Ketchup, Tag Type Only Variable ....................................................... 207
Motor Oil-Motor Oil, Tag Type Only Variable .................................................. 207

The Effect of Tag Type in the Package Effect Test .................................................... 209
Two Way Interactions ............................................................................................. 209
Product by Tag Type, Tag Type Only Variable .................................................. 209

Tag Type by Speed, Tag Type Only Variable .................................................... 209
Three-Way Interactions ........................................................................................... 210
MSWFC-MSWFC, Tag Type Only Variable ..................................................... 210
Plastic Tubs-Plastic Tubs, Tag Type Only Variable ........................................... 210

The Effect of Tag Type in the Package Shape Effect Test ......................................... 212
Two Way Interactions ............................................................................................. 212
Product by Tag Type, Tag Type Only Variable .................................................. 212

Tag Type by Speed, Tag Type Only Variable .................................................... 213
Three-Way Interactions ........................................................................................... 214
Bottle-Bottle , Tag Type Only Variable .............................................................. 214
Can-Can Interaction, Tag Type Only Variable ................................................... 214
Tin-Tin Interaction, Tag Type Only Variable ..................................................... 214
Thoughts and Conclusions .......................................................................................... 216
REFERENCES ................................................................................................................ 219

viii

LIST OF TABLES

Table 1: Tests of Fixed Effects for ProduEffect ................................................. 66
Table 2: Average Number of Tag Reads Per Trial and Difference Between Means, by
Product Type ......................................................................................... 67
Table 3: Average Number of Tag Reads Per Trial and Difference Between Means, by
Conveyor Speed ..................................................................................... 68
Table 4: Average number of tag reads per trial and difference between means by tag
generation ............................................................................................. 68
Table 5: Average Number of Tag Reads Per Trial, Product by Speed ....................... 69
Table 6: Difference in Average Number of Tag Reads Per Trial Between Product by
Speed .................................................................................................. 70
Table 7: Average Number of Tag Reads Per Trial, Product by Tag Generation ............ 72
Table 8: Difference in Average Number of Tag Reads Per Trial Between Product by Tag
Generation ............................................................................................ 73
Table 9: Average Number of Tag Reads Per Trial, Tag Generation by Speed ............. 74
Table 10: Difference In Average Number of Tag Reads Per Trial Between Tag
Generation by Speed ................................................................................ 75
Table 11: Average Number of Tag Reads Per Trial, Product by Tag Generation by Speed
......................................................................................................... 77
Table 12: Difference in Average Number of Tag Reads Per Trial Between Product
(Ketchup) by Tag Generation by Speed .......................................................... 79
Table 13: Difference in Average Number of Tag Reads Per Trial Between Product
(Ketchup and Motor Oil) by Tag Generation by Speed ..................................... 82,83
Table 14: Difference in Average Number of Tag Reads Per Trial Between Product
(Motor Oil) by Tag Generation by Speed ....................................................... 88
Table 15: Tests of Fixed Effects for Package Effect .......................................... 92

Table 16: Average Number of Tag Reads Per Trial and Difference Between Means, by
Product Type ......................................................................................... 93

Table 17: Average Number of Tag Reads Per Trial and Difference Between Means by
Conveyor Speed .................................................................................... 94

Table 18: Average Number of Tag Reads Per Trial and Difference Between Means by
Tag Type ............................................................................................. 94

ix

Table 19: Average Number of Tag Reads Per Trial, Product by Speed ..................... 95
Table 20: Difference in Average Number of Tag Reads Per Trial Between Product by

Speed ................................................................................................... 96
Table 21: Average Number of Tag Reads Per Trial, Product by Tag Generation... ........97
Table 22: Difference In Average Number of Tag Reads Per Trial Between Product by
Tag Type ............................................................................................. 98
Table 23: Average Number of Tag Reads Per Trial, Tag Generation by Speed .......... 100
Table 24: Difference In Average Number of Tag Reads Per Trial Between Tag
Generation by Speed .............................................................................. 101
Table 25: Average Number of Tag Reads Per Trial, Product by Tag Generation by Speed
........................................................................................................ 102
Table 26: Difference in Average Number of Tag Reads Per Trial Between Product
(MSWFC) by Tag Generation by Speed ........................................................ 104
Table 27: Difference in Average Number of Tag Reads Per Trial Between Product
(Plastic Tubs and MSWF C) by Tag Generation by Speed .............................. 107,108
Table 28: Difference In Average Number of Tag Reads Per Trial Between Product
(Plastic Tubs) by Tag Generation by Speed ................................................... 113
Table 29: Tests of Fixed Effects for Product Shape Effect .................................. 117
Table 30: Average Number of Tag Reads Per Trial by Product Type ...................... 118
Table 31: Difference Average Number of Reads Per Trial by Product Type .............. 118
Table 32: Average Number of Tag Reads Per Trial and Difference Between Means by
Conveyor Speed .................................................................................... 119
Table 33: Average Number of Tag Reads Per Trial and Difference Between Means by
Tag Type ............................................................................................ 120
Table 34: Average Number of Tag Reads Per Trial, Product by Speed .................... 120
Table 35: Difference in Average Number of Tag Reads Per Trial Between Product by
Speed ........................................................................................... 122,123

Table 36: Average Number of Tag Reads Per Trial, Product by Tag Generation. . . . . ....126

Table 37: Difference In Average Number of Tag Reads Per Trial Between Product by
Tag Type ............................................................................................ 128

Table 38: Average Number of Tag Reads Per Trial, Product by Tag Generation. . . . . 1 31

 

Table 39: Difference In Average Number of Tag Reads Per Trial Between Tag Type by
Speed ................................................................................................ 132

Table 40: Average Number of Tag Reads Per Trial, Product by Tag Generation by
Speed ................................................................................................ 134

Table 41: Difference in Average Number of Tag Reads Per Trial Between Product
(Bottles) by Tag Generation by Speed .......................................................... 136

Table 42: Difference in Average Number of Tag Reads Per Trial Between Product (Cans)
by Tag Generation by Speed ..................................................................... 139

Table 43: Difference in Average Number of Tag Reads Per Trial Between Product (Tins)
by Tag Generation by Speed ..................................................................... 142

Table 44: Difference in Average Number of Tag Reads Per Trial Between Two Products
(Bottles and Cans) by Tag Generation by Speed ......................................... 145,146

Table 45: Difference in Average Number of Tag Reads Per Trial Between Two Products
(Bottles and Tins) by Tag Generation by Speed .......................................... 150,151

Table 46: Difference in Average Number of Tag Reads Per Trial Between Two Products
(Cans and Tins) by Tag Generation by Speed ............................................ 155,156

Table 47: Average Advantage in Tag Reads at 300 fpm Over 600 fpm .................... 161

Table 48: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 162

Table 49: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed .................................................................................. 162

Table 50: Ketchup-Ketchup Three-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 164

Table 51: Ketchup-Motor Oil Three-Way Interaction, Difference in Number of Tag
Reads Caused by Speed ........................................................................... 166

Table 52: Motor Oil-Motor Oil Three-Way Interaction, Difference in Number of Tag
Reads Caused by Speed .......................................................................... 168

Table 53: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 170

Table 54: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 170

Table 55: MSWFC-MSWFC Three-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 171

Table 56: MSWFC-Plastic Tubs Three-Way Interaction, Difference in Number of Tag
Reads Caused by Speed ........................................................................... 174

xi

Table 57: Plastic Tubs-Plastic Tubs Three-Way Interaction, Difference in Number of Tag

Reads Caused by Speed ........................................................................... 176
Table 58: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 178
Table 59: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Speed ................................................................................... 179
Table 60: Bottle-Bottle Three-Way Interaction, Difference in Number of Tag Reads
Caused by Speed .................................................................................. 179
Table 61: Can-Can Three-Way Interaction, Difference in Number of Tag Reads Caused
by Speed ............................................................................................. 180
Table 62: Tin-Tin Three-Way Interaction, Difference in Number of Tag Reads Caused by
Speed ................................................................................................ 181
Table 63: Bottle-Can Three-Way Interaction, Difference in Number of Tag Reads Caused
by Speed ............................................................................................. 184
Table 64: Average Number of Tag Reads Per Trial and Difference Between Means, by
Product Type ....................................................................................... 185
Table 65: Difference in Average Number of Tag Reads Per Trial Between Product by
Speed ................................................................................................ 186
Table 66: Difference in Average Number of Tag Reads Per Trial Between Product by
Tag Generation ..................................................................................... 187
Table 67: Difference in Average Number of Tag Reads Per Trial Between Product
(Ketchup and Motor Oil) by Tag Generation (Constant) by Speed (Constant) ............ 188
Table 68: Difference in Average Number of Tag Reads Per Trial Between Product
(Ketchup and Motor Oil) by Tag Generation by Speed (Constant) ......................... 188
Table 69: Average Number of Tag Reads Per Trial and Difference Between Means, by
Product Type ....................................................................................... 189
Table 70: Difference in Average Number of Tag Reads Per Trial Between Product by
Speed ................................................................................................ 190
Table 71: Difference In Average Number of Tag Reads Per Trial Between Product by
Tag Type ............................................................................................ 191
Table 72: Difference in Average Number of Tag Reads Per Trial Between Product
(Plastic Tubs and MSWFC) by Tag Type (Constant) by Speed (Constant) ............... 192
Table 73: Difference in Average Number of Tag Reads Per Trial Between Product
(Plastic Tubs and MSWFC) by Tag Generation by Speed (Constant) ..................... 192
Table 74: Difference Average Number of Reads Per Trial by Product Type .............. 193

xii

Table 75: Difference in Average Number of Tag Reads Per Trial Between Product by
Speed ................................................................................................ 196

Table 76: Difference in Average Number of Tag Reads Per Trial Between Product by
Tag Type ............................................................................................ 198

Table 77: Difference in Average Number of Tag Reads Per Trial Between Product
(Bottles and Tins) by Tag Generation by Speed, Where Product Was Not the Dominant
Variable ............................................................................................. 199

Table 78: Difference in Average Number of Tag Reads Per Trial Between Product
(Bottles and Tins) by Tag Generation by Speed, Where Product Was Not the Dominant
Variable ............................................................................................. 201

Table 79: Difference in Average Number of Tag Reads Per Trial Between Product (Cans
and Tins) by Tag Generation by Speed, Where Product Was Not the Dominant Variable

........................................................................................................ 203
Table 80: Average Difference Between Number of Tag Reads Per Trail for Gen 1 Tag
versus Gen 2 Tag .................................................................................. 204
Table 81: Product-Tag Type Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 206
Table 82: Product-Tag Type Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 206
Table 83: Ketchup-Ketchup Three-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 208
Table 84: Motor Oil-Motor Oil Three-Way Interaction, Difference in Number of Tag
Reads Caused by Tag Type ....................................................................... 208
Table 85: Product-Tag Type Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 209
Table 86: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 210
Table 87: MSWFC-MSWFC Three-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 211

Table 88: Plastic Tubs-Plastic Tubs Three-Way Interaction, Difference in Number of Tag
Reads Caused by Tag

Table 89: Product-Tag Type Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 213

Table 90: Two-Way Interaction Between Case of Bottles With Gen 2 Tag versus Case of
Cans With Gen 1 Tag .............................................................................. 213

xiii

 

Table 91: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag Reads

Caused by Tag Type .............................................................................. 213
Table 92: Bottle-Bottle Three-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type .............................................................................. 215
Table 93: Can-Can Three-Way Interaction, Difference in Number of Tag Reads Caused
by Tag Type ........................................................................................ 215
Table 94: Tin-Tin Three-Way Interaction, Difference in Number of Tag Reads Caused by
Tag Type ............................................................................................ 215
xiv

 

 

LIST OF FIGURES
Images in this thesis are presented in color

Figure 1: ALL-9440 “Gen 2 Squiggle TM” Class 1 Tag with Description of Parts ........... 10
Figure 2: (A) Case of Ketchup, (B) Lay-out of Bottles in Case, (C) Individual Bottle... 33
Figure 3: (A) Case of Motor Oil, (B) Lay-out of Bottles in Case, (C) Individual Bottle. 34
Figure 4: (A) Case of Potato Chips, (B) Lay-out of Tubs in Case, (C) Individual Tub 35

Figure 5: (A) Case of Potato Chips, (B) Lay-out of MSWFC in Case, (C) Individual
MSWFC ............................................................................................................................ 36

Figure 6: (A) Tray of Bottles, (B) Lay-out of Bottles in Tray, (C) Individual Bottle ...... 37

Figure 7: (A) Case of Cans, (B) Lay-out of Cans in Case, (C) Individual Can ................ 38
Figure 8: (A) Case of Tins, (B) Lay-out of Tins in Case, (C) Individual Tin ................... 39
Figure 9: ALR-9780 Reader ............................................................................................. 40
Figure 10: ALR-9610-AC Antenna .................................................................................. 40
Figure 11: (A) Back of Gen 1 Tag (top) and Gen 2 Tag (bottom). (B) Front of Gen 1 Tag
(top) Gen 2 Tag (bottom) .................................................................................................. 41
Figure 12: Antennae Locations Surrounding Conveyor ................................................... 42
Figure 13: Case and Anna Set-up for Cape Systems RFID Tag Locator Soﬁware .......... 44
Figure 14: Example Grid Placed on Front Face of Case ................................................... 45
Figure 15: Example Tag Location on Package Face ......................................................... 46
Figure 16: Warehouse with Conveyor and RFID Set-up .................................................. 48
Figure 17: Example RF ID'hotspot test outcome for one side of a case ............................ 51
Figure 18: Tag Location on Ketchup Case ....................................................................... 52
Figure 19: Hotspot Test Output for Case of Ketchup, (A) Length and Width View of
Case, (B) Length View of Case, (C) Width View of Case ............................................... 52
Figure 20: Tag Location on Motor Oil Case ..................................................................... 53
Figure 21: Hotspot Test Output for Case of Motor Oil, (A) Length and Width View of
Case, (B) Length View of Case, (C) Width View of Case ............................................... 54
Figure 22: Tag Location on Case of Chips in Plastic Tubs .............................................. 54

XV

 

 

Figure 23: Hotspot Test Output for Case of Potato Chips in Plastic Tubs, (A) Length and

Width View of Case, (B) Length View of Case, (C) Width View of Case ...................... 55
Figure 24: Tag Location on Case of Chips in MSWFC .................................................... 56
Figure 25: Hotspot Test Output for Case of Potato Chips in MSWFC, (A) Length and
Width View of Case, (B) Length View of Case, (C) Width View of Case ...................... 56
Figure 26: Tag Location on Case of Cans ......................................................................... 57
Figure 27: Hotspot Test Output for Case of Cans, (A) Length and Width View of Case,
(B) Length View of Case, (C) Width View of Case ......................................................... 58
Figure 28: Tag Location for Tray of Bottles ..................................................................... 59
Figure 29: Hotspot Test Output for Tray of Bottles, (A) Length and Width View of Case,
(B) Length View of Case, (C) Width View of Case ......................................................... 60
Figure 30: Tag Location for Case of Tins ......................................................................... 61

Figure 31: Hotspot Test Output for Case of Tins, (A) Length and Width View of Case,
(B) Length View of Case, (C) Width View of Case ......................................................... 62

xvi

 

CHAPTER 1 — INTRODUCTION

Radio Frequency Identiﬁcation (RFID) technology is a tool that enables
companies to efficiently track products in their supply chain. From raw material through
the entire life of the product, RFID can provide valuable information regarding inventory
data, shipment locations, and even product temperature.

RFID is a data collection technology that is able to communicate information
from 3 tagged item to a computer system. For an RFID system to function properly it
must be equipped with four items: a reader (also known as an interrogator), an antenna, a
computer equipped with the proper software, and an RFID transponder (tag) placed on an
item. Radio waves transfer data between the RFID tag and the reader, which are tuned to
the same frequency.

The use of RFID in supply chain applications is currently organized by a
worldwide standards organization known as EPCglobal. Undoubtedly one of the most
important achievements of this organization was the completion of RFID Ultra-High
Frequency (UHF) Generation 2 (Gen 2) protocol, which solved the readability problems
associated with Generation 1. This protocol made it possible to read any UHF RFID tag
using any UHF RF ID equipment. Previously, when using Generation 1 protocol, tag
readability was dependent on using one of the two main passive UHF classes, Class 0 or
Class 1. A Class 1 tag could not be detected on a Class 0 reader/antenna set up and vice
versa.

In the United States, RF ID use is currently monitored by the Federal

Communications Commission (FCC). The FCC determines both the frequency

bandwidth and the power level allocated for use. These allocations have a direct effect
on equipment performance and effectiveness.

The use of RFID for supply chain applications has grown immensely since the
Wal-Mart mandate to their suppliers became effective January 2005. The mandate stated
that the company’s top 100 suppliers had to begin shipping select products headed for
particular distribution centers (DCs) with RF ID tags on each case and pallet. The reason
for the mandate was simple: by improving product availability on the store shelves, and
by being able to track the whereabouts of expected deliveries, Wal-Mart can improve
store operations and increase proﬁts. The Department of Defense (DOD) followed Wal-
Mart’s mandate with similar RFID requirements for their suppliers, recognizing the
technology’s ability improve product visibility especially in tracking dangerous and
expensive supplies. Additionally, the Food and Drug Administration (FDA) has issued
statements to both food retailers and the pharmaceutical industry asserting the desire for
improved security within supply chains, to help prevent possible bio-terrorism attacks
and to effectively recall products in the event of an emergency.

The use of RFID technology is by no means limited to the United States, as
numerous retailers in countries around the world are also using RFID for supply chain
applications. Companies in Europe have issued mandates similar to those of Wal-Mart
and the DOD to their own suppliers. Tesco, Marks and Spencer, and Metro Group are
some of Europe’s largest retailers working with RFID technology in their stores and DCs.
Additionally, Asian retailers and organizations throughout Japan, Singapore, South

Korea, and China are working with RF ID systems.

 

As is the case in the United States, both retailers and suppliers overseas are
learning about some of the set backs inherent in RF ID technology. For some
implementers, RFID performance is hampered by regulations limiting the operation and
frequency available for RFID readers. The frequency band allocated for UHF RFID
differs from country to country and the amount of availability between 866 MHz to 956
MHz UHF band can have drastic system performance implications. Others working with
the technology may ﬁnd that their product or package absorbs or reﬂects radio waves,
which can hamper their ability to meet mandates effectively, as well as their ability to use
all the beneﬁts RF ID can offer.

There is no one speciﬁc method for using RF ID technology, nor is there one
speciﬁc solution to be applied across industries. For RFID to work most advantageously
within an organization’s supply chain, the implementer must think of each product on an
individual level. An RF ID tagged product, Product A will not function equally to a
tagged Product B if there are differences in the product composition and the package
system. Additionally, optimal tag type, tag location and orientation, antenna location and
orientation, reader location and broadcast strength, and even the cord length between the
reader and the antennae, are among the variables that implementers of the technology
need to consider when trying to optimize the readability of an RF ID tagged product.

In the event that a retailer has implemented RFID mandates to suppliers, both
shipper and receiver must work collectively to optimize performance and usefulness in
the supply chain. Speciﬁcally, the read location (where the tag is to be detected) can
have a signiﬁcant impact on successful RF ID utilization. Suppliers will have to ensure

that they have tested their product at each read point in the supply chain process in order

to avoid failing to meet mandates which could lead to ﬁnancial losses. Traditional read
locations for RFID tags are warehouse dock doors, stretch wrappers and fork trucks, or
conveyors. This thesis focuses on conveyors, since retailers such as Wal-Mart use UHF
RFID systems to track product traveling on conveyors in their DCs from arrival, to
storage, and to their eventual exit, bound for their destination point. With product
traveling upwards of 600 feet per minute on conveyor lines, the amount of time for
interaction between the mounted RFID antennae and the tagged cases is minimal. The
purpose of this research is to determine the effect of conveyor speed on the readability of
RFID tagged case goods in a typical warehouse environment. The null hypotheses for
this research are:
0 Conveyor speed will not have an inﬂuence on the average amount of tag reads per
trial for RFID transponders
0 Product will not have an inﬂuence on the average amount of tag reads per trial for
RF ID transponders on packages moving on a conveyor
0 Packaging materials will not have an inﬂuence on the average amount of tag reads
per trial for RFID transponders on packages while moving on a conveyor.
0 Package shape will not have an inﬂuence on the average amount of tag reads per
trial for RFID transponders on packages moving on a conveyor
0 Tag type will not have an inﬂuence on the average amount of tag reads per trial
for packages moving on a conveyor.
Although the effects of package, product and type of tag on read rates have been
relatively well documented, the effect of conveyor speed, and its potential interaction

with these variables has not been researched in detail. The outcome of this research will

 

provide RFID users, whether they are implementing the technology into their own supply
chain or merely meeting the mandates of retailers, with valuable information to consider

when working with RFID tagged product on conveyors.

CHAPTER 2 — LITERATURE REVIEW
The Electronic Product Code

One of the most important advances aiding RF ID technology and its use in
industry was the development of the Auto-ID center. The Auto-ID center was a “non-
proﬁt collaboration between private companies and academia that pioneered the
development of an Intemet-like infrastructure for tracking goods globally through the use
of RFID tags carrying Electronic Product Codes [EPC]”.“] When the Auto-ID center
closed in September 2003, EPCglobal, a non-proﬁt organization, was set up to continue
the work of developing the use of RF ID to produce more visibility and efficiency
throughout the supply chain. EPCglobal is achieving this goal through the Electronic
Product Code Network”. This network is to serve “as the global standard for
immediate, automatic, and accurate identiﬁcation of any item in the supply chain of any
company, in any industry, anywhere in the world”. [2]

The EPC is the primary information of concern stored on the RF ID tag’s
microchip. Used for recognition, the EPC assigns a numeric identiﬁcation to each
packaging unit whether it is an item, case, or pallet. The number used in the EPC
consists of four parts: a Header, a Manager Number, an Object Class, and a Serial
Number. The Header identiﬁes length, type, structure, version, and generation of EPC.
The Manager Number identiﬁes the company that owns the product. The Object Class
Number represents the stock-keeping unit (SKU), while the Serial Number is assigned to
identify each individual case. [3] The EPC by itself gives no more information about a
product than a car’s license plate tells you about the car. To decode information

contained in a particular EPC the computer is directed to information located at an

intemet address. The Object Name Service (ONS) is an automated networking service
that points computers to sites on the World Wide Web. Once the information is located

it can be forwarded to a company’s inventory or supply chain datam

Generations

Assembled by more than 60 of the world’s leading technology companies,
EPCglobal recently presented an RFID UHF Gen 2 protocol. This protocol describes the
core capabilities required to meet the performance needs set by the end user
community”

Until the recent ratiﬁcation of a Gen 2 standard, various vendors (and their end
user customers) had adopted incompatible EPC technologies (EPC Version 1), including
EPC Class 0 and Class 1.5] There are two separate components in a tag class: the air
protocol (how the tags communicate) and the programming technique Grow the tags get
their data). Class 0 and Class 1 tags use different methods in their approach to RFID
communication, and thus have different performance capabilities.[6] Class 0 tags use air
protocol developed by Matrics, Inc, (Now Symbol Technologies), and Class 1 tags use air
protocol developed by Alien Technologies. The implications of having two different
protocols are straightforward: simply because a product is equipped with an RFID tag
does not mean that it can be recognized by an RFID system. A Class 0 tag will not read
on a RFID system designed for Class 1 tags. Gen 2 is an Air to Air Protocol that allows
communication between the tag and reader irrespective of the equipment manufacturer.

The beneﬁt of Gen 2 is that it offers speciﬁcations and regulations that can be

applied across the world. Gen 2 offers RFID users the assurance that no matter which

type of tag comes through their doors, their readers will be able to detect it. The new
EPC Gen 2 standard supports an increased frequency range and regulatory requirements
promoting adoption in the US, Europe and Asia. Gen 2 includes password-protected
mechanisms to ensure data safety and a kill feature to ensure consumer privacy
protection. [7] Gen 2 has the ability to work in dense reader environments, making it
optimal for distribution centers loaded with inventory. Additionally, Gen 2 allows users
to read and write data multiple times to the same RF ID tag. [8]

It is thought that with a global standard set, Gen 2 will create more competition in
the market place, hence lowering the price of RFID equipment and tags. Additionally, in
some cases, costs of RF ID materials are decreasing. “Physically, the silicon [with Gen 2]
becomes smaller, and that’s one of the assets that makes it cheaper,” said Scott Medford,
vice president of RFID at Interrnec Technologies Corp., an RFID hardware and software
manufacturer that participated in Gen 2 speciﬁcation development. With Gen 2 comes
more than just cost beneﬁts. “It has a lot more selectivity, which means that there is less
reader on reader interference. In addition, claims have been made hyping Gen 2 to be ﬁve
to 10 times faster than Gen 1”. [8] In most cases Gen 2 readers are capable of reading
Class 0, Class 1 and Gen 2 tags, easing the transition for those moving away from Gen 1
systems.[91

Although Gen 2 offers many positive features necessary for a global RF ID
deployment, some companies are reluctant to move on from Gen 1 at short notice. “In
practical applications, Gen 1 and Gen 2 hardware performance will be comparable. Gen
2 has gained a lot of intrigue and interest, but it hasn’t changed the ﬁlndamental laws of

physics. It is still UHF, with all the beneﬁts and drawbacks that that brings,” says Tom

Pounds, vice president of corporate development at tag and reader maker Alien
Technologies!”

Another avid participant in the Gen 2 speciﬁcation development, Impinj, a
Seattle— based chip manufacturer, states that Gen 2 buyers need to be aware of what type
of reader is right for their application. EPCglobal plans to certify three different levels of
Gen 2 compliant readers. “At the lowest level, readers will be certiﬁed to work only
when there are no other readers within a 1 km radius. The next level will be for readers
capable of being deployed with several readers within a 1 km radius. The highest level
will be certiﬁed to work alongside 50 or more readers within a 1 km radius”. Irnpinj’s
founder and chairman, Chris Diorio, is convinced that users have to get educated that not
all EPC compliance is the same. If you can upgrade only to the lowest grade, you are not
going to get multi-reader performance.“01

The Gen 2 standard put forth by EPCglobal is royalty-free. The organization
engaged legal counsel to examine claims made by Intermec Technologies, a RFID
systems provider, which claims the Gen 2 spec contains intellectual property that it has
patented. It was concluded that Interrnec’s patents are not essential to implementing the
Gen 2 standard. However, Intermec President Tom Miller concluded, “it is important to
remember the claim of a royalty-free protocol does not mean the UHF RFID products
will be royalty-free, we believe companies who offer UHF RFID products will still
require a license to use Intermec intellectual property” .[11] It is the hope of EPCglobal
and RFID companies alike that possible intellectual property battles with Intermec will

not slow down the development of Gen 2 technology and implementation.

 

RFID Tag Classiﬁcations

A wide variety of RFID tag types currently exist, each with different capabilities.
Although current RF ID mandates focus on passive RFID tags, other tag types such as
semi-passive and active tags exist. As opposed to passive tags, semi-passive and active
tags contain their own source of power, a battery. These types of tags tend to be used for
tracking large or expensive items. In contrast, passive tags are powered by the
electromagnetic waves transmitted from the RF ID reader. These RF waves carrying data
induce a current in the tag’s antenna, powering up the accompanying microchip which
contains information that is sent back to the reader (Figure 1). Passive tags are currently
being used for case and pallet application for inventory tracking purposes, but can also be
used for tracking children in amusement parks, skiers on mountains, luggage in airports,

and sports race timing. “21

Microchip

 

Figure l: ALL-9440 “Gen 2 Squiggle TM” Class 1 Tag with Description of Parts
Multiple variations of passive tags exist. To help understand an RFID tag’s
capabilities, the tags are placed into certain generations and classes. Gen 1 Class 0 RFID
tags are the most basic passive tags, arriving to the end user factory programmed and in a
read-only format. “Read-only is a term used to describe RFID tags that contain data that

cannot be changed unless the microchip is reprogrammed electronically”. I” Gen 1 Class

10

 

0+ and Class 1 RF ID tags are known as Write Once Read Many (WORM) passive tags,
which allow end users to program the tag as opposed to the tag arriving preprogrammed.
Gen 1 Class 0, 0+, and 1 tags have all been approved for meeting the mandates put forth
by Wal-Mart and the DOD. Gen 1 Class 0 tags can contain either 64 or 96 bits of
memory, while Gen 1 Class 0+ contain 96 bits of memory. Gen 1 Class 1 tags contain 96
bits of user programmable memory. “3’ 14] With the current movement towards Gen 2 it
is important to note the emergence of Gen 2 tags. Gen 2 tags offer 128 bits of user

programmable memoryml

RFID in the United States

RFID technology and its use in the United States is monitored and regulated by
the FCC. The FCC has set aside rules for the use of RFID in the Code of Federal
Regulations, speciﬁcally 47 CFR Part 15, which is reserved for low power devices.
Since these devices are low power, owners and operators do not require a license to
operate the machinery. However, it is required that RFID readers must meet the FCC’s
emissions limitations and power restrictions. The FCC classiﬁes RFID readers as
intentional radiators, and therefore they require certiﬁcation by the FCC. This
certiﬁcation process is generally performed by the manufacturer of the reader. After the
certiﬁcation process the reader is properly labeled and can be marketed and operated in
the US. “5]

Should the purchaser of the certiﬁed reader make changes to its operational
capabilities, FCC rules state that the modifying party automatically becomes responsible

for complying with the FCC standards. Should the modifying party make changes to the

11

 

 

power level or otherwise alter the equipment, they are subject to sanctions and monetary
forfeitures if the reader is not recertiﬁed.

The certiﬁcation process is important because it ensures the performance of RFID
technology from being hampered by readers interfering with one another. The more
unlicensed devices in operation, the more likely it is that interference will occur. To
encourage licensing, the FCC has imposed sizeable monetary ﬁnes on various parties
who operated noncompliant Part 15 devices.['6]

UHF RFID systems operating in the US. use a frequency allocation between 902
MHz and 928 MHz, providing 26 MHz of bandwidth. UHF readers are allowed to
operate at 1 watt and can go up to 4 watt if they have directional antennae and if they hop
across at least 50 channels.“51

With the use of RFID increasing in the US, both federal and state governments
have introduced RFID privacy bills. Most bills deal with requiring retailers to notify

customers when RF ID tags are on the products they are purchasing, and to remove tags at

the point of sale.

The Wal—Mart Effect

Wal-Mart was the ﬁrst retailer to realize the possible cost savings that could be
attained by using RF ID technology in its supply chain and distribution centers. In June of
2003, Wal-Mart mandated that its top 100 suppliers would have to ship selected pallets
and cases RFID-tag-equipped beginning January 2005. In this trial run, Wal-Mart
selected 3 of its 99 US. distribution centers to receive these tagged shipments. [17] By

late 2005, the company extended its trial run to include a total of ﬁve distribution centers.

12

Later, the mandate was expanded to include an additional 200 suppliers to ship their
pallets and cases RF ID-tag-equipped by January 2006.

The product receiving process begins at Wal-Mart’s regional distribution centers.
When tagged cases and pallets arrive they are scanned by a reader and antennae set up
near the distribution center’s dock doors. Data about the shipment are collected and sent
to Wal-Mart’s operations and merchandising teams. Additionally, the supplier is notiﬁed
within 30 minutes, through a website that links retailers to real time data, that a speciﬁc
shipment has arrived.“81 The next step that occurs at the distribution center is pallet
disassembly. Cases are removed from their pallets, placed onto conveyors, sent through
another RFID read point on the conveyor line, and placed into storage. When product is
needed at a Wal-Mart retail center, order pickers in DCs gather the required cased goods,
and place them back onto conveyors at the end of which products are re-palletized and
shipped out via truck. This re-palletizing of a variety of products allows the Wal-Mart
retail center placing the order to receive exactly what they need for restocking purposes.
Wal-Mart aims to read 100 percent of all tagged pallets entering through distribution
center and store dock doors, as well as 100 percent of all tagged cases on conveyors
within the distribution centers.“91

After more than a year of receiving tagged shipments from suppliers, Wal-Mart
determined that the technology provided a 16 percent reduction in out-of-stock
merchandise and a 70 percent drop in the time it takes to receive new shipments from
suppliers. The key to the vast improvements arose from in-store inventory tracking.

Prior to receiving RFID tagged shipments, knowing when to restock shelves at Wal-Mart

was based on visual observation. Now, Wal-Mart associates receive shelf restocking data

13

 

that are linked to real time product sales. Ensuring that Wal-Mart stores are receiving the
desired product ﬁ'om their distribution center is critical to avoiding out—of-stocks and
empty shelves. Using RFID technology allows Wal-Mart to know speciﬁc details about
when product arrives at their distribution centers, and how long it takes for the product to
be redirected. “Wal-Mart was able to determine that a particular product arrived at its
distribution center on Aug. 4, that it was put on the conveyor system ﬁve days later and
that it departed shortly thereafter. Upon arrival at the store (12 hours after it left the
distribution center), the product was whisked to the store's back room and moved to the
sales ﬂoor the following day”.[2°] This type of information could help improve inventory
turnover, determine distribution center efﬁciency, and track bottlenecks in the supply
chain.

With the arrival of Gen 2 equipment and tags, Wal-Mart has decided that as of
July 2006 they will no longer accept EPC Gen 1 tags on cases and pallets received from
suppliers. Internal tests performed at Wal-Mart determined that EPC Class 1 Gen 2 tags
showed improved performance compared to their Gen 1 predecessors. Wal-Mart polled
suppliers and concluded that most could be ready to deploy Gen 2 tags by the second

quarter of 2006.911

What’s the Business Value of Reading or Not Reading a Tag?

The business-value of reading (or detecting) tags at checkpoints throughout the
supply chain are numerous. For the supplier merely meeting the mandates, tag
readability on both pallet and case loads are critical to ensure they receive payment for
their shipments. If the retailer doesn’t know it has received a supplier’s shipment (the

RFID tag on the pallet or case load isn’t detected), the retailer doesn’t know it has

14

received any product, and therefore, would not likely pay for the product. Additionally,
tag reads are crucial for the retailer because it makes them aware of product availability,

and therefore, can help to avoid dreaded out-of-stock situations.

Meeting the Mandate

Suppliers to retailers such as Wal—Mart have more often than not applied a “slap
and ship” approach to tagging their products with RFID tags. In this approach, placing
RFID tags on the case and pallet is done in the ﬁnal stages of the manufacturing and
distribution process. With this “slap and ship” approach, at no point does the supplier
attempt to use the RFID-tagged goods for their own internal purposes, thus creating
nothing more than an additional cost to the supplier. This cost ranges from 7.9 cents for
an inlay to 12.9 cents for a self-adhesive tag for orders of 1 million or moreml Smart
suppliers will try to take advantage of the beneﬁts of RF ID, especially by tagging product
early in the manufacturing process.

One such example comes from Paramount Farms, the world’s largest supplier of
almonds and pistachio nuts. Paramount owns 50,000 acres of orchards and processing
facilities, and is responsible for growing about 60 percent of the US. pistachio crop. To
meet its pistachio needs, Paramount also networks with nearly 400 grower partners. “In
our average harvest season, incoming green product totals a half a billion pounds over a
six-week period. Given this time constraint, as we increase our production goals, our
efficiency and productivity must likewise increase,” said Andy Anzaldo, director of

grower relations at Paramount [23].

15

 

After forming a committee within Paramount to brainstorm ideas about
technology features necessary to make the processing job easier, it was determined that
the Grower Receiving System would need to provide multi-site distribution of
information from a central server at Paramount Farms. This could easily be done in a
web based environment, and allows the company’s two pistachio processing plants, as
well as its sales and marketing ofﬁces in Southern California, to access the system with
only an Intemet-compliant browser.

Productivity was enhanced by providing growers with handheld computers,
access points, and RF ID tag readers. Load processing time at Paramount Farms’ pistachio
nut farms was improved by 60 percent. As processing time decreased, Paramount
noticed an increase in revenue. The receiving department became so efficient at
equipment logistics that it reduced leased trailer usage by 30 percent. “We are more
conﬁdent than ever of our data system’s integrity and the accuracy of the information,
since more of the data is collected using radio-frequency tags and barcode scanners,” said
Mr. Anzaldo. "-31

Paramount receives 20 million pounds of product per day for recording, weighing,
pre-cleaning, sampling, and processing. RFID-tagged trailers ﬁlled with pistachios arrive
and are interrogated by a reader. The reader captures the tag’s unique identiﬁcation
number and wirelessly transmits it to the central server. The database relays the pre-
recorded proﬁle of the identiﬁed trailer back to the scale house worker’s mobile
computer. Now the worker knows the trailer’s net weight, license plate number,
equipment number, and owner name. Next, the scale house workers take the product load

details. The grower name, ranch, ﬁeld, product temperature, and harvest method are all

16

 

 

sent wirelessly to the database. The trailer’s gross weight is automatically retrieved from
the truck weigh scale and a weight certiﬁcation is printed. During processing, the nuts
are cleaned. Sifters remove foreign debris such as leaves and branches. The Grower
Receiving System automatically weighs the debris, subtracts it from the original load
weight and sends the corrected weight to the database. [23]
Both weight and quality play a role in the amount a grower gets paid. An
automatic sampler scoops 20 pounds from each 50,000-pound load for quality testing.
While this sample is peeled, hulled, dried and tested, the rest of the load travels on to the
main processing line, where it is mixed with the rest of the day’s harvest. This mass of
nuts is processed and stored within 24 hours. Sample testing determines the grade of the
nut, and the pay rate to the grower. “It’s imperative for us to ensure that the volume and
quality we pay for is the volume and quality we receive. Our new Grower Receiving
System helps us do all these things” [23].
The Paramount Farms RFID implementation is an excellent demonstration of a
supplier employing RFID early in the manufacturing process to attain a return on
investment (ROI). If the Wal-Mart mandate were to include products from Paramount
Farms, the company’s familiarity with the technology and ability to attain a proﬁt from
its use would make tagging shipments not nearly as painful as it would be to a company
utilizing a “slap and ship” approach. In fact, tagging retailer-bound shipments will
provide even more traceability to the Paramount Farms outgoing supply chain, shedding

light on other areas of the product life cycle that could be improved. A likely impact of

incorporating the use of RFID into a supply chain will be RF ID working its way into the

17

 

 

supplier's own vendor supply chain system. This will increase the demand for RFID tags

and result in lowering the cost per tagml

The Department of Defense

The United States DOD also recognized the beneﬁts that RFID technology could
provide in terms of logistics support, asset management, and overall supply chain
optimization. Another advantage is hands-free data capture, which allows efficient
recording of material transactions. In July of 2004, the DOD released requirements to
their suppliers. The Defense Federal Acquisition Regulations Supplement (DFARS)
became effective in November of 2005 and required suppliers to afﬁx passive RF ID tags
at the case and pallet level for shipments of certain commodities to two speciﬁc locations:
Susquehanna, PA, and San Joaquin, CA. The commodities included four classes of
supplies: Shipments of Packaged Operational Rations, Clothing/Equipment/Tools,
Personal Demand Items, and Weapon System Repair Parts?“ In 2006 the tagging
requirements added three more classes of supplies and an additional 19 locations. By

2007 all locations will be instrumented and all classes of supply will require RFID

tagging.

The Food and Drug Administration

The FDA is asking food retailers for help against the war on terrorism by keeping
detailed data about the food shipments in their supply chain. The agency announced that
it is the responsibility of everyone in the food supply chain to keep logs of where they

received food from and where they shipped it to.

18

If contamination in the food supply chain takes place, companies must be able to
make their records available within 24 hours if the FDA has reason to believe that an
article of food presents a serious threat. During his resignation speech, Tommy
Thompson, Secretary of Health and Human Services, may have prompted the new policy
when he made reference to the ease of attacking the country’s food supply. [26]

RF ID stands to play an important role in helping food retailers improve their
supply chain records. “When you take a supply chain like ours, operating close to 100
facilities with products going through multiple facilities, it will obviously require some
additional technology,” says Scott McWilliams, CEO of Ozbum-Hessey Logistics, which
deals heavily in food storage relations. With the ratiﬁcation of Gen 2, RFID solution
providers should be excited about the new FDA mandates. RFID is capable of being the
technology of choice for food retailers, allowing them to meet FDA standards while
offering improved traceability of their products in the supply chain. Food retailers,
manufactures and distributors have until January 2006 to bring their operations into
compliance with the ruling. [27]

The FDA has also been vocal in their desire for the pharmaceutical supply chain
to become more secure, and has endorsed the use RFID to combat the growth of
counterfeit drugs. A Finnish drug maker, Orion Pharma, recently performed a trial
tracking passive tags on the cartons of individual bottles of drugs as they moved through
the supply chain. The test stemmed from the anticipation of stricter policy in the United
States with respect to tracking medication?” The FDA desires that each product moving

through the supply chain have an electronic pedigree (e—pedigree) that shows each

bottle’s chain of custody. Speciﬁcally, an e-pedigree is a secure ﬁle that stores data

19

about each move a product makes through the supply chain, thus helping reduce
counterfeiting of drugs while improving supply chain safety. It is the goal of the FDA
that RFID technology be used widely throughout the pharmaceutical industry by 2007 to
improve security and safety. Until the Orion Pharma trial run, the only pharmaceutical
companies to actively test and report data were Purdue Pharma and Pﬁzer. Purdue
Pharma performed “the pharmaceutical industry’s ﬁrst electronic drug pedigree using
RFID tags to match each bottle of a drug with a corresponding record detailing the drug’s
movement through the supply chain”.[29] Pﬁzer’s use of RFID tags concentrated on
allowing wholesalers and pharmacies to verify that the product they were receiving was

genuine, but did not focus on the tracking aspect utilizing e-pedigrees. [30]

European Adoption:

RFID technology is not limited to suppliers and retailers in the United States. In
Europe, companies like Tesco, Marks and Spencer, and Metro Group have implemented
RF ID technology into their supply chains. European companies have been using RF ID
for tracking reusable containers for years, albeit utilizing both low frequency and high
frequency. Since UHF is the accepted frequency for most pallet and case level supply
chain applications, a bandwidth of 866 MHz to 956 MHz is available for use. Since
regulations governing the use of the radio spectrum differ across the globe, ease of
implementation and use varies accordingly. For UHF applications, the European Radio
Communications Office (ERO) and the European Telecommunications Standard Institute
(ETSI) of the European Union has speciﬁed a range from 865.6 MHz to 867.6 MHz. The
FCC determined that companies in the United States are able to use between 902 MHz

and 928 MHz. Although 26 MHz of bandwidth in the United States versus 2 MHz of

20

bandwidth in Europe may seem insigniﬁcant, “think of data from tags as cars driving on a
two-lane highway in Europe, compared to a 26-lane highway in North America. Simply
put, European companies are not going to get the same performance from their UHF
systems as their North American competitors”.[3 '1 John Clarke, chief technology officer
of Tesco, claims the European deployment of EPC RFID is slowed greatly by regulations
limiting the operation and frequency available for RF ID readers. European readers use a
listen-before-talk function that can limit the time a reader can operate if there is too much
activity or noise in the same radio frequency spectrum. Another aspect that impedes
European RF ID technology functionality are the lower power limits which reduce an
antenna’s read ﬁeld.[321

For US. companies planning to deploy RFID in Europe, testing the reader at the
European frequency can be difﬁcult while on US. soil, since the spectrum is currently
used for police telecommunication. Conversely, European companies testing RF ID
readers operating at the United State’s frequency can encounter interference with wireless
phone handsetsml
European Retailers:

In April 2004 Tesco, the United Kingdom’s largest retailer, started tagging cases
of non-food items at its distribution center, and tracking them to their retail stores. The
company’s approach differs slightly from that of Wal-Mart, who had their suppliers
provide the tagged product for tracking purposes. Tesco plans to have suppliers ship
their goods tagged but has not set a deadline when all suppliers must tag their cases. As

of April 7‘h 2006, 40 of 1400 Tesco stores are equipped with RF ID technology. Tesco

has stated that complications from using UHF RFID under European Union Regulations

21

have slowed its attempts to make full use of the technology. Nevertheless, Tesco’s
research proved to them that RFID could provide “greater supply chain visibility and
simpler processes for its staff, while resulting in improved product availability, better
service and cheaper prices for its customers”.[33]

Marks and Spencer (M&S), a United Kingdom retailer of clothing, food, and
home products, began testing RFID’s capabilities in 2003. The preliminary trial
concentrated on placing tags on clothing items, speciﬁcally men’s suits, shirts, and ties.
By 2004, M&S expanded the operation to nine stores, but decided to concentrate only on
tagging men’s suits. After three years of testing, as of spring 2006, M&S has decided to
extend the RFID trial to 53 stores and encompass additional articles of clothing. M&S
has determined that by using RFID they are more aware of their inventory and have
reduced the time it takes to record inventory by 7 hours per week for a single store.
Additionally, constant inventory updates ensure that a full range of sizes is available for
any product. In addition to ﬁnding the right size, customers are provided with an
informational label advising consumers that the RF lD tag on the clothing is being used by
M&S for stock—control purposes. In addition to informative labels, as well as pamphlets
posted around the store, M&S offers to remove the RF ID tag at the checkout counter.
These methods appease most issues raised by consumer privacy groups in regards to the
RFID tagging of products. M&S surveys have concluded that most consumers do not
even recognize the RFID tag on the items, but recognize improved product
availability.[34]

The third largest retailer world-wide, Metro Group, began utilizing RF ID in its

supply chain in November 2004. Metro Group wanted to focus on tracking incoming and

22

outgoing shipments, and automatic reconciliation of shipments with shipping documents
across three retail sales divisions. The three divisions bring in a variety of products.
Metro’s Cash and Carry (groceries/general merchandise), Kauﬂrof (department store),
and Real (hyperrnarket) began receiving tagged shipments from 20 suppliers in total.
These shipments included groceries, general merchandise, textiles, and apparel.[35] In
addition to using RFID in supply chain trials, Metro recently took up more than 30,000
square feet at a German electronics fair, allowing attendees to see the real life
applications of RFID in their everyday life. Metro Group simulated a future store, which
contained RF ID technology on shopping carts, scales, clothing racks, and check-out
stations. Other areas demonstrated RF ID technology’s ability to help out the consumer at
home. RFID equipped washing machines, microwaves, and refrigerators were all a part
of the future home demonstration, all designed to make everyday chores less time

consumingm’]

RFID in Asia:

Countries in Asia have also begun allocating UHF ranges for RFID. Through the
Ministry of Public Management, Home Affairs, Posts, and Telecommunications, Japan
has allocated 6 MHz of bandwidth from 950 MHz to 956 MHz for UHF reader operation.
One particular Japanese electronics ﬁrm, NEC Tokin, plans to sell EPC Gen 2 readers for
use in the Japanese marketml The company worked in part with Impinj, a Seattle-based
semiconductor manufacturer. Companies in Japan have been eager to show off the future
applications that RFID technology can offer. In January 2006, Mitsukoshi Ginza

department store (owned by Fujitsu) performed a pilot in which 5000 pairs of jeans were

23

tagged for inventory management and improvement of store operations. The jeans were
then placed on smart shelves, which allowed employees to monitor what is available for
the customer, and also what sizes they had in stock rooms. Additionally, the pilot
included six smart ﬁtting rooms, which provided the customer with information about the
clothing they were trying on, what sizes were available, as well as outﬁt ideas and
accessories that the customer might be interested in based on what they had brought into
the ﬁtting room.[38]

South Korea is another Asian country that is supporting the growth of RFID and
its use in industry. Currently South Korea has set aside between 910 MHz and 914 MHz
for UHF reader operation. In 2004, GS] Korea organized an RFID pilot project with
Samsung Tesco, a large Korean retailer. The objective of the pilot was to determine
technical reliability using EPC standards. Beneﬁts identiﬁed in the pilot included
accelerated receiving and shipping with less human intervention.[391 Other RFID progress
in South Korea is currently being provided by Sun Microsystems. The company is
developing a RFID test center in Busan, South Korea, in collaboration with Busan
National University (BNU). BNU is known for the focus on manufacturing in their
curriculum, and Sun Microsystems hopes that the university will attract local
manufacturers to the RFID test center.[40]

Singapore has recently enacted legislation that increased the spectrum for UHF
RF ID systems from 920 MHz through 925 MHz. Infocomm Development Authority
(IDA) believes the added bandwidth will improve the performance of RF ID technology

in Singapore by reducing read-errors because systems will be able to select from more

channels to achieve less interference. IDA has reported that over 25 companies use RFID

24

in their supply chain, all of which have combined to invest over 16.5 million dollars in
RFID projectsl‘m.

The Chinese government has shown great interest in RF ID technology, but is
undecided on whether to cooperate with non- Chinese standards organizations. The
Standardization Administration of China currently has plans to make its own RFID
standard to protect its information security and enterprise interests, but will consider
compatibility between its own and foreign standards. The Chinese government has not
fully authorized frequencies for RFID use in Chinaml While standards talks continue,
RFID research is being carried out throughout China. Dan Dingyi, deputy director of
China Logistics Network Alliance, says that RFID had been widely adopted in a large
variety of ﬁelds including anti-counterfeiting systems, traffic monitoring, logistics, and
manufacturing. [43 ]

Retailers are also becoming involved with RFID in China. Bailian Group, one of
China’s largest retailers, has developed plans for the second phase of the China
Implementation Reference Project. The program seeks to expand usage of EPCs and
RFID. The second phase will track the movement of actual products tagged with EPC-
enabled RFID tags. Bailian is talking to suppliers based on their level of interest in
integrating RFID technology into the supply chain. Participating suppliers will tag
shipments in their originating distribution centers, and send those shipments to a Bailian
Group distribution center in Shanghai. The movement of these products will be recorded,

enabling all participants to gain a more accurate view of inventory levels.[441

25

RFID Read Locations:

It is evident that retailers, suppliers, and organizations are working with RFID
systems across the globe. However, at what point in the supply chain implementers
determine to set up an RFID system to gather data, known as read point or location,
varies. In retail applications, an obvious point to set up an RFID system and collect data
is by a dock door. The purpose of a dock door is to provide a medium where all products
arrive and depart, making it an excellent location for tracking inventory. Generally, a
portal contains, but is not limited to, four antennae positioned in various locations around
the door, in order to accurately detect tags entering and exiting the facility.

A shrink wrap station is another typical data collection point. This is an excellent
data capture point because most stretch wrappers offer 360 degrees of visibility of the
product (more importantly, of the tags), and, the stretch wrapping process takes more
time than walking through a dock door, which increases the chance that all tags are
detected. Multiple stretch wrap machines exist, and how they operate determines where
the RFID system will be set up. In some instances, a portal similar to that used in the
dock door situation can be placed around the stretch wrapper. Other stretch wrap
machines are equipped with an arm that rotates around the pallet, in which case an RF ID
antenna can be afﬁxed to the arm itself.

Fork trucks, used to transport product into, out of, and within warehouse facilities,
provide another medium for an RFID system placement. An antenna placed on the front
of the fork truck is capable of reading the tags located on the pallet load, ensuring that
drivers are carrying the correct product and placing it in the desired area, whether it is the

back of a destination-bound truck, or in storage.

26

Of all the examples discussed, dock doors, stretch wrappers, and fork truck read
locations focus on the detection of products on a unit load, like a pallet. Unit loads
generally contain a large number of case loads, which contain the individual product(s).
During the distribution process these case loads either need to be placed onto, or removed
from, the unit load. DCs generally utilize conveyors, deﬁned as “a mechanical apparatus
that transports materials, packages, or items to be assembled from one place to another”
to move the case loads to and from storageml Conveyors offer an excellent location for
a RFID read point by conﬁrming that the correct case has been pulled from storage, and
is bound for the correct re-palletization area.

Types of Conveyors

Conveyors generally operate by using either gravity or power to move an object
from point to point. A wide variety of conveyors exist, each performing a specialized
function. Some of the most common types of conveyor are those containing either a belt
or roller bars.

A belt conveyor is composed of either “fabric, rubber, plastic, leather, or metal
which operates over suitable drive, tail end and bend terminals”.[46] Belt conveyors are
versatile, provide a continuous ﬂow of product, and are low maintenance. They are
mainly used for carrying units, cartons, and bags. However, a modem-day example is the
use of the belt conveyor as a people mover in high trafﬁc areas such as airport terminals.

Roller conveyors tend to use gravity for product movement. On a roller conveyor
the load is supported over a series of rolling bars, turning on ﬁxed bearings that are
mounted between side rails at ﬁxed intervals. Product moving on a roller conveyor

requires three rollers under the load at all times. Product movement is controlled by

27

gravity; therefore, heavy loads on roller conveyors can be dangerous, they could
accelerate beyond control. Slides in parks for children are often built in roller bar
conveyor form, because the acceleration due to gravity can be a source of excitementl46].
A wide variety of other conveyor types exists, including bucket, chain, chute,
pneumatic, screw, vibrating, and wheel conveyors. Although they are mainly used for
material handling, conveyors also function throughout society as people movers. Ski

chair lifts on mountains are another functional examplem].

Conveyor Use and Research
It is almost inevitable that a package will travel on a conveyor at some point

during the manufacturing and distribution process. Since tracking product movement is
one of the key aspects of RFID, it is important to determine if RFID antennae are able to
track tagged packages on conveyors. Cases on a Wal-Mart conveyor can travel up to 600
feet per minute. Being able to read tagged cases on conveyors is important because it
ensures that Wal-Mart stores are receiving exactly what they ordered from their
distribution centers. Supply chain analysts have concluded that achieving 100 percent
readability on conveyors is not always possible, especially when the cases contain
product or packaging with high water or metal contentml Recent ﬁgures indicate that
Wal-Mart is currently reading 95% of tagged cases traveling on conveyor, which is 5%
less than their goal.

In a study performed by the RF ID Alliance laboratory at Kansas University, RFID
researchers brought tagged cases into an operational Kansas City distribution center and

studied case readability on a conveyor. Testing was conducted on ﬁve cased goods:

28

paper, bottled water, canned food, dishwasher detergent, and liquid hand soap. Tags
were placed on cases at a position that gave the best tag performance although how this
was determined was not discussed. The test was performed on a conveyor equipped with
a ﬂat metal sled covered by a belt, which operated at 204 feet per minute (fpm). For this
test, the researchers placed three different tags on three different cases of the same
product. They randomly selected the tags but placed them at the same optimal position.
Two types of readers were used for the test, the Matrics AR400 reader and the Alien
9780 reader. Each reader was connected to three antennae arranged so that one was on
each side of the conveyor 5’ apart and one was 42” above the conveyor pointing down.
The reason the antennae were set up in this manner was to closely simulate what is being
done by major retailers.

The results of the study were not released; however, the RFID Alliance Group did
make some general hypotheses about the relationship between conveyor line speed and
the readability of RFID tags. The group hypothesizes that, as line speed increases, tag
performance will decrease because the tag is in the reader ﬁeld for a shorter period of
time. Speciﬁcally, the group claims that, to achieve the same read rates at 600 fpm as at
200 fpm, the reader’s power settings would need to be increased by as much as 4 to 5

dB.[48]

Previous Testing and Research Gap
Over the past ﬁve years a number of theses focusing on various aspects of RFID
have been completed. Some have focused on the RF ID tag itself: its performance

capabilities design and implementation, failure modes, and enhanced anti-collision

29

protocol.[49'53] Research has also been performed on the effect of RFID in supply chain
management, in a warehouse environment, and for mobile object tracking. [54'56]

As knowledge of tag functionality increased, it became important to understand
the interaction between the tag and the product it was placed on. At Michigan State
University, John Onderko studied the performance of an RFID transponder placed on
reﬁigerated and frozen beef loin packages. The testing compared RFID systems
operating at different frequencies (13.56 MHz and 915 MHz) in combination with two
beef loin muscle conditions (frozen versus reﬁigerated). The author concluded that in the
915 MHz frequency range RFID systems communicated through frozen packages, but
experienced loss of data when communicating through a reﬁigerated package. He also
concluded that the 13.56 MHz frequency range experienced no loss of data when
transmitting through a package of refrigerated or frozen beef. This research determined
that the product and its state both have an effect on the readability of RFID tags.[57]

Another Michigan State Graduate, Jeffrey Tazelaar, expanded on Onderko’s work
by studying the effect of tag orientation and package content on the readability of RFID
tags. Tazelaar “utilized a pallet load of 48 corrugated cases, each tagged with a UHF
Class 0 RF ID tag”.[22] He compared ﬁve different tag locations and 5 different products,
walking the pallet load through a simulated warehouse RF ID portal. His research
determined that both tag orientation and product type have a signiﬁcant effect on tag
readability.[58]

As knowledge on the effect of product on readability of RFID tags grew, another

area of research that had not been explored was the effect of the number of antennae and

their conﬁguration had on the readability of cased goods. Thomas Crawforth of

30

Michigan State University determined that “there was a signiﬁcant difference in the read
rate of reads for both number and location of antennae” when encompassing case
contents (empty cases, rice and water), and tag type (Alien Class 1 and Symbol Class 0)
as variables in his researchml

Among the research being done across industry and at universities, there appears
to be a noticeable gap in published literature. This gap involves the relationship between
conveyor speed and the effect of both the package and the product on the readability of
RFID transponders. Conveyor speed and package readability have been tested in one
study only (RF ID Alliance lab at Kansas University discussed previously), but the
producers of the work determined “there were too many unconstrained variables to make
the results transferable to other environments”[48].

This research focuses on two conveyor speeds commonly used in industry; 600
feet per minute and 300 feet per minute. Determining merely the inﬂuence of speed on
the readability of RF ID tags is a test in itself; however, for suppliers and retailers of
goods, more valuable information can be gathered by placing the tags on actual products.
By using multiple products, testing of three additional variables is possible: product
effect (ketchup and motor oil), package effect (potato chips in plastic tub and potato chips
in metalized tub) and package shape (metal bottles, metal cans and metal tins).
Additionally, the type of tag used was the ﬁfth and ﬁnal variable in the research
performed, comparing a Gen 1 tag to a Gen 2 tag. These two tag types were chosen

because at the time this research was performed Gen 2 tags were available, however,

most mandates pertaining to RFID in the supply chain only required a Gen 1 tag.

31

CHAPTER 3 — METHODOLOGY

The following section discusses the materials and equipment necessary to perform
the aforementioned research testing. In total, seven different consumer goods products
were used to evaluate the effect of the ﬁve variables tested. To test the product effect a
case of ketchup was comparted to a case of motor oil. To test the package effect, a case
of potato chips in a metalized spiral wound ﬁberboard container was compared to a case
of potato chips in plastic tubs. To test the effect of package shape, three products were
used; a case of metal cans, a case of metal tins, and a tray of metal bottles.

Throughout the testing of the seven products, both the tag type and equipment
used were held constant throughout the entire testing procedure. Two procedures were
performed on the cases. Prior to the converyor testing, it was necessary to determine the
optimal tag location on each package. After this was complete the consumer goods
products were tested on the conveyor set-up to determine the tagged product’s readability
while moving down the conveyor at speed. Upon completion of testing, a statistical
analysis of variables was performed to determine the overall effect of the ﬁve variables
on case readability.

Materials

Product Effect

The case of ketchup was made from B ﬂute corrugated board, with the outside
dimensions of the case measuring 16” x 9.5” x 7.5”.The box manufacturer tested the case
determining a 32 pounds per inch (lbs/in) edge crush test, a size limit of 75” and a gross

weight limit of 65 pounds (lbs). The case held 16 bottles, each containing 20 ounces (oz)

32

of ketchup packaged in a polyethylene terephthalate (PETE bottle) with a polypropylene

(PP) lid (Figure 2).

IO I.‘ L I (1
KETCHUP

 

C

Figure 2: (A) Case of Ketchup, (B) Lay-out of Bottles in Case, (C) Individual Bottle

The case of motor oil was made from B Flute corrugated board, with the outside
dimensions of the case measuring 13.25” x 9.75” x 9.25”. The box manufacturer tested
the box determining a 44 lbs/in edge crush test, a size limit of 95” and a gross weight
limit of 95 lbs. The case contained 12 one- quart bottles of motor oil packaged in high

density polyethylene (HDPE) with a PP lid (Figure 3).

33

Mlllﬂll llll

 

 

C
Figure 3: (A) Case of Motor Oil, (B) Lay-out of Bottles in Case, (C) Individual

Bottle

34

 

Package Effect

The case containing the potato chips in plastic tubs consisted of B ﬂute 1/8” thick
corrugated board with the outside dimensions of the case measuring 14.75” x 9” x 6.75”.
The box manufacturer tested the box determining a 26 lbs/in edge crush test, with a size
limit of 50” and a gross weight limit of 35 lbs. The case contained 36 tubs, each holding

0.81 oz. of potato chips. (29.16 total oz) (Figure 4).

 

C
Figure 4: (A) Case of' Potato Chips, (B) Lay-out of Tubs in Case, (C) Individual Tub

35

The case of potato chips in metalized spiral wound ﬁberboard containers
(MSWFC) was made of E ﬂute 1/ 16” thick corrugated board with the outside dimensions
of the case measuring 16” x 9” x 9.5”. The box manufacturer tested the box determining
a 23 lbs/in edge crush test, with a size limit of 40” and a gross weight limit of 20 lbs. The
case contained 14 MSWFC each holding 6 oz. of potato chips per tub (84 oz total)

(Figure 5).

 

V.

V“.
GIN“ ,
W 4’
" 09:13; {a

C

 

Figure 5: (A) Case of Potato Chips, (B) Lay-out of MSWFC in Case, (C) Individual

MSWFC

36

Package Shape Effect

The aluminum bottles were packaged in an E ﬂute tray with the outside
dimensions of the tray measuring 10.5” x 8” x 3”, with the bottles extending 8” above the
tray, which was shrink wrapped with a polyoleﬁn ﬁlm. The tray held 12 bottles

containing 18 ﬂuid 02. of a liquid energy drink (Figure 6).

 

Figure 6: (A) Tray of Bottles, (B) Lay-out of Bottles in Tray, (C) Individual Bottle

37

 

The aluminum cans were packaged in a paperboard case (1.6 points) with the
outside dimensions of the case measuring 10.25” x 7.75” x 5”. The case held 12 cans

each containing 12 ﬂuid oz. of a lemon lime soda (Figure 7).

 

Figure 7: (A) Case of Cans, (B) Lay-out of Cans in Case, (C) Individual Can

The aluminum tins were packaged in a B ﬂute corrugated board case with the
outside dimensions of the case measuring 13” x 10” x 10”. The case contained 12 boxes
of facial tissue wrapped in aluminum foil (.0007” thick) to simulate a traditional candy

tin (Figure 8).

38

 

 

Figure 8: (A) Case of Tins, (B) Lay-out of Tins in Case, (C) Individual Tin

Equipment and Software:

The reader was an Alien ALR-9780 Reader (Alien Technologies, Morgan Hill,
Ca) (Figure 9). It is a Gen 2 reader, able to read and send data from Electronic Product
Code (EPC) Gen 1 and Gen 2 Class 1 tags to a computer for analysis using a RS-232
computer interconnection cable. The ALR-978O is a 4-port reader, capable of connecting
to four ultra high frequency (UHF) antennae. During the data collection process, the
software used was Alien Gateway V2.15.08. To determine the optimal tag location for

each product, RFID Tag Locator software V01 .0004 from Cape Systems (South

39

Plainﬁeld, NJ) was used. The results of the optimal tag location testing is discussed in

Chapter 4.

 

Figure 9: ALR-9780 Reader
Four Alien ALR-96lO-AC circularly polarized antennae were used because they
are less sensitive to tag orientation, and the read distance required was not large enough

to require linear antennae (Figure 10).

 

Figure 10: ALR-9610-AC Antenna

The tags were an EPC Class 1 Gen 1 ALL-9340-02 "SquiggleTM 2" and an EPC
Class 1 Gen 2 ALL-9440 "Gen2 Squigglem". Each tag measured 4” x 1/2”. These two

tags types were chosen because they offered a comparison between Gen 1 and Gen 2 tag

40

capabilities. Speciﬁcations for both tags claim the ability to work well on most
packaging products (corrugated board, plastic, and paper), while the Gen 2 tag is claimed
to perform well when used with package systems involving metal and/or water (Figure

11).

(«mean
‘ lama

 

Figure 11: (A) Back of Gen 1 Tag (top) and Gen 2 Tag (bottom). (B) Front of Gen 1
Tag (top) Gen 2 Tag (bottom)

The conveyor (Buschman Conveyors (Cincinnati, OH) was 10’ long and 31.5”
wide. Conveyor speed was controlled by a converted Weslo treadmill (Colorado Springs,
CO). A skate-wheel conveyor was positioned at the end of the Buschman conveyor to
slow and stop the product. The speed of the conveyor was monitored by a Computak
tachometer Model 8203 by Cole-Parmer Instrument Company (Vernon Hills, IL).

The antennae were placed on the left (1), right (3), top (2) (facing down towards
the conveyor), and bottom (4) (facing up towards the conveyor) sides of the conveyor
(Figure 12). The horizontal distance from the center of the conveyor belt to the center of
the side antennae was 20.75”. The vertical distance from the top of the conveyor to the

center of the side antennae was 8”. Both side antennae were angled 30 degrees down

41

towards the conveyor. The vertical distance from the top of the conveyor belt to the top

antenna was 30”, and to the bottom antennae was 12”.

1‘1} «2‘

___.__m.

 

Figure 12: Antennae Locations Surrounding Conveyor

42

Procedure for Determining Optimal Tag Location

Testing was conducted using one Alien ALR-9610-AC circularly polarized
antenna mounted on a wooden stand 36” from the center of the antenna to the ﬂoor. Each
of the products tested was placed on top of a 30.5” stand, composed of 3 empty
corrugated boxes stacked on top of one another, and located at 90 degrees and 30” away
from the antenna (Figure 13). This corrugated stand provided adequate line of sight
between the antenna and the tagged product, in addition to being a medium in which RF
waves are neither absorbed nor reﬂected, thus ensuring the stand had no effect on the
outcome of the test. With each product tested, the face of the case and the front of the
antenna were kept 30” apart, following the instructions for case testing provided in the

Cape Systems user manual (version V01.00.04).

43

 

Figure 13: Case and Anna Set-up for Cape Systems RFID Tag Locator Software
For each product, two sides of the case were selected to determine an optimal tag
location: a front face of the case (representing the width of the case) and a side face of
the case (representing the length of the case). Each face to be tested was equipped with a
1” x 1” grid drawn on a piece of paper that was taped to the face of the case to be tested.
The center of the tag was placed at the intersection of each horizontal and vertical line.

The tag was moved from intersection to intersection for each new trial run (Figure 14).

44

 

Figure 14: Example Grid Placed on Front Face of Case

Once the case and antenna were set up, the dimensions of the case were entered in
the software’s Case Setup page. The Hotspot test option, which brings up a 3-
dirnensional version of the product, was selected. The software creates a 1” x 1” grid on
each face of the case (Figure 15). The face representing the width of the case and the
closest size tag (1” x 4”) were selected from the on screen options. On the 3-dimensional
on-scrcen image, an intersection was selected that allowed for the tag to ﬁt completely on
the case without overhang, and the actual tag was placed in the same location on the

product to be tested.

45

 

 

 

Figure 15: Example Tag Location on Package Face

The tag was placed on the package vertically, the antenna was activated, and
results were recorded at each grid intersection. When each intersection had been tested,
the tag was moved to the length side of the case, and the test was repeated. After
completion of both sides of the case, the tag was repositioned horizontally on the case,
and both the width and length side of the case was tested again. Upon determining the
optimal tag location, a pin was used to penetrate through the grid and mark the box. The
grid was then removed from the package and the pin hole represented the place for which

the center of the tag was placed during conveyor testing.

Procedure for Testing RF ID Tagged Case Loads on Conveyor

The seven types of product were tested in an off-campus warehouse (Figure 16).
Besides the product, package and shape variations in each case, two additional variables
were tested: speed (600 and 300 feet per minute (fpm)) and tag type (a Generation 1 tag
and a Generation 2 tag). Each test, which consisted of 30 trials, began with the activation
of the RFID equipment through the Alien software. Next, the tagged case located outside
of the antennae read ﬁeld was placed on the moving conveyor belt operating at speed.

The orientation of the case on the belt was such that the tag on the case was in the direct

46

 

 

line of sight with one of the side antennae. The product traveled down the conveyor,
passed through the antenna portal read ﬁeld, and eventually moved out of the read ﬁeld
when it was swept onto the skate wheel conveyor and brought to a halt. Each product
underwent 120 trials using two conveyor speeds and two tag types. Two types of results
were recorded; whether the tag was detected or not, and the number of times the tag was

detected. These results were stored in a Microsoft Excel ﬁle.

47

Antennae
Portal

. :ﬁ'
‘ 3
'3

_ Tl
1
TI
..-

 

Figure 16: Warehouse with Conveyor and RFID Set-up

48

Statistical Analysis

With the assistance of the Center for Statistical Training and Consulting (CSTAT)
a 3 way Analysis of Variance (ANOVA) model was used to test the normality of
residuals. SAS Glimmix was the version of software used. For a valid comparison to be
made, it was necessary to use the data showing the number of times the product was
detected. It would not be statistically valid, nor very interesting, if comparisons were
made by assessing whether the product was detected or not, being that six out of seven
products were detected in every trial of every test.

The research performed in the 3 tests had 3 main effectors. One of the effectors
was conveyor speed, consisting of two levels (300fpm and 600fpm). The second effecter
was tag type, consisting of two levels (Genl and Gen2). The third effecter was the
package or product itself. For the product effect test, the effecter was the product itself,
consisting of two levels (ketchup and motor oil). For the package effect test, the effecter
was the package, consisting of two levels (plastic tubs and MSWFC). For the effect of
package shape, the effecter was shape, consisting of three levels (metal cans, metal

bottles and metal tins).

49

 

 

 

 

CHAPTER 4 — RESULTS

Hotspot Test Results

The Hotspot test determines whether a tag is in a good or bad location by
measuring the level of attenuation at which a tag responds. Signal attenuation is deﬁned
as “the weakening of RF energy from an RF ID tag or a reader”.m The higher the
attenuation value recorded when a tag responds, the lower the amount of power being
delivered to the tag ﬁom the antenna. The range of power used by the software to detect
the passive tag ranged from 15 dB to 0 dB, with 15 dB representing an optimal situation
in which the antenna had to put out very little power to receive a signal back from the tag
and 0 dB representing the tag not sending a response signal. As the tag is moved ﬁ'om
intersection to intersection within the grid, attenuation values are recorded for each tag
location. The value then correlates to a certain color which appears on the screen at the
intersection of interest (Figure 17). A tag attenuation response between 0-8 dB provides
a color response ranging from bright red to bright white. A response between 8-15 dB
provides a color response ranging from bright white moving to bright green. In general, a
red response is a poor location to place a tag, a white response is an okay position to
place a tag, and a green response is an excellent location to place a tag. The variance in
color is due to a variety of interference possibilities due to packaging materials or product

content.

50

 

Figure 17: Example RF ID hotspot test outcome for one side of a case

Results for each of the seven products were recorded, testing two tag orientations
on two adjoining faces of each case, a width and a length. From the data obtained, an
optimal tag location and orientation was determined for each product. The results for the
Hotspot test revealed similarities regarding each of the seven products. Although testing
was performed with two tag orientations, the vertical tagging proved to be optimal for
each product. Additionally, the width face, either end 5 or end 6 according to ASTM D
775, Standard Test Method for Drop Test for Loaded Boxes, of the case proved to be an
equal or better tag location than length face of the case. To keep the testing consistent,

the tag was placed on the width face of the case for each of the seven products.

Product Effect Results from Hotspot Testing

For the case of ketchup the center of the tag was afﬁxed on end 6 of the case, 8”
from uppermost left comer of the width face, and down 3” from the top of the width face

(Figure 18).

51

 

Figure 18: Tag Location on Ketchup Case
This location was determined to be optimal because it produced an attenuation level of 12

dB, which the software depicted as a bright green in that location (Figure 19).

 

 

 

 

 

 

 

Figure 19: Hotspot Test Output for Case of Ketchup, (A) Length and Width View of
Case, (B) Length View of Case, (C) Width View of Case

52

For the case of motor oil the center of the tag was affixed on end 6 of the case, 1”
from uppermost left comer of the width face, and down 2” from the top of the width face

(Figure 20).

 

Figure 20: Tag Location on Motor Oil Case

This location was determined to be optimal because it produced an attenuation level of 10

dB, which the software depicted as a very light green in that location (Figure 21).

53

 

C

Figure 21: Hotspot Test Output for Case of Motor Oil, (A) Length and Width View
of Case, (B) Length View of Case, (C) Width View of Case

Package Effect Results From Hotspot Testing
For the case of potato chips in plastic tubs the center of the tag was affixed on end
5 of the case, 4” from uppermost left corner of the width face, and down 2” from the top

of the width face (Figure 22).

 

Figure 22: Tag Location on Case of Chips in Plastic Tubs

54

This location was determined to be optimal because it produced an attenuation level of 10

dB, which the software depicted as a very light green in that location (Figure 23).

 

C

Figure 23: Hotspot Test Output for Case of Potato Chips in Plastic Tubs, (A) Length
and Width View of Case, (B) Length View of Case, (C) Width View of Case

For the case of potato chips in metalized spiral-wound ﬁberboard containers
(MSWFC) the center of the tag was afﬁxed on end 6 of the case, 4” fi'om uppermost left

comer of the width face, and down 3” from the top of the width face (Figure 24).

55

 

Figure 24: Tag Location on Case of Chips in MSWFC
This location was determined to be optimal because it produced an attenuation level of 13

dB, which the software depicted as a bright green in that location (Figure 25).

--Iw-

.‘uﬁ .. _
l-II-It

l-Il-I"
..._,

 

C

Figure 25: Hotspot Test Output for Case of Potato Chips in MSWFC, (A) Length
and Width View of Case, (B) Length View of Case, (C) Width View of Case

56

 

Package Shape Effect Results From Hotspot Testing
For the case of aluminum cans the center of the tag was affixed on end 6 of the
case, 5” ﬁ'om uppermost left corner of the width face, and down 2” from the top of the

width face (Figure 26).

 

Figure 26: Tag Location on Case of Cans

This location was determined to be optimal because it produced an attenuation level of 4

dB, which the software depicted as a pinkish white in that location (Figure 27).

57

 

C

Figure 27: Hotspot Test Output for Case of Cans, (A) Length and Width View of
Case, (B) Length View of Case, (C) Width View of Case

For the tray of aluminum bottles the center of the tag was affixed on end 6 of the
tray, 3” from uppermost left corner of the width face, and down 6” from the top of the

width face (Figure 28).

58

 

u- rz-rsrrozaaaaz ;
Him . maruuuurimmmm

Figure 28: Tag Location for Tray of Bottles
This location was determined to be optimal because it produced an attenuation level of 8

dB, which the software depicted as a bright white in that location (Figure 29).

59

 

Figure 29: Hotspot Test Output for Tray of Bottles, (A) Length and Width View of
Case, (B) Length View of Case, (C) Width View of Case

For the case of aluminum tins the center of the tag was afﬁxed on end 6 of the
case, 9” from uppermost left corner of the width face, and down 2” from the top of the

width face (Figure 30).

60

 

Figure 30: Tag Location for Case of Tins
At no point during the Hotspot test was there ever a dB level greater than 0
recorded, therefore the tag location was determined by using a location on the case that
had been a traditional tag location in previous research testing, the upper right hand

comer of the case (Figure 31).

61

 

 

C

Figure 31: Hotspot Test Output for Case of Tins, (A) Length and Width View of

Case, (B) Length View of Case, (C) Width View of Case

Statistical Analysis of Conveyor Effect

To statistically analyze the results of the conveyor test, results (number of reads
per trial) were taken from the Microsoft excel ﬁle, and put into the SAS Glimmix
software. The results for each case tested were calculated and compared.

In the tables depicted below there are a variety of terms and corresponding values
displayed. For the purpose of this research, the values of most signiﬁcance are located

under the Estimate column, and either the Pr > F or Pr > |t| columns.

62

The values under the Estimate column refer to either the average number of tag
reads per trial or, in the case the table is titled “Differences of Product Least Square
Means”, Estimate is deﬁned as the difference in the average number of tag reads per trail
between the two compared products.

The values under either the Pr > F or Pr > |t| represent the statistical signiﬁcance
of the data. A P-value of 0.01 or lower indicates statistical signiﬁcance.

Within the tables product names are shortened to a more concise term, the list of
the product and the shortened term are shown below, in addition to deﬁnitions for terms

of table column headings.

Terminology

Chips in Plastic Tubs: Plastic

Chips in MSWFC: Metal

Den DF: Degrees of freedom of the denominator

DF (Degrees of Freedom): Any of the unrestricted, independent random variables
that constitute a statistic.

Effect: The variable or variables affecting the number of tag reads per trial
Estimate: 1.) The average number of tag reads per trial, or 2.) In the case the
table is titled “Differences of Product Least Square Means”, Estimate is deﬁned
as the difference in the average number of tag reads per trail between the two
compared products.

Fixed Effects:

0 Product (Ketchup, Motor oil)

0 Speed (300 feet per minute (fpm), 600 fpm)

0 Tag (Gen 1, Gen 2)

0 Product x Speed (Ketchup (300 fpm, 600 fpm)) x (Motor Oil (300 fpm, 600
fpm))

0 Product x Tag (Ketchup (Gen 1, Gen 2)) x (Motor Oil (Gen 1, Gen 2))

63

0 Tag x Speed (300 fpm (Gen 1, Gen 2)) x (600 fpm (Gen 1, Gen 2))

0 Product x Speed x Tag (Ketchup (Gen 1, Gen 2) (300 fpm, 600 fpm)) x
(Motor Oil (Gen 1, Gen 2) (300 fpm, 600 fpm)

F-Value: Mean Square for effect divided by mean square for error

Ketchup: Ketchup

Least Square Means: average, this is the best method for calculating the average
statistically

MD: Mean Difference between the average number of tag reads per trial between
the subjects being compared.

Metal Bottles: Bottles

Metal Cans: Cans

Metal Tins: Tins

Motor Oil: Oil

Num DF: Degrees of freedom of the numerator

Pr > F: Calculation of the P-value (probability to reject null hypothesis) for F -
test.

Product: One of the consumer goods being tested

Product: The consumer good being compared to the ﬁrst product

Pr > |t|: Calculation of the P-value for the t-test.

Standard Error: The estimated standard deviation of the sample mean
t-Value: The ratio of the difference between the sample mean and the given

number to the standard error of the mean

Product Effect Results:

Ketchup

In Trial 1, the case traveled on the conveyor at 300 feet per minute (fpm)

equipped with a Gen 1 tag. Over the 30 trials, there were 2,875 instances in which the

tag was detected. On average there were 95.83 tag reads per trial with a standard

deviation of 12.9.

64

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 1,206 instances in which the tag was detected. On
average there were 40.2 tag reads per trial with a standard deviation of 12.1.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with 3 Gen 2
tag. Over the 30 trials, there were 3,831 instances in which the tag was detected. On
average there were 127.7 tag reads per trial with a standard deviation of 13.6.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1,023 instances in which the tag was detected. On

average there were 34.1 tag reads per trial with a standard deviation of 7.4.

Motor Oil

In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 5,974 instances in which the tag was detected. On
average there were 199.1 tag reads per trial with a standard deviation of 40.2.

In Trial 2 the case traveled on the conveyor at 600 fpm equipped with a Gen 1 tag.
Over the 30 trials, there were 2,243 instances in which the tag was detected. On average
there were 74.4 tag reads per trial with a standard deviation of 18.5.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 6,685 instances in which the tag was detected. On
average there were 222.8 tag reads per trial with a standard deviation of 14.5.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1,803 instances in which the tag was detected. On

average there were 60.1 tag reads per trial with a standard deviation of 8.1.

65

Statistical Analysis of the Product Effect Test
Table 1 depicts the results from testing the ﬁxed effects involved in testing the
product effect on tag readability.

Table 1: Tests of Fixed Effects for Product Effect

 

 

 

 

 

 

 

 

 

 

Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
Product 1 232 704.85 <.0001
Speed 1 232 2008.70 <.0001
Tag 1 232 13.00 0.0004
Product*Speed l 232 201.20 <.0001
Product*Tag l 232 2.83 0.0937
Speed*Tag l 232 60.85 <.0001
Product*Speed*Tag 1 232 0.00 0.9945

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 1.

Row 1. Statistical analysis output concluded that product (ketchup or motor oil) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 2. Statistical analysis output concluded that speed (300 fpm or 600 fpm) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 3. Statistical analysis output concluded that tag type (Gen 1 or Gen 2) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 4. Statistical analysis output concluded that product by speed had a signiﬁcant

interaction effect on the average number of tag reads per trial.

66

Row 5. Statistical analysis output concluded that product by tag did not have a
signiﬁcant interaction effect on the average number of tag reads per trial.
Row 6. Statistical analysis output concluded that speed by tag had a signiﬁcant
interaction effect on the average number of tag reads per trial.
Row 7. Statistical analysis output concluded that product by speed by tag did not have a
signiﬁcant interaction effect on the average number of tag reads per trial.

Table 2 shows the statistical analysis output concluded that product type had a
signiﬁcant effect on the average amount of tag reads per trial, with the case of motor oil
averaging 64.7 more tag reads per trial than the case of ketchup

Table 2: Average Number of Tag Reads Per Trial and Difference Between Means,
by Product Type

 

Product Least Squares Means

 

 

 

 

Product Estimate Standard DF t Value Pr > |t|
Error
Ketchup 74.4583 1.7223 232 43.23 <.0001
Oil 139.13 1.7223 232 80.78 <.0001
MD -64.6667 2.4358 232 -26.55 <.0001

 

 

 

 

 

 

 

 

Table 3 shows that statistical analysis output concluded that conveyor speed had a
signiﬁcant effect on the average number of tag reads per trial, with 300 fpm averaging

109.2 more tag reads per trial than 600 fpm.

67

Table 3: Average Number of Tag Reads Per Trial and Difference Between Means,

by Conveyor Speed

 

 

 

 

 

Speed Least Squares Means
Speed Estimate Standard Error DF t Value Pr > |t|
300 161.38 1.7223 232 93.70 <.0001
600 52.2083 1.7223 232 30.31 <.0001
MD 109.17 2.4358 232 44.82 <.0001

 

 

 

 

 

 

 

 

Table 4 shows that statistical analysis output concluded that tag type had a
signiﬁcant effect on the average number of tag reads per trial, with Gen 2 averaging 8.8
more tag reads per trial than Gen 1.

Table 4: Average number of tag reads per trial and difference between means by tag

 

 

 

 

 

generation
Tag Least Squares Means
Tag Estimate Standard Error DF t Value Pr > |t|
Genl 102.40 1.7223 232 59.45 <.0001
Gen2 111.18 1.7223 232 64.55 <.0001
MD -8.7833 2.4358 232 -3.61 0.0004

 

 

 

 

 

 

 

 

Table 5 shows the statistical analysis output of the interaction between the product

and a conveyor speed.

68

Table 5: Average Number of Tag Reads Per Trial, Product by Speed

 

Product*Speed Least Squares Means

 

Product Speed Estimate Standard DF tValue Pr>|t|
Error

 

Ketchup 300 111.77 2.4358 232 45.89 <.0001

 

Ketchup 600 37.1500 2.4358 232 15.25 <.0001

 

Oil 300 210.98 2.4358 232 86.62 <.0001

 

Oil 600 67.2667 2.4358 232 27.62 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 5.
0 Row 1. The case of ketchup moving at 300 fpm averaged 111.8 reads per trial
0 Row 2. The case of ketchup moving at 600 fpm averaged 37.2 reads per trial
0 Row 3. The case of motor oil moving at 300 fpm averaged 211 reads per trial

0 Row 4. The case of motor oil moving at 600 fpm averaged 67.3 reads per trial

Table 6 shows the statistical analysis output of the interaction between the two

way interaction between the two products and the two speeds.

69

Table 6: Difference in Average Number of Tag Reads Per Trial Between Product by

 

 

 

 

 

 

 

 

 

Speed
Differences of Product*Speed Least Squares Means
Product Speed _Product _Speed Estimate Standard DF t Value Pr > |t|
Error
Ketchup 300 Ketchup 600 74.6167 3.4447 232 21.66 <.0001
Oil 300 Oil 600 143.72 3.4447 232 41.72 <.0001
Ketchup 300 Oil 300 -99.2167 3.4447 232 -28.80 <.0001
Ketchup 300 Oil 600 44.5000 3.4447 232 12.92 <.0001
Ketchup 600 Oil 300 - l 73.83 3 .4447 232 -50.46 <.0001
Ketchup 600 Oil 600 -30.1 167 3.4447 232 -8.74 <.0001

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 6.

0 Row 1. Comparing the case of ketchup moving at 300 fpm versus the case of
ketchup moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of ketchup at 300 fpm averaging 74.6 more tag
reads per trial.

0 Row 2. Comparing the case of motor oil moving at 300 fpm versus the case of
motor oil moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with. the case of motor oil at 300 fpm averaging 143.7 more
tag reads per trial.

0 Row 3. Comparing the case of ketchup moving at 300 fpm versus the case of
Motor oil moving at 300 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of motor oil at 300 fpm averaging 99.2 more

tag reads per trial.

70

0 Row 4. Comparing the case of ketchup moving at 300 fpm versus the case of
Motor oil moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of ketchup at 300 fpm averaging 44.5 more tag
reads per trial.

0 Row 5. Comparing the case of ketchup moving at 600 fpm versus the case of
Motor oil moving at 300 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of motor oil at 300 fpm averaging 173.8 more
tag reads per trial.

0 Row 6. Comparing the case of ketchup moving at 600 fpm versus the case of
motor oil moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of motor oil at 600 fpm averaging 30.1 more

tag reads per trial.

Table 7 shows the statistical analysis output of the interaction between the product

and the tag type.

71

Table 7: Average Number of Tag Reads Per Trial, Product by Tag Generation

 

I

Product*Tag Least Squares Means

 

Product Tag Estimate Standard DF tValue Pr>|t|
Error

 

Ketchup Genl 68.0167 2.4358 232 27.92 <.0001

 

Ketchup Gen2 80.9000 2.4358 232 33.21 <.0001

 

Oil Genl 136.78 2.4358 232 56.16 <.0001

 

Oil Gen2 141.47 2.4358 232 58.08 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 7.
0 Row 1. The case of ketchup with a Gen 1 tag averaged 68 reads per trial.
0 Row 2. The case of ketchup with a Gen 2 tag averaged 81 reads per trial.
0 Row 3. The case of motor oil with a Gen 1 tag averaged 136.8 reads per trial.

0 Row 4. The case of motor oil with a Gen 2 tag averaged 141.5 reads per trial.

Table 8 shows the statistical analysis output of the interaction between the two

way interaction between the two products and the two tag types.

72

Table 8: Difference in Average Number of Tag Reads Per Trial Between Product by

Tag Generation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error

Ketchup Genl Ketchup Gen2 -12.8833 3.4447 232 -3.74 0.0002

Oil Genl Oil Gen2 —4.6833 3.4447 232 -l .36 0.1753
Ketchu p Genl Oil Genl -68.7667 3.4447 232 419.96 <.0001
Ketchup Genl Oil Gen2 -73.4500 3.4447 232 -2 l .32 <.0001
Ketchup Gen2 Oil Genl -55.8833 3.4447 232 -l6.22 <.0001
Ketchup Gen2 Oil Gen2 -60.5667 3.4447 232 -l7.58 <.0001

L_—.-=== ===r = = j—

 

 

The following results refer speciﬁcally to each row of results in Table 8.

0 Row 1. Comparing the Gen 1 tag on ketchup versus the Gen 2 tag on ketchup,
there was a signiﬁcant difference between the average number of tag reads per
trial, with the Gen 2 tag on ketchup averaging 12.9 more tag reads per trial.

0 Row 2. Comparing the Gen 1 tag on motor oil versus the Gen 2 tag on motor oil,
there ELM a signiﬁcant difference between the average number of tag reads
per trial, with the Gen 2 tag on motor oil averaging 4.7 more tag reads per trial.

0 Row 3. Comparing the Gen 1 tag on ketchup versus the Gen 1 tag on motor oil,
there was a signiﬁcant statistical difference in the number of tag reads with the
Gen 1 tag on motor oil averaging 68.8 more tag reads per trial.

0 Row 4. Comparing the Gen 1 tag on ketchup versus the Gen 2 tag on motor oil,
there was a signiﬁcant statistical difference in the number of tag reads with the

Gen 2 tag on motor oil averaging 73.5 more tag reads per trial.

73

0 Row 5. Comparing the Gen 2 tag on ketchup versus the Gen 1 tag on motor oil,

there was a signiﬁcant statistical difference in the number of tag reads with the

Gen 1 tag on motor oil averaging 55.9 more tag reads per trial.

0 Row 6. Comparing the Gen 2 tag on ketchup versus the Gen 2 tag on motor oil,

there was a signiﬁcant statistical difference in the number of tag reads with the

Gen 2 tag on motor oil averaging 60.6 more tag reads per trial.

Table 9 shows the statistical analysis output of the interaction between tag type and

conveyor speed.

Table 9: Average Number of Tag Reads Per Trial, Tag Generation by Speed

 

 

 

 

 

 

 

 

 

 

 

 

 

a
Speed*Tag Least Squares Means
Tag Speed Estimate Standard DF t Value Pr > |t|
Error
Genl 300 147.48 2.4358 232 60.55 <.0001
Gen2 300 175.27 2.4358 232 71.96 <.0001
Genl 600 57.3167 2.4358 232 23.53 <.0001
Gen2 600 47.1000 2.4358 232 19.34 <.0001

 

 

 

The following results refer speciﬁcally to each row of results in Table 9.

0 Row 1. The case with a Gen 1 tag moving 300 fpm averaged 147.5 reads per trial.

0 Row 2. The case with a Gen 2 tag moving 300 fpm averaged 175.3 reads per trial.

0 Row 3. The case with a Gen 1 tag moving 600 fpm averaged 57.3 reads per trial.

0 Row 4. The case with a Gen 2 tag moving 600 fpm averaged 47.1 reads per trial.

74

Table 10 shows the statistical analysis output of the two way interaction between

the tag type and conveyor speed.

Table 10: Difference In Average Number of Tag Reads Per Trial Between Tag

Generation by Speed

 

 

 

 

 

 

 

 

 

- Differences of Speed*Tag Least Squares Means

Tag Speed _Tag _Speed Estimate Standard Error DF t Value Pr > |t|
Genl 300 Gen2 300 -27.7833 3.4447 232 -8.07 <.0001
Genl 300 Genl 600 90.1667 3.4447 232 26.18 <.0001
Genl 300 Gen2 600 100.38 3 .4447 232 29.14 <.0001
Gen2 300 Gen 1 600 1 17.95 3.4447 232 34.24 <.0001
Gen2 300 Gen2 600 128.17 3.4447 232 37.21 <.0001
Genl 600 Gen2 600 10.2167 3.4447 232 2.97 0.0033

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 10.

Row 1. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving
at 300 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 2 tag at 300 fpm averaging 27.8 more tag reads per trial.
Row 2. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 1 tag at 300 fpm averaging 90.2 more tag reads per trial.
Row 3. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a significant statistical difference in the number of tag

reads, with the Gen 1 tag at 300 fpm averaging 100.4 more tag reads per trial.

75

0 Row 4. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 2 tag at 300 fpm averaging 118 more tag reads per trial.

0 Row 5. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 2 tag at 300 fpm averaging 128.2 more tag reads per trial.

0 Row 6. Comparing the Gen 1 tag moving at 600 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag

reads, with the Gen 1 tag at 600 fpm averaging 10.2 more tag reads per trial.

Table 11 shows the statistical analysis output of the interaction between the product,

tag type, and conveyor speed.

76

Table 11: Average Number of Tag Reads Per Trial, Product by Tag Generation by

 

 

 

 

 

 

 

 

 

 

Speed
Product*Speed*Tag Least Squares Means
Product Tag Speed Estimate Standard Error DF t Value Pr > |t|
Ketchup Genl 300 95.8333 3.4447 232 27.82 <.0001
Ketchup Gen2 300 127.70 3.4447 232 37.07 <.0001
Ketchup Genl 600 40.2000 3.4447 232 1 1.67 <.0001
Ketchup Gen2 600 34.1000 3 .4447 232 9.90 <.0001
Oil Genl 300 199.13 3.4447 232 57.81 <.0001
Oil Gen2 300 222.83 3.4447 232 64.69 <.0001
Oil Genl 600 74.4333 3.4447 232 21.61 <.0001
Oil Gen2 600 60.1000 3.4447 232 17.45 <.0001

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 11.

0 Row 1. The case of ketchup with a Gen 1 tag moving 300 fpm averaged 95.8
reads per trial.

0 Row 2. The case of ketchup with a Gen 2 tag moving 300 fpm averaged 127.7
reads per trial.

0 Row 3. The case of ketchup with a Gen 1 tag moving 600 fpm averaged 40.2
reads per trial.

0 Row 4. The case of ketchup with a Gen 2 tag moving 600 fpm averaged 34.1
reads per trial.

0 Row 5. The case of motor oil with a Gen 1 tag moving 300 fpm averaged 199.1

reads per trial.

77

0 Row 6. The case of motor oil with a Gen 2 tag moving 300 fpm averaged 222.8
reads per trial.

0 Row 7. The case of motor oil with a Gen 1 tag moving 600 fpm averaged 74.4
reads per trial.

0 Row 8. The case of motor oil with a Gen 2 tag moving 600 fpm averaged 60.1

reads per trial.

Ketchup-Ketchup Interactions
Table 12 shows the statistical analysis output of the three-way interaction between

the case of ketchup, tag type, and conveyor speed.

78

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M: _Nc mm; Nmm m _ $6. aaa _ .e ace ~50 9533— cae :50 9583—
_cca.v _m.o_ mmm m _ S; acae. mo ace ~50 92.83— can N50 a283—
_ccc.v em;— Nmm m _ Se ccamSw cce :30 9533— aan «EU 9583—
Soov 5.9 mm m _ 2:. m3. 3 8... «50 3.3.3. 8». :50 9:33.
38v m3 2 mmm m _ 2:. 38.3. 8e :50 9.53. can .50 9533.
_aac.v ewe- mmm m _ Saw seem. _ m- can «:00 9:33.: acm :50 9383—
LOP—m—
E A .5 33> . ma Ewe—_Sm 235mm 63%| uahl 828.51 3on map. 835.:—
mnaez mega—Em .234 makreoamtgeoi ac 89:95.:5

 

 

6025 .3 5:20:00 web .3 3.58on «2595 season ETC. .3.— naaom nah. .«e hoe—=52 owaao>< E agony—ma "S 035—.

79

The following results refer speciﬁcally to each row of results in Table 12.

Row 1. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of ketchup with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 2 tag moving at 300 fpm averaging 31.8 more tag reads per
trial.

Row 2. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of ketchup with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 1 tag moving at 300 fpm averaging 55.6 more tag reads per
trial.

Row 3. Comparing the case of ketchup with 3 Gen 1 tag moving at 300 fpm
versus the case of ketchup with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 1 tag moving at 300 fpm averaging 61.7 more tag reads per
trial.

Row 4. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm
versus the case of ketchup with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 2 tag moving at 300 fpm averaging 87.5 more tag reads per
trial.

Row 5. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm

versus the case of ketchup with a Gen 2 tag moving at 600 fpm, there was a

80

signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 2 tag moving at 300 fpm averaging 93.6 more tag reads per
trial.

0 Row 6. Comparing the case of ketchup with a Gen 1 tag moving at 600 fpm
versus the case of ketchup with a Gen 2 tag moving at 600 fpm, there M a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 1 tag moving at 600 fpm averaging 6.1 more tag reads per

trial.

Ketchup Oil Interactions
Table 13 shows the statistical analysis output of the three-way interaction between

the two products, a tag, and a conveyor speed.

81

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_ccc.v cad- ~m~ £5... acach cce ~50 =0 ace :50 9.5.3—
_acc.v mas- ~m~ 256 33.3”- cce _50 :0 ace ~50 9.5.3—
.ccc.v 9.6m- ~m~ 356 3.3—. can ~50 =0 cce ~50 9.5.3—
_ccc.v ne.~m- ~n~ m_ 56 made..- can :50 :0 cce _50 9383—
_caa.v ee.v_- ~m~ 55... 9.3;». can ~50 :0 can ~50 9583—
_cac.v wad— ~m~ 25.... ccceHe ace ~50 :0 can ~50 9.55!
_acc.v moc— ~m~ 25... beeﬁnm cce —50 :0 can ~50 9.5.3—
~aac.v mde ~n~ m-nwé mam—.mo- acm ~50 :0 can ~50 9:550—
.cac.v um.» ~m~ 23.9 nmnnwm cce ~50 :0 can ~50 9.5.3—
.acc.v an... ~n~ m Fad cccv. _~ ace ~50 :0 can _ 50 9.55:
.acc.v bc.e~- ~m~ 25.». cc.\.~_- can ~50 :0 can _50 9.5.3—
_cca.v c~._~- ~m~ 256 cm.mc_- can .50 :0 can «50 9.5.3—
Sta
2 A E 33> . “E 2.65% 3555 .3an “El .262..- 825 w; 8355
2:52 523$ .33 ashraoeawaeu=u2m he 5o55t5

 

 

ask. .3 EC .582 can 933on 535.5 5955mm _ath 5m 35% wan. .«e 5:352 0353.. E 355.59 "Mm 935.

525 .3 559550

82

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

_aaa.v nae- ~m~ m—bwé aaaa.e~- ace ~50 =0 ace ~50 9.55!
—aac.v w~.a.. ~n~ m—nwé manna..- ace _50 :0 cce ~50 c.533—
_caa.v Exam- ~n~ m _ bad mama T can ~50 =0 ace ~50 9.55!
_aac.v maﬁa- ~m~ m—naé ma.me_- can _50 :0 cce ~50 9.55!
nun “ﬂ“ r1111... H H H

 

€3.83 2 uses

 

 

83

The following results refer speciﬁcally to each row of results in Table 13.

Row 1. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of oil

with a Gen 1 tag moving at 300 fpm averaging 103.3 more tag reads per trial.

Row 2. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 2 tag moving at 300 fpm averaging 127 more tag reads per trial.
Row 3. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 1 tag moving at 300 fpm averaging 21.4 more tag reads per
trial.

Row 4. Comparing the case of ketchup with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 1 tag moving at 300 fpm averaging 35.7 more tag reads per
trial.

Row 5. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor

oil with 3 Gen 2 tag moving at 300 fpm averaging 95.1 more tag reads per trial.

34

Row 6. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 2 tag moving at 300 fpm averaging 53.3 more tag reads per
trial.

Row 7. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with 3 Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
ketchup with a Gen 2 tag moving at 300 fpm averaging 67.6 more tag reads per
trial.

Row 8. Comparing the case of ketchup with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 300 fpm averaging 71.4 more tag reads per trial.
Row 9. Comparing the case of ketchup with a Gen 1 tag moving at 600 fpm
versus the case of motor oil with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 300 fpm averaging 158.9 more tag reads per trial.
Row 10. Comparing the case of ketchup with a Gen 1 tag moving at 600 fpm
versus the case of motor oil with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor

oil with a Gen 2 tag moving at 300 fpm averaging 182.6 more tag reads per trial.

85

Row 11. Comparing the case of ketchup with a Gen 1 tag moving at 600 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 600 fpm averaging 34.2 more tag reads per trial.
Row 12. Comparing the case of ketchup with a Gen 1 tag moving at 600 fpm
versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 2 tag moving at 600 fpm averaging 19.9 more tag reads per trial.
Row 13. Comparing the case of ketchup with a Gen 2 tag moving at 600 fpm
versus the case of motor oil with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with 3 Gen 1 tag moving at 300 fpm averaging 165 more tag reads per trial.
Row 14. Comparing the case of ketchup with a Gen 2 tag moving at 600 fpm
versus the case of motor oil with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 2 tag moving at 300 fpm averaging 188.7 more tag reads per trial.
Row 15. Comparing the case of ketchup with a Gen 2 tag moving at 600 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 600 fpm averaging 40.3 more tag reads per trial.
Row 16. Comparing the case of ketchup with 3 Gen 2 tag moving at 600 fpm

versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a

86

signiﬁcant statistical difference in the number of tag reads, with the case of motor

oil with a Gen 2 tag moving at 600 fpm averaging 26 more tag reads per trial.

Oil-Oil Interactions

Table 14 shows the statistical analysis output of the three-way interaction between

the case of motor oil, a tag, and a conveyor speed.

87

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

enaac vo.~ ~n~ m. 2; nmmmé. ace ~50 =0 cce _50 :0
—caa.v Slum ~m~ 2.5..» Q..~e_ ace ~50 :0 can ~50 =0
_caa.v 3cm ~m~ $5.». ave: cce _50 :0 can ~50 =0
_ccc.v vm.a~ ~n~ m _ 5... madm— cce ~50 :0 can _ 50 =0
_aca.v ce.m~ ~m~ 25... ab.v~_ cce _50 :0 can ~50 =0
_aaa.v 5.? ~m~ m _ 5... cccn.m~- cam ~50 :0 can :50 :0
Sta
E A .5 25> . ha Ewe—Sum San—3mm 55ml wail “Seek—I 525 m3. 835.:—
n=52 531cm 35‘— waheaooamtusaci .8 5955C:—

 

 

5on .3 533050 man. .3 EC .835: 535.5 5935mm 3:. 5h 35% man. he 5eE=Z 05.53.. 5 855.5: u: 035.

88

The following results refer speciﬁcally to each row of results in Table 14.

Row 1. Comparing the case of motor oil with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 2 tag moving at 300 fpm averaging 23.7 more tag reads per trial.
Row 2. Comparing the case of motor oil with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 300 fpm averaging 124.7 more tag reads per trial.
Row 3. Comparing the case of motor oil with a Gen 1 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 1 tag moving at 300 fpm averaging 139 more tag reads per trial.
Row 4. Comparing the case of motor oil with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with a Gen 2 tag moving at 300 fpm averaging 148.4 more tag reads per trial.
Row 5. Comparing the case of motor oil with a Gen 2 tag moving at 300 fpm
versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of motor
oil with 3 Gen 2 tag moving at 300 fpm averaging 162.7 more tag reads per trial.
Row 6. Comparing the case of motor oil with a Gen 1 tag moving at 600 fpm

versus the case of motor oil with a Gen 2 tag moving at 600 fpm, there was a

89

signiﬁcant statistical difference in the number of tag reads, with the case of motor

oil with a Gen 1 tag moving at 600 fpm averaging 14.3 more tag reads per trial.

Package Effect Results

Chips in Plastic Tubs

In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 6,681 instances in which the tag was detected. On
average there were 222.7 tag reads per trial with a standard deviation of 23.96.

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 1,869 instances in which the tag was detected. On
average there were 62.3 tag reads per trial with a stande deviation of 22.14.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 6,753 instances in which the tag was detected. On
average there were 225.1 tag reads per trial with a standard deviation of 15.59.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1,750 instances in which the tag was detected. On

average there were 58.3 tag reads per trial with a standard deviation of 9.49.

Chips in MSWF C

In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 2,739 instances in which the tag was detected. On

average there were 91.3 tag reads per trial with a standard deviation of 14.9.

90

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 969 instances in which the tag was detected. On
average there were 32.3 tag reads per trial with a standard deviation of 14.7.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 2,694 instances in which the tag was detected. On
average there were 89.8 tag reads per trial with a standard deviation of 7.5.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 820 instances in which the tag was detected. On

average there were 27.3 tag reads per trial with a standard deviation of 6.7.

Statistical Analysis of the Package Effect Test

Table 15 depicts the results from testing the ﬁxed effects involved in testing the

product effect on tag readability.

91

Table 15: Tests of Fixed Effects for Package Effect

 

 

 

 

 

 

 

 

 

Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
Product 1 232 1664.89 <.0001
Speed 1 232 3120.44 <.0001
Tag 1 232 1.00 0.3182
Product*Speed 1 232 656.00 <.0001
Product*Tag 1 232 0.37 0.5424
Speed*Tag 1 232 1.50 0.2221
Product*Speed*Tag l 232 0.13 0.7184

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 15.

Row 1. Statistical analysis output concluded that package type (plastic tubs, MSWFC)
had a signiﬁcant effect on the average number of tag reads per trial.

Row 2. Statistical analysis output concluded that speed (300 fpm or 600 fpm) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 3. Statistical analysis output concluded that tag type (Gen 1 or Gen 2) did not have
a signiﬁcant effect on the average nrunber of tag reads per trial.

Row 4. Statistical analysis output concluded that product by speed had a signiﬁcant effect
on the average number of tag reads per trial.

Row 5. Statistical analysis output concluded that product by tag did not have a
signiﬁcant interaction effect on the average number of tag reads per trial.

Row 6. Statistical analysis output concluded that speed by tag did not have signiﬁcant

interaction effect on the average number of tag reads per trial.

92

Row 7. Statistical analysis output concluded that product by speed by tag did not have a

signiﬁcant interaction effect on the average number of tag reads per trial.

Table 16 shows the statistical analysis output concluded that package type had a
signiﬁcant effect on the average amount of tag reads per trial, with the case of potato
chips in plastic tubs averaging 81.9 more tag reads per trial than the case of MSWFC.

Table 16: Average Number of Tag Reads Per Trial and Difference Between Means,
by Product Type

 

Product Estimate Standard Error DF tValue Pr>|t|

 

 

 

Metal 60.1833 1.4197 232 42.39 <.0001
Plastic 142.1 1 1.4197 232 100.09 <.0001
MD -81.9250 2.0078 232 -40.80 <.0001

 

 

 

 

 

 

 

 

Table 17 shows that statistical analysis output concluded that conveyor speed had
a signiﬁcant effect on the average number of tag reads per trial, with 300 fpm averaging

1 12.2 more tag reads per trial than 600 fpm.

93

Table 17: Average Number of Tag Reads Per Trial and Difference Between Means

 

 

 

 

 

by Conveyor Speed
Speed Least Squares Means
Speed Estimate Standard DF t Value Pr > |t|
Error
300 157.23 1.4197 232 110.74 <.0001
600 45.0667 1.4197 232 31.74 <.0001
MD 112.16 2.0078 232 55.86 <.0001

 

 

 

 

 

 

 

 

Table 18 shows that statistical analysis output concluded that tag type did not
have a signiﬁcant effect on the average number of tag reads per trial, with Gen 1
averaging 2 more tag reads per trial than Gen 2.

Table 18: Average Number of Tag Reads Per Trial and Difference Between Means

 

 

 

 

 

 

by Tag Type
Tag Least Squares Means
Tag Estimate Standard DF t Value Pr > |t|
Error
Genl 102.15 1.4197 232 71.95 <.0001
Gen2 100.14 1.4197 232 70.54 <.0001
MD 2.0083 2.0078 232 1.00 0.3182

 

 

 

 

 

 

 

Table 19 shows the statistical analysis output of the interaction between the

product and a conveyor speed.

94

Table 19: Average Number of Tag Reads Per Trial, Product by Speed

 

Product*Speed Least Squares Means

 

Product Speed Estimate Standard DF tValue Pr>|t|
Error

 

Metal 300 90.5500 2.0078 232 45.10 <.0001

 

Metal 600 29.8167 2.0078 232 14.85 <.0001

 

Plastic 300 223.90 2.0078 232 11 1.51 <.0001

 

Plastic 600 60.3167 2.0078 232 30.04 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 19.
0 Row 1. The case of MSWFC moving at 300 fpm averaged 90.6 reads per trial.
0 Row 2. The case of MSWFC moving at 600 fpm averaged 29.8 reads per trial.
0 Row 3. The case of plastic tubs moving at 300 fpm averaged 223.9 reads per trial.

0 Row 4. The case of plastic tubs moving at 600 fpm averaged 60.3 reads per trial.

Table 20 shows the statistical analysis output of the two way interaction between

the two products and the two speeds.

95

Table 20: Difference in Average Number of Tag Reads Per Trial Between Product

by Speed

 

Differences of Product*Speed Least Squares Means

 

 

 

 

 

 

 

Product Speed _Product _Speed Estimate Standard DF t Value Pr > It]
Error
Metal 300 Metal 600 60.7333 2.8395 232 21.39 <.0001
Plastic 300 Plastic 600 163.58 2.8395 232 57.61 <.0001
Metal 300 Plastic 300 - 1 33 .35 2.8395 232 -46.96 <.0001
Metal 300 Plastic 600 30.2333 2.8395 232 10.65 <.0001
Metal 600 Plastic 300 -1 94.08 2.8395 232 -68.35 <.0001
Metal 600 Plastic 600 -30.5000 2.8395 232 -10.74 <.0001

 

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 20.

0 Row 1. Comparing the case of MSWF C moving at 300 fpm versus the case of
MSWFC moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of MSWFC at 300 fpm averaging 60.7 more
tag reads per trial.

0 Row 2. Comparing the case of plastic tubs moving at 300 fpm versus the case of
plastic tubs moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of Plastic Tubs at 300 fpm averaging 163.6
more tag reads per trial.

0 Row 3. Comparing the case of MSWFC moving at 300 fpm versus the case of
plastic tubs moving at 300 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of plastic tubs at 300 fpm averaging 133.4

more tag reads per trial.

96

 

a Row 4. Comparing the case of MSWFC moving at 300 fpm versus the case of
plastic tubs moving at 600 fpm, there was a signiﬁcant statistical difference in the

number of tag reads, with the case of MSWFC at 300 fpm averaging 30.2 more

tag reads per trial.

0 Row 5. Comparing the case of MSWFC moving at 600 fpm versus the case of
plastic tubs moving at 300 fpm, there was a signiﬁcant statistical difference in the

number of tag reads, with the case of plastic tubs at 300 fpm averaging 194.08

more tag reads per trial.

0 Row 6. Comparing the case of MSWFC moving at 600 fpm versus the case of
plastic tubs moving at 600 fpm, there was a signiﬁcant statistical difference in the

number of tag reads, with the case of plastic tubs at 600 fpm averaging 30.5 more

tag reads per trial.

Table 21 shows the statistical analysis output of the interaction between product and

tag type.

Table 21: Average Number of Tag Reads Per Trial, Product by Tag Generation

 

 

 

 

 

 

 

 

 

 

 

 

 

Product*Tag Least Squares Means
Product Tag Estimate Standard DF t Value Pr > It]
Error
Metal Genl 61.8000 2.0078 232 30.78 <.0001
Metal Gen2 58.5667 2.0078 232 29.17 <.0001
Plastic Genl 142.50 2.0078 232 70.97 <.0001
Plastic Gen2 141.72 2.0078 232 70.58 <.0001

 

97

 

The following results refer speciﬁcally to each row of results in Table 21.
0 Row 1. The case of MSWFC with a Gen 1 tag averaged 61.8 reads per trial.
0 Row 2. The case of MSWFC with a Gen 2 tag averaged 58.6 reads per trial.
0 Row 3. The case of plastic tubs with a Gen 1 tag averaged 142.5 reads per trial.

0 Row 4. The case of plastic tubs with a Gen 2 tag averaged 141.7 reads per trial.

Table 22 shows the statistical analysis output of the two way interaction between
the product and tag type.

Table 22: Difference In Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

 

 

 

by Tag Type
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error
Metal Genl Metal Gen2 3.2333 2.8395 232 1.14 0.2560
Plastic Genl Plastic Gen2 0.7833 2.8395 232 0.28 0.7829
Metal Genl Plastic Genl —80.7000 2.8395 232 -28.42 <.0001
Metal Genl Plastic Gen2 ~79.9167 2.8395 232 —28. 14 <.0001
Metal Gen2 Plastic Genl -83.9333 2.8395 232 -2956 <.0001
Metal Gen2 Plastic Gen2 -83. 1500 2.8395 232 -29.28 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 22.

0 Row 1. Comparing the Gen 1 tag on a MSWFC versus the Gen 2 tag on a

MSWFC, there was not a signiﬁcant difference between the average number of

tag reads per trial, with the Gen 1 tag on MSWFC averaging 3.2 more tag reads

per trial.

98

 

0 Row 2. Comparing the Gen 1 tag on plastic tubs versus the Gen 2 tag on plastic
tubs, there mn_ot a signiﬁcant difference between the average number of tag
reads per trial, with the Gen 1 tag on plastic tubs averaging 0.78 more tag reads
per trial.

0 Row 3. Comparing the Gen 1 tag on MSWFC versus the Gen 1 tag on plastic
tubs, there was a signiﬁcant statistical difference in the number of tag reads with
the Gen 1 tag on plastic tubs averaging 80.7 more tag reads per trial.

0 Row 4. Comparing the Gen 1 tag on MSWFC versus the Gen 2 tag on plastic
tubs, there was a signiﬁcant statistical difference in the number of tag reads with
the Gen 2 tag on plastic tubs averaging 79.9 more tag reads per trial.

0 Row 5. Comparing the Gen 2 tag on MSWFC versus the Gen 1 tag on plastic
tubs, there was a signiﬁcant statistical difference in the number of tag reads with
the Gen 1 tag on plastic tubs averaging 83.9 more tag reads per trial.

0 Row 6. Comparing the Gen 2 tag on MSWFC versus the Gen 2 tag on plastic
tubs, there was a signiﬁcant statistical difference in the number of tag reads with

the Gen 2 tag on plastic tubs averaging 83.2 more tag reads per trial.

Table 23 shows the statistical analysis output of the interaction between tag type and

conveyor speed.

99

Table 23: Average Number of Tag Reads Per Trial, Tag Generation by Speed

 

 

 

 

 

 

Speed*Tag Least Squares Means
Tag Speed Estimate Standard DF t Value Pr > |t|
Error
Genl 300 157.00 2.0078 232 78.19 <.0001
Gen2 300 157.45 2.0078 232 78.42 <.0001
Genl 600 47.3000 2.0078 232 23.56 <.0001
Gen2 600 42.8333 2.0078 232 21.33 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 23.
0 Row 1. The case with a Gen 1 tag moving 300 fpm averaged 157 reads per trial.
0 Row 2. The case with a Gen 2 tag moving 300 fpm averaged 157.5 reads per trial.
0 Row 3. The case with a Gen 1 tag moving 600 fpm averaged 47.3 reads per trial.

0 Row 4. The case with a Gen 2 tag moving 600 fpm averaged 42.8 reads per trial.

Table 24 shows the statistical analysis output of the two way interaction between

the tag type and conveyor speed.

100

Table 24: Difference In Average Number of Tag Reads Per Trial Between Tag

 

 

 

 

 

 

 

 

Generation by Speed
Differences of Speed*Tag Least Squares Means

Tag Speed _Tag _Speed Estimate Standard DF t Value Pr > |t|

Error
Genl 300 Gen2 300 -0.4500 2.8395 232 -0.16 0.8742
Genl 300 Genl 600 109.70 2.8395 232 38.63 <.0001
Genl 300 Gen2 600 1 14.17 2.8395 232 40.21 <.0001
Gen2 300 Genl 600 110.15 2.8395 232 38.79 <.0001
Gen2 300 Gen2 600 1 14.62 2.8395 232 40.37 <.0001
Genl 600 Gen2 600 4.4667 2.8395 232 1.57 0.1 171

 

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 24.
0 Row 1. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving

at 300 fpm, there was not a signiﬁcant statistical difference in the number of tag

 

reads, with the Gen 2 tag at 300 fpm averaging 0.5 more tag reads per trial.

0 Row 2. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 1 tag at 300 fpm averaging 109.7 more tag reads per trial.

0 Row 3. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 1 tag at 300 fpm averaging 114.2 more tag reads per trial.

0 Row 4. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag

reads, with the Gen 2 tag at 300 fpm averaging 110.2 more tag reads per trial.

101

0 Row 5. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 2 tag at 300 fpm averaging 114.6 more tag reads per trial.

0 Row 6. Comparing the Gen 1 tag moving at 600 fpm versus the Gen 2 tag moving
at 600 fpm, there mtg a signiﬁcant statistical difference in the number of tag

reads, with the Gen 1 tag at 600 fpm averaging only 4.5 more tag reads per trial.

Table 25 shows the statistical analysis output of the interaction between the product,
tag type, and conveyor speed.

Table 25: Average Number of Tag Reads Per Trial, Product by Tag Generation by

 

 

 

The following results refer speciﬁcally to each row of results in Table 25.

 

 

 

 

 

 

 

 

Speed
Product*Speed*Tag Least Squares Means
Product Tag Speed Estimate Standard Error DF t Value Pr > |t|
Metal Genl 300 91.3000 2.8395 232 32.15 <.0001
Metal Gen2 300 89.8000 2.8395 232 31.63 <.0001
Metal Genl 600 32.3000 2.8395 232 1 1.38 <.0001
Metal Gen2 600 27.3333 2.8395 232 9.63 <.0001
Plastic Genl 300 222.70 2.8395 232 78.43 <.0001
Plastic Gen2 300 225.10 2.8395 232 79.28 <.0001
Plastic Genl 600 62.3000 2.83 95 232 21.94 <.0001
Plastic Gen2 600 58.3333 2.8395 232 20.54 <.0001

 

 

 

 

 

 

 

 

Row 1. The case of MSWFC with a Gen 1 tag moving 300 fpm averaged 91.3

reads per trial.

102

 

0 Row 2. The case of MSWFC with a Gen 2 tag moving 300 fpm averaged 89.8
reads per trial.

0 Row 3. The case of MSWF C with a Gen 1 tag moving 600 fpm averaged 32.3
reads per trial.

0 Row 4. The case of MSWFC with a Gen 2 tag moving 600 fpm averaged 27.3
reads per trial.

0 Row 5. The case of plastic tubs with a Gen 1 tag moving 300 fpm averaged 222.7
reads per trial.

0 Row 6. The case of plastic tubs with a Gen 2 tag moving 300 fpm averaged 225.].
reads per trial.

0 Row 7. The case of plastic tubs with a Gen 1 tag moving 600 fpm averaged 62.3
reads per trial.

0 Row 8. The case of plastic tubs with a Gen 2 tag moving 600 fpm averaged 58.3

reads per trial.

Metal-Metal Interactions
Table 26 shows the statistical analysis output of the three-way interaction between

the case of MSWFC, tag type, and conveyor speed.

103

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

an . ~.c em. ~m~ em . c... Sec... cce ~50 :52 cce .50 .802
_ooc.v emf mmm em 5... weevme ace ~50 :52 can ~50 :52
.cac.v mm... ~m~ em _ ca cccmsm cce .50 :52 can ~50 .30.).
_ooc.v mow. ~m~ em _ c... neeo. me ace ~50 :52 can —50 .35.).
.coc.v oeé. mm m em. _ a... 089% ace _50 :52 can ~50 :52
Sand two mmm e23. cacm. can ~50 :52 can .50 :52
Stu
E A .5 2...; . no 5.655. 3.253... 525i mahl ens—coil .5on we... «26?:—
n=52 83:—em 3.5,. whack—$652,. .o moo—55:5

 

 

50am .3 5.5550 man. .3 Sham—20 535.5 50.5.5 3...... 5.. 35¢ wary me 53:57. 0920.3. 5 355.05 ue~ 2...;

104

The following results refer speciﬁcally to each row of results in Table 26.

Row 1. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of MSWFC with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 1 tag moving at 300 fpm averaging 1.5 more tag reads per
trial.

Row 2. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of MSWFC with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with 3 Gen 1 tag moving at 300 fpm averaging 59 more tag reads per
trial.

Row 3. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of MSWFC with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 1 tag moving at 300 fpm averaging 63.9 more tag reads per
trial.

Row 4. Comparing the case of MSWF C with a Gen 2 tag moving at 300 fpm
versus the case of MSWFC with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 2 tag moving at 300 fpm averaging 57.5 more tag reads per
trial.

Row 5. Comparing the case of MSWF C with a Gen 2 tag moving at 300 fpm

versus the case of MSWFC with a Gen 2 tag moving at 600 fpm, there was a

105

signiﬁcant statistical difference in the number of tag reads, with the case of
MSWF C with a Gen 2 tag moving at 300 fpm averaging 62.4 more tag reads per
trial.

0 Row 6. Comparing the case of MSWFC with a Gen 1 tag moving at 600 fpm
versus the case of MSWFC with a Gen 2 tag moving at 600 fpm, there M a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 1 tag moving at 600 fpm averaging 5 more tag reads per

trial.

Metal-Plastic Interactions

Table 27 shows the statistical analysis output of the three-way interaction between

the two products, a tag, and a conveyor speed.

106

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.aca.v med..- ~n~ em.c.v 3.3.. can .50 0.8a... cce ~50 .80.).
.aca.v ave- ~n~ ems... mmna.e~- ace ~50 3.5.x. cce .50 .50.).
.aca.v 3...- ~n~ ems... cccc.cm- aae .50 0:5... ace .50 .50.).
.aca.v .cdv- ~m~ em .a... cc.~c . - cam ~50 0.3.0... cce .50 .30.).
.aca.v 3.....- ~m~ ems... 3.3.- can .50 0:5... ace .50 .80.).
.aca.v c..nn- ~n~ em.c.v cc.~m.- can .50 3.5.x. can ~50 .30.).
.aca.v 3..» ~m~ em.c.v reef; ace ~50 0:5... can ~50 .302
.aca.v nae ~m~ ems... cacm...~ ace .50 0:3... can ~50 .50.).
.aca.v cede- ~m~ ems... amen .- can ~50 0:3... can ~50 .30.).
.aca.v .~.c ~n~ ems... neeo.~n ace ~50 2.5.“. can .50 .50.).
.aca.v -... ~m~ em 5... cccc.o~ cce .50 035?. can .50 .50.).
.aca.v ~m.mm- ~n~ ems... cad»..- can ~50 0:5... can .50 .30.).
.aca.v ~r.~m- ~m~ ems... 3.7:..- can .50 0:3... can .50 .502
Stu
E A k. 0:.a> . mm. FEE—5m 0.05.3. c00aml mahl .2595! 50.5 mm... 8.69:.
050—). 55:—cm .50. wahecgamtggcm .o 5055....5

 

 

ma... .3 A0359). can 2...... 0.95.... 535.... 50.50». .02... 5.. 0.50”. we... .5 5....57. 035.5. a. 0055...... K~ 030...

50.5 .3 5.5550

107

 

 

 

 

.aca.v Nb..- Nmn cmSé 2...? cce ":00 03mm?— 25 n=00 .502
.aca.v .5..- SUN cm 3... hooém- .59 .:00 0.307. .50 ”:00 .50.).
.aca.v mud? "nu em :30 REE- can N=00 0:007. .50 ~50 .50.).

 

 

 

 

 

 

 

 

 

 

 

 

8.2.8. 2 2%..

108

The following results refer speciﬁcally to each row of results in Table 27.

Row 1. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 131.4 more tag reads per trial.
Row 2. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 2 tag moving at 300 fpm averaging 133.8 more tag reads per trial.
Row 3. Comparing the case of MSWFC with 3 Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 1 tag moving at 300 fpm averaging 29 more tag reads per
trial.

Row 4. Comparing the case of MSWFC with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWF C with a Gen 1 tag moving at 300 fpm averaging 33 more tag reads per
trial.

Row 5. Comparing the case of MSWFC with a Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 300 1pm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic

tubs with a Gen 2 tag moving at 300 fpm averaging 135.3 more tag reads per trial.

109

 

Row 6. Comparing the case of MSWFC with 3 Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 2 tag moving at 300 fpm averaging 27.5 more tag reads per
trial.

Row 7. Comparing the case of MSWFC with a Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of
MSWFC with a Gen 2 tag moving at 300 fpm averaging 31.5 more tag reads per
trial.

Row 8. Comparing the case of MSWF C with a Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 132.9 more tag reads per trial.
Row 9. Comparing the case of MSWFC with a Gen 1 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 190.4 more tag reads per trial.
Row 10. Comparing the case of MSWFC with a Gen 1 tag moving at 600 fpm
versus the case of Plastic Tubs with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic

tubs with a Gen 2 tag moving at 300 fpm averaging 192.8 more tag reads per trial.

110

 

 

 

Row 11. Comparing the case of MSWFC with a Gen 1 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 600 fpm averaging 30 more tag reads per trial.
Row 12. Comparing the case of MSWFC with a Gen 1 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 2 tag moving at 600 fpm averaging 26 more tag reads per trial.
Row 13. Comparing the case of MSWFC with a Gen 2 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 195.4 more tag reads per trial.
Row 14. Comparing the case of MSWFC with a Gen 2 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 2 tag moving at 300 fpm averaging 197.8 more tag reads per trial.
Row 15. Comparing the case of MSWFC with a Gen 2 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 600 fpm averaging 35 more tag reads per trial.
Row 16. Comparing the case of MSWFC with a Gen 2 tag moving at 600 fpm

versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there was a

111

signiﬁcant statistical difference in the number of tag reads, with the case of plastic

tubs with a Gen 2 tag moving at 600 fpm averaging 31 more tag reads per trial.

Plastic-Plastic Interactions

Table 28 shows the statistical analysis output of the three-way interaction between

the case of plastic tubs, a tag, and a conveyor speed.

112

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9.3.5 3.: nnu 3:; Seal” cce ~50 932:.— 89 :50 23a..—
Sccv mm. 5 mm” em 5... $.3— 8» «so 9:3:— ccm ~50 33a..—
Secv vmdv NMN cmSé cad: coo :50 9:3..— ecn N50 9:2...—
Secv 3.3. a? 3:3. unio— oce ~50 9:3?— ccm :50 9:3.“—
SSV 35m «MN em _ a... $62 26 :59 puma—m com :50 puma—m
3mg. 86. unu em 56 33d- com. ~50 9:3..— eom :56 32...:—
E A .5 33> a “a .Stm FEE—Sm Saﬁzmm 33ml nah} 8.695.. 3on Man. 338.:—
mcauE 83:—cm :54 wahtuoonmtuauoi u: 80:29.:5

 

 

.5on .3 guano-BU man. 3 Awash. gag—Av 835.5 5953— 3th com 33% nah. .8 moan—=2 0333‘ 5 09:20.5: "mu 933.

113

The following results refer speciﬁcally to each row of results in Table 28.

Row 1. Comparing the case of plastic tubs with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 300 fpm, there EMS!
a signiﬁcant statistical difference in the number of tag reads, with the case of
plastic tubs with 3 Gen 2 tag moving at 300 fpm averaging 2.4 more tag reads per
trial.

Row 2. Comparing the case of plastic tubs with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with 3 Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 160.4 more tag reads per trial.
Row 3. Comparing the case of plastic tubs with a Gen 1 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 1 tag moving at 300 fpm averaging 164.4 more tag reads per trial.
Row 4. Comparing the case of plastic tubs with a Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with a Gen 1 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic
tubs with a Gen 2 tag moving at 300 fpm averaging 162.8 more tag reads per trial.
Row 5. Comparing the case of plastic tubs with a Gen 2 tag moving at 300 fpm
versus the case of plastic tubs with 3 Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of plastic

tubs with a Gen 2 tag moving at 300 fpm averaging 166.7 more tag reads per trial.

114

0 Row 6. Comparing the case of plastic tubs with a Gen 1 tag moving at 600 fpm
versus the case of plastic tubs with a Gen 2 tag moving at 600 fpm, there 3%
a signiﬁcant statistical difference in the number of tag reads, with the case of
plastic tubs with a Gen 1 tag moving at 600 fpm averaging 4 more tag reads per

trial.

Package Shape Effect Results

Aluminum Cans:

In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 3,015 instances in which the tag was detected. On
average there were 100.5 tag reads per trial with a standard deviation of 13.5.

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 1,235 instances in which the tag was detected. On
average there were 41.2 tag reads per trial with a standard deviation of 11.7.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 4,601 instances in which the tag was detected. On
average there were 153.4 tag reads per trial with a standard deviation of 12.61.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1,273 instances in which the tag was detected. On

average there were 42.4 tag reads per trial with a standard deviation of 15.8.

115

Aluminum Bottles:

In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 2,719 instances in which the tag was detected. On
average there were 90.6 tag reads per trial with a standard deviation of 17.3.

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with a Gen 1
tag. Over the 30 trials, there were 1,319 instances in which the tag was detected. On
average there were 44 tag reads per trial with a standard deviation of 14.9.

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 3,166 instances in which the tag was detected. On
average there were 105.5 tag reads per trial with a standard deviation of 13.4.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1102 instances in which the tag was detected. On

average there were 36.7 tag reads per trial with a standard deviation of 12.7.

Aluminum Tins:
In Trial 1, the case traveled on the conveyor at 300 fpm equipped with a Gen 1 tag. Over
the 30 trials, there were 36 instances in which the tag was detected. On average there
were 1.2 tag reads per trial with a standard deviation of 1.9.

In Trial 2, the case traveled on the conveyor at 600 fpm equipped with 3 Gen 1
tag. Over the 30 trials, there were 13 instances in which the tag was detected. On

average there were 0.43 tag reads per trial with a standard deviation of 1.2.

116

In Trial 3, the case traveled on the conveyor at 300 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 1,324 instances in which the tag was detected. On
average there were 44.1 tag reads per trial with a standard deviation of 9.1.

In Trial 4, the case traveled on the conveyor at 600 fpm equipped with a Gen 2
tag. Over the 30 trials, there were 112 instances in which the tag was detected. On

average there were 3.7 tag reads per trial with a standard deviation of 2.3.

Statistical Analysis of the Package Shape Effect Test
Table 29 depicts the results from testing the ﬁxed effects involved in testing the

product effect on tag readability.

Table 29: Tests of Fixed Effects for Product Shape Effect

 

 

 

 

 

 

 

 

 

 

Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
Product 2 348 1230.39 <.0001
Speed 1 348 1901.54 <.0001
Tag 1 348 207.68 <.0001
Product*Speed 2 348 224.12 <.0001
Product*Tag 2 348 33.00 <.0001
Speed*Tag 1 348 228.69 <.0001
Product*Speed*Tag 2 348 1 1.72 <.0001

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 29.
Row 1. Statistical analysis output concluded that package shape (bottles, cans, tins) had a

signiﬁcant effect on the average number of tag reads per trial.

117

Row 2. Statistical analysis output concluded that speed (300 fpm or 600 ﬁrm) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 3. Statistical analysis output concluded that tag type (Gen 1 or Gen 2) had a
signiﬁcant effect on the average number of tag reads per trial.

Row 4. Statistical analysis output concluded that product by speed had a signiﬁcant effect
on the average number of tag reads per trial.

Row 5. Statistical analysis output concluded that product by tag had a signiﬁcant
interaction effect on the average number of tag reads per trial.

Row 6. Statistical analysis output concluded that speed by tag had signiﬁcant interaction
effect on the average number of tag reads per trial.

Row 7. Statistical analysis output concluded that product by speed by tag had a

signiﬁcant interaction effect on the average number of tag reads per trial.

Table 30 shows the statistical analysis output of the average number of tag reads
per trial by product.

Table 30: Average Number of Tag Reads Per Trial by Product Type

 

 

 

 

 

 

 

 

 

 

 

Product Least Squares Means
Product Estimate Standard DF t Value Pr > It]
Error
Bottle 69.2167 1.0820 348 63.97 <.0001
Can 84.3667 1.0820 348 77.97 <.0001
Tin 12.3750 1.0820 348 11.44 <.0001

 

The following results refer speciﬁcally to each row of results in Table 30.

0 Row 1. The case of bottles, averaged 69.2 tag reads per trial,

118

 

Row 2. The case of cans averaged 84.4 tag reads per trial

Row 3. The case of tins averaged 12.4 tag reads per trial.

Table 31 shows the statistical analysis output of the comparison of the average

number of reads per trial by product.

Table 31: Difference Average Number of Reads Per Trial by Product Type

 

 

 

 

 

 

 

 

 

 

 

 

Differences of Product Least Squares Means
Product _Product Estimate Standard DF t Value Pr > |t|
Error
Bottle Can -15.1500 1.5302 348 —9.90 <.0001
Bottle Tin 56.8417 1.5302 348 37.15 <.0001
Can Tin 71.9917 1.5302 348 47.05 <.0001

 

 

The following results refer speciﬁcally to each row of results in Table 31.

0 Row 1. Comparing the case of bottles versus the case of cans, there was a
signiﬁcant statistical difference in the number of tag reads, with the cans
averaging 15.2 more tag reads per trial.

0 Row 2. Comparing the case of bottles versus the case of tins, there was a
signiﬁcant statistical difference in the number of tag reads, with the bottles
averaging 56.8 more tag reads per trial.

0 Row 3. Comparing the case of cans versus the case of tins, there was a signiﬁcant
statistical difference in the number of tag reads, with the cans averaging 72 more

tag reads per trial.

119

Table 32 shows statistical analysis output concluded that conveyor speed had a
signiﬁcant effect on the average number of tag reads per trial, with 300 fpm averaging
54.5 more tag reads per trial than 600 fpm.

Table 32: Average Number of Tag Reads Per Trial and Difference Between Means

 

 

 

 

 

by Conveyor Speed
Speed Least Squares Means
Speed Estimate Standard DF t Value Pr > |t|
Error
300 82.5611 0.8835 348 93.45 <.0001
600 28.0778 0.8835 348 31.78 <.0001
MD 54.4833 1.2494 348 43.61 <.0001

 

 

 

 

 

 

 

Table 33 shows statistical analysis output concluded that tag type had a signiﬁcant

effect on the average number of tag reads per trial, with Gen 2 averaging 18 more tag

reads per trial than Gen 1.

Table 33: Average Number of Tag Reads Per Trial and Difference Between Means

 

 

 

 

 

 

 

 

 

 

 

by Tag Type
Tag Least Squares Means
Tag Estimate Standard DF t Value Pr > |t|
Error
Genl 46.3167 0.8835 348 52.43 <.0001
Gen2 64.3222 0.8835 348 72.81 <.0001
MD -18.0056 1.2494 348 -14.41 <.0001

 

120

 

 

Table 34 shows the statistical analysis output of the interaction between the
product and a conveyor speed.

Table 34: Average Number of Tag Reads Per Trial, Product by Speed

 

Product*Speed Least Squares Means

 

Product Speed Estimate Standard DF tValue Pr>|t|
Error

 

Bottle 300 98.0833 1.5302 348 64.10 <.0001

 

Bottle 600 40.3500 1.5302 348 26.37 <.0001

 

Can 300 126.93 1.5302 348 82.95 <.0001

 

Can 600 41.8000 1.5302 348 27.32 <.0001

 

Tin 300 22.6667 1.5302 348 14.81 <.0001

 

 

 

 

 

 

 

 

 

Tin 600 2.0833 1.5302 348 1.36 0.1743

 

The following results refer speciﬁcally to each row of results in Table 34.
0 Row 1. The case of bottles moving at 300 fpm averaged 98.1 reads per trial,
0 Row 2. The case of bottles moving at 600 fpm averaged 40.4 reads per trial
0 Row 3. The case of cans moving at 300 fpm averaged 126.9 reads per trial,
0 Row 4. The case of cans moving at 600 fpm averaged 41.8 reads per trial
0 Row 5. The case of tins moving at 300 fpm averaged 22.7 reads per trial
0 Row 6. The case of tins moving at 600 fpm averaged 2.1 reads per trial, this 125

r_ro_t a statistically signiﬁcant amount.

Table 35 shows the statistical analysis output of the interaction between the two

way interaction between the three products and the two speeds.

121

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

38v m3: ”3 :e. .N Era Se 5 =8 :5
58v 3% 3m :2 .N 22 .2 8n iv 8... =5
58v 3% w: :2 .N 22.9 8» .5 8n :5
88¢ 2.3. a: :2 .N 8.x: 8... .5 8n 55
58v 3% xx :2 .m 2: .3 8... :5 8n 56
58..» 3: ”3 SEN $83 Se .5 Se 2:8
527 a _ .w mam :8. .N 2%.: 8n .5. 80 2:8
202 So- as. .8 _ .m 83.. _- 8... =5 25 2:8.
58». 3.3- w: 2.2 .m 333- 3m 50 25 2:8
38v 3.3. mam :63 88.8 26 5 8m 2.8:
527 3.3 3. SEN SEE 8n 5 8n 2:8.
527 3.8 ”2 SEN 53% 8... 56 8m 2:3.
38.,» 2.2- m2 :2 .m 83%- 8... 5.0 8m 2:8
52? 38 ”E :r: .m 23% :8 2:2. 8m 2:8.
LOP-m
E A E 3.5 . “E Eagem 3253 88m- 835...: 325 .262.—
232 3:35 3qu uoonmtusuem no moo—_Eota

 

 

3on .3 335.5 :uoEom 3:. non— muaom man. he SAS—.2 09293” 5 353.59 ”mm 035.

122

 

_ _ooov — _md — 31m = Soﬁa __ mmwmdm = cce = E... __ can

 

 

a _

 

3.283 mm as;

123

The following results refer speciﬁcally to each row of results in Table 35.

Row 1. Comparing the case of bottles moving at 300 fpm versus the case of
bottles moving at 600 fpm, there was a signiﬁcant statistical difference in the
number of tag reads, with the case of bottles at 300 fpm averaging 57.7 more tag
reads per trial.

Row 2. Comparing the case of bottles moving at 300 fpm versus the case of cans
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 300 fpm averaging 28.9 more tag reads per trial.
Row 3. Comparing the case of bottles moving at 300 fpm versus the case of cans
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles at 300 fpm averaging 56.3 more tag reads per
trial.

Row 4. Comparing the case of bottles moving at 300 fpm versus the case of tins
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles at 300 fpm averaging 75.4 more tag reads per
trial.

Row 5. Comparing the case of bottles moving at 300 fpm versus the case of tins
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles at 300 fpm averaging 96 more tag reads per
trial.

Row 6. Comparing the case of bottles moving at 600 fpm versus the case of cans
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of

tag reads, with the case of cans at 300 fpm averaging 86.6 more tag reads per trial.

124

Row 7. Comparing the case of bottles moving at 600 fpm versus the case of cans

moving at 600 fpm, there was not a signiﬁcant statistical difference in the number

 

of tag reads, with the case of cans at 600 fpm averaging 1.5 more tag reads per
trial.

Row 8. Comparing the case of bottles moving at 600 fpm versus the case of tins
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles at 600 fpm averaging 17.7 more tag reads per
trial.

Row 9. Comparing the case of bottles moving at 600 fpm versus the case of tins
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles at 600 fpm averaging 38.3 more tag reads per
trial.

Row 10. Comparing the case of cans moving at 300 fpm versus the case of cans
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 300 fpm averaging 85.1 more tag reads per trial.
Row 11. Comparing the case of cans moving at 300 fpm versus the case of tins
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 300 fpm averaging 104.3 more tag reads per
trial.

Row 12. Comparing the case of cans moving at 300 fpm versus the case of tins
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 300 fpm averaging 124.9 more tag reads per

trial.

125

0 Row 13. Comparing the case of cans moving at 600 fpm versus the case of tins
moving at 300 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 600 fpm averaging 19.1 more tag reads per trial.

0 Row 14. Comparing the case of cans moving at 600 fpm versus the case of tins
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of
tag reads, with the case of cans at 600 fpm averaging 39.7 more tag reads per trial.

0 Row15. Comparing the case of tins moving at 300 fpm versus the case of tins
moving at 600 fpm, there was a signiﬁcant statistical difference in the number of

tag reads, with the case of tins at 300 fpm averaging 20.6 more tag reads per trial.

Table 36 shows the statistical analysis output of the interaction between product and

tag type.

Table 36: Average Number of Tag Reads Per Trial, Product by Tag Generation

 

 

 

 

 

 

 

 

 

Product*Tag Least Squares Means
Product Tag Estimate Standard DF t Value Pr > |t|
Error

Bottle Gen 1. 67.3000 1.5302 348 43 .98 <.0001
Bottle Gen2 71.1333 1.5302 348 46.49 <.0001
Can Genl 70.8333 1.5302 348 46.29 <.0001
Can Gen2 97.9000 1.5302 348 63.98 <.0001
Tin Genl 0.8167 1.5302 348 0.53 0.5939
Tin Gen2 23.9333 1.5302 348 15.64 <.0001

 

 

 

 

 

 

 

126

 

The following results refer speciﬁcally to each row of results in Table 36.
0 Row 1. The case of bottles with 3 Gen 1 tag averaged 67.3 reads per trial.
0 Row 2. The case of bottles with a Gen 2 tag averaged 71.3 reads per trial.
0 Row 3. The case of cans with a Gen 1 tag averaged 70.8 reads per trial.
0 Row 4. The case of cans with a Gen 2 tag averaged 97.9 reads per trial.
0 Row 5. The case of tins with a Gen 1 tag averaged 0.8 reads per trial; this guy;
a statistically signiﬁcant amount.

0 Row 6. The case of tins with a Gen 2 tag averaged 23.9 reads per trial.

Table 37 shows the statistical analysis output of the two way interaction between

the product and tag type.

127

Table 37: Difference In Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

by Tag Type
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error

Bottle Genl Bottle Gen2 -3.8333 2.1641 348 -1.77 0.0774
Bottle Genl Can Genl -3.5333 2.1641 348 -1.63 0.1034
Bottle Genl Can Gen2 -30.6000 2.1641 348 -l4. 14 <.0001
Bottle Genl Tin Genl 66.4833 2.1641 348 30.72 <.0001
Bottle Genl Tin Gen2 43.3667 2.1641 348 20.04 <.0001
Bottle Gen2 Can Genl 0.3000 2.1641 348 0.14 0.8898
Bottle Gen2 Can Gen2 -26.7667 2.1641 348 -12.37 <.0001
Bottle Gen2 Tin Genl 70.3167 2.1641 348 32.49 <.0001
Bottle Gen2 Tin Gen2 47.2000 2.1641 348 21.81 <.0001
Can Genl Can Gen2 -27.0667 2.1641 348 -12.51 <.0001
Can Genl Tin Genl 70.0167 2.1641 348 32.35 <.0001
Can Genl Tin Gen2 46.9000 2.1641 348 21.67 <.0001
Can Gen2 Tin Genl 97.0833 2.1641 348 44.86 <.0001
Can Gen2 Tin Gen2 73.9667 2.1641 348 34.18 <.0001
Tin Genl Tin Gen2 -23.1 167 2.1641 348 ~10.68 <.0001

 

The following results refer speciﬁcally to each row of results in Table 37.

 

0 Row 1. Comparing the case of bottles with a Gen 1 tag versus the case of bottles
with a Gen 2 tag, there was not a signiﬁcant statistical difference in the number of
tag reads, with the case of bottles with a Gen 2 tag averaging 3.8 more tag reads

per trial.

128

Row 2. Comparing the case of bottles with a Gen 1 tag versus the case of cans
with a Gen 1 tag, there moi a signiﬁcant statistical difference in the number of
tag reads, with the case of cans with a Gen 1 tag averaging 3.5 more tag reads per
trial.

Row 3. Comparing the case of bottles with a Gen 1 tag versus the case of cans
with a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 2 tag averaging 30.6 more tag reads per
trial.

Row 4. Comparing the case of bottles with a Gen 1 tag versus the case of tins with
a Gen 1 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of bottles with a Gen 1 tag averaging 66.5 more tag reads per
trial.

Row 5. Comparing the case of bottles with a Gen 1 tag versus the case of tins with
a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of bottles with 3 Gen 1 tag averaging 43.4 more tag reads per
trial.

Row 6. Comparing the case of bottles with a Gen 2 tag versus the case of cans
with a Gen 1 tag, there wn—ot signiﬁcant statistical difference in the number of
tag reads, with the case of bottles with a Gen 2 tag averaging 0.3 more tag reads
per trial.

Row 7. Comparing the case of bottles with a Gen 2 tag versus the case of cans

with a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag

129

reads, with the case of cans with a Gen 2 tag averaging 26.8 more tag reads per
trial.

Row 8. Comparing the case of bottles with a Gen 2 tag versus the case of tins with
a Gen 1 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of bottles with a Gen 2 tag averaging 70.3 more tag reads per
trial.

Row 9. Comparing the case of bottles with a Gen 2 tag versus the case of tins with
a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of bottles with a Gen 2 tag averaging 47.2 more tag reads per
trial.

Row 10. Comparing the case of cans with a Gen 1 tag versus the case of cans with
a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 2 tag averaging 27.1 more tag reads per
trial.

Row 11. Comparing the case of cans with a Gen 1 tag versus the case of tins with
a Gen 1 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 1 tag averaging 70 more tag reads per
trial.

Row 12. Comparing the case of cans with a Gen 1 tag versus the case of tins with
a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 1 tag averaging 46.9 more tag reads per

trial.

130

0 Row 13. Comparing the case of cans with a Gen 2 tag versus the case of tins with
a Gen 1 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 2 tag averaging 97.1 more tag reads per
trial.

0 Row 14. Comparing the case of cans with a Gen 2 tag versus the case of tins with
a Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag
reads, with the case of cans with a Gen 2 tag averaging 74 more tag reads per
trial.

0 Row 15. Comparing the case of tins with a Gen 1 tag versus the case of tins with a
Gen 2 tag, there was a signiﬁcant statistical difference in the number of tag reads,

with the case of tins with a Gen 2 tag averaging 23.1 more tag reads per trial.

Table 38 shows the statistical analysis output of the interaction between product and

tag type.

Table 38: Average Number of Tag Reads Per Trial, Product by Tag Generation

 

Speed*Tag Least Squares Means

 

Tag Speed Estimate Standard DF tValue Pr>|t|
Error

 

Genl 300 64.1111 1.2494 348 51.31 <.0001

 

Gen2 300 101.01 1.2494 348 80.85 <.0001

 

Genl 600 28.5222 1.2494 348 22.83 <.0001

 

Gen2 600 27.6333 1.2494 348 22.12 <.0001

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 38.

0 Row 1. A case with a Gen 1 tag moving 300 fpm averaged 64.1 reads per trial,

131

0 Row 2. A case with a Gen 2 tag moving 300 fpm averaged 101 reads per trial, a

statistically signiﬁcant amount.

0 Row 3. A case with a Gen 1 tag moving 600 fpm averaged 28.5 reads per trial, a

statistically signiﬁcant amount.

0 Row 4. A case with a Gen 2 tag moving 600 fpm averaged 27.6 reads per trial, a

statistically signiﬁcant amount.

Table 39 shows the statistical analysis output of the two way interaction between

 

 

 

 

 

 

 

 

the tag type and speed
Table 39: Difference In Average Number of Tag Reads Per Trial Between Tag Type
by Speed
Differences of Speed*Tag Least Squares Means
Tag Speed _Tag _Speed Estimate Standard Error DF t Value Pr > |t|
Genl 300 Gen2 300 -36.9000 1.7670 348 -20.88 <.0001
Genl 300 Genl 600 35.5889 1.7670 348 20.14 <.0001
Genl 300 Gen2 600 36.4778 1.7670 348 20.64 <.0001
Gen2 300 Genl 600 72.4889 1.7670 348 41.02 <.0001
Gen2 300 Gen2 600 73 .3778 1.7670 348 41.53 <.0001
' Genl 600 Gen2 600 0.8889 1.7670 348 0.50 0.6152

 

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 39.
0 Row 1. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving
at 300 fpm, there was a signiﬁcant statistical difference in the number of tag

reads, with the Gen 2 tag at 300 fpm averaging 36.9 more tag reads per trial.

132

0 Row 2. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 1 tag at 300 fpm averaging 35.6 more tag reads per trial.

0 Row 3. Comparing the Gen 1 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 1 tag at 300 fpm averaging 36.5 more tag reads per trial.

0 Row 4. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 1 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the nmnber of tag
reads, with the Gen 2 tag at 300 fpm averaging 72.5 more tag reads per trial.

0 Row 5. Comparing the Gen 2 tag moving at 300 fpm versus the Gen 2 tag moving
at 600 fpm, there was a signiﬁcant statistical difference in the number of tag
reads, with the Gen 2 tag at 300 fpm averaging 73.4 more tag reads per trial.

0 Row 6. Comparing the Gen 1 tag moving at 600 fpm versus the Gen 2 tag moving
at 600 fpm, there wot a signiﬁcant statistical difference in the number of tag

reads, with the Gen 1 tag at 600 fpm averaging only 0.9 more tag reads per trial.

Table 40 shows the statistical analysis output of the interaction between the product,

tag type, and conveyor speed.

133

 

Table 40: Average Number of Tag Reads Per Trial, Product by Tag Generation by

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Speed
Product*Speed*Tag Least Squares Means

Product Tag Speed Estimate Standard Error DF t Value Pr > |t|
Bottle Genl 300 90.6333 2.1641 348 41.88 <.0001
Bottle Gen2 300 105.53 2.1641 348 48.77 <.0001
Bottle Genl 600 43.9667 2.1641 348 20.32 <.0001
Bottle Gen2 600 36.7333 2.1641 348 16.97 <.0001
Can Genl 300 100.50 2.1641 348 46.44 <.0001
Can Gen2 300 153.37 2.1641 348 70.87 <.0001
Can Genl 600 41.1667 2.1641 348 19.02 <.0001
Can Gen2 600 42.4333 2.1641 348 19.61 <.0001
Tin Genl 300 1.2000 2.1641 348 0.55 0.5796
Tin Gen2 300 44.1333 2.1641 348 20.39 <.0001
Tin Genl 600 0.4333 2.1641 348 0.20 0.8414
Tin Gen2 600 3.7333 2.1641 348 1.73 0.0854

 

 

 

 

 

 

 

 

 

 

The following results refer speciﬁcally to each row of results in Table 40.

0 Row 1. The case of bottles with a Gen 1 tag moving 300 fpm averaged 90.6 tag
reads per trial.

0 Row 2. The case of bottles with a Gen 2 tag moving 300 fpm averaged 105.5 tag
reads per trial.

0 Row 3. The case of bottles with a Gen 1 tag moving 600 fpm averaged 44 tag

reads per trial.

134

0 Row 4. The case of bottles with a Gen 2 tag moving 600 fpm averaged 36.7 tag
reads per trial.

0 Row 5. The case of cans with a Gen 1 tag moving 300 fpm averaged 100.5 tag
reads per trial.

0 Row 6. The case of cans with a Gen 2 tag moving 300 fpm averaged 153.4 tag
reads per trial.

0 Row 7. The case of cans with a Gen 1 tag moving 600 fpm averaged 41.2 tag
reads per trial.

0 Row 8. The case of cans with a Gen 2 tag moving 600 fpm averaged 42.4 tag
reads per trial.

0 Row 9. The case of tins with a Gen 1 tag moving 300 fpm averaged 1.2 tag reads

per trial, this was not a statistically signiﬁcant amount.

 

0 Row 10. The case of tins with a Gen 2 tag moving 300 fpm averaged 44.1 tag
reads per trial.
0 Row 11. The case of tins with a Gen 1 tag moving 600 fpm averaged 0.43 tag

reads per trial, this was not a statistically signiﬁcant amount.

 

0 Row 12. The case of tins with 3 Gen 2 tag moving 600 fpm averaged 3.7 tag reads

per trial.

Bottle-Bottle Interaction
Table 41 shows the statistical analysis output of the three-way interaction between

the case of bottles, tag type, and conveyor speed.

135

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S _ od 3N mam wooed mmmms :8 ~50 0.39m coo :30 p.23—
Soov wean mam wooed oocw we :8 ~50 2.8: can ~50 2:3—
_ooo.v 9 .om wqm mecca Sew. _ o :8 :50 232. 9% N50 «:8:
_ooo.v 3.: mam wooed 892mm :8 N50 2:2.— gm 1:5 2:3—
_ooo.v mmf wvm wooed Sconce cce :50 2.8m can :30 «:8:
_ooov 23.- 3m 38. m cocci . can N50 2:3— ccn :50 2:2.
E A .5 2:5» . he Btu Eaveﬁm 2:523— voonml wahl 835.5! 3on ma... .2.—5.:—
mcaoE mega—Em 3.3: wahruooamtuaeoum he 89:95:;—

 

 

uoonm .3 cataconow Man. .3 Ame—tomb .2.—8.5 :ooEom _aE. 3m 33% want no .555: Z omauo>< 5 35.559 3v 033—.

136

The following results refer speciﬁcally to each row of results in Table 41.

Row 1. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of bottles with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 14.9 more tag reads per trial.

Row 2. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of bottles with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 46.7 more tag reads per trial.

Row 3. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of bottles with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 53.9 more tag reads per trial.

Row 4. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of bottles with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 61.6 more tag reads per trial.

Row 5. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of bottles with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 68.8 more tag reads per trial.

Row 6. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm versus

the case of bottles with a Gen 2 tag moving at 600 fpm, there was not a

137

signiﬁcant statistical difference in the number of tag reads, with the case of bottles

with a Gen 1 tag moving at 600 fpm averaging 7.2 more tag reads per trial.

Can-Can Interaction

Table 42 shows the statistical analysis output of the three-way interaction between

the case of cans, tag type, and conveyor speed.

138

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«3.0.: .96. awn 33$ 53;- cce ~50 :aU cce :50 .50
.aca.v muén wvn 38M. 8...: cce «EU 55 com «50 .50
58V 36m men 32:” 3.2 _ c8 :56 55 con N50 55
53V 3.3 new wooed 5.5.3 25 «EU 55 can :50 .50
58V and. awn mcocd nnmmdm =3 :50 .30 can .50 55
38V 3.5- wvm meccd Sowﬁm- com ~50 56 can :56 55
E A .5 33> . “a not”... unseaﬂm 82:5”.— uooaml web: aggil 3on man. 3355
2:32 8:35 .39... wahreooamtuacoum _o 825.5:5

 

 

.5on .3 55.22.00 wan. .3 Amuauv 335.5 5053— 32,—. .5.— mvaom man. he hos—~52 $303.. E 35.5%.: "NV 035-.

139

The following results refer speciﬁcally to each row of results in Table 42.

Row 1. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of cans with 3 Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 52.9 more tag reads per trial.

Row 2. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of cans with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 59.3 more tag reads per trial.

Row 3. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 58.1 more tag reads per trial.

Row 4. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus
the case of cans with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 112.2 more tag reads per trial.

Row 5. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 110.9 more tag reads per trial.

Row 6. Comparing the case of cans with a Gen 1 tag moving at 600 fpm versus

the case of cans with a Gen 2 tag moving at 600 fpm, there was not a signiﬁcant

 

140

 

statistical difference in the number of tag reads, with the case of cans with a Gen

2 tag moving at 600 fprn averaging 1.30 more tag reads per trial.

Tin-Tin Interaction

Table 43 shows the statistical analysis output of the three-way interaction between

the case of tins, tag type, and conveyor speed.

141

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Emno no.7 we». moood ooond- ooo ~50 :E. ooo :50 Eh.
_ooov ond— wvn mooosn ooovov ooo «55 E... oon «:00 :2.
_ooov «Ni— wvm moood ooozmv ooo :50 5h oom ~59 E...
£33 and. wvm moood nmmm.~- ooo ~50 a; oon .50 E...
muooo mud own moood rooﬁo ooo :50 E... oon :30 :2.
_ooo.v no.1- wen mooo.m 23.3- oom «50 E... oon :50 Eh.
E A an— o=_a> . h: Stu ouacuaom 9353mm Boa—ml nah: $3351 68.5 wan. .2.—5.:—
meoE noun—Em “may: unbroounmtoﬂxzm he 39:28.5

 

 

325 .3 :oﬂauouow nah. .3 35.5 335.5 nooEom _aih. hem moaom wan. no hoe—52 on???» E 35.-£me "mv 03.;

142

The following results refer speciﬁcally to each row of results in Table 43.

Row 1. Comparing the case of tins with a Gen 1 tag moving at 300 fpm versus the
case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant statistical
difference in the number of tag reads, with the case of tins with a Gen 2 tag
moving at 300 fpm averaging 42.9 more tag reads per trial.

Row 2. Comparing the case of tins with a Gen 1 tag moving at 300 fpm versus the
case of tins with a Gen 1 tag moving at 600 fpm, there M a signiﬁcant
statistical difference in the number of tag reads, with the case of tins with a Gen 1
tag moving at 300 fpm averaging 0.7 more tag reads per trial.

Row 3. Comparing the case of tins with a Gen 1 tag moving at 300 fpm versus the
case of tins with a Gen 2 tag moving at 600 fpm, there M a signiﬁcant
statistical difference in the number of tag reads, with the case of tins with a Gen 2
tag moving at 600 fpm averaging 2.5 more tag reads per trial.

Row 4. Comparing the case of tins with a Gen 2 tag moving at 300 fpm versus the
case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant statistical
difference in the number of tag reads, with the case of tins with a Gen 2 tag
moving at 300 fpm averaging 43.7 more tag reads per trial.

Row 5. Comparing the case of tins with a Gen 2 tag moving at 300 fpm versus the
case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant statistical
difference in the number of tag reads, with the case of tins with a Gen 2 tag
moving at 300 fpm averaging 40.4 more tag reads per trial.

Row 6. Comparing the case of tins with a Gen 1 tag moving at 600 fpm versus the

case of tins with a Gen 2 tag moving at 600 fpm, there was not a signiﬁcant

143

statistical difference in the number of tag reads, with the case of tins with a Gen 2

tag moving at 600 fpm averaging 3.3 more tag reads per trial.

Bottle-Can Interactions
Table 44 shows the statistical analysis output of the three-way interaction between

the bottles and cans, tag type, and conveyor speed.

144

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.ooo.v ...wn- wvn moood moo. .- oon «:00 ....0 ooo «.50 0.38.
.ooo.v vwom- ova moood rooﬁmo. oom .50 .30 ooo ~50 0.23—
536 on... own moood nnmn. ooo «.50 .30 ooo .50 e.82.
ooo»: o .oo ovn moood ooomd ooo 2.00 5.0 ooo .50 0.3.5
.ooo.v whmm- own good 3.2..- oon «:90 5.0 ooo .50 0.33.
.ooo.v 5.3- own moood mmmmdm- oon .50 .30 ooo .50 2:3.
ooo. .o no. 35. moood mnmo.m oom .50 .50 oon N50 2:2.
.ooo.v noom own moood ooo . .n o ooo n50 .30 oon ~50 2:3.
.ooo.v no. . N own moood roonéo ooo .50 .50 oom ~50 v.22.
.ooo.v no.2. o3. moood nnnwﬁ..- oom N50 5.0 oom ~50 2:2.
.ooo.v who. 8.». moood ooo~.mv ooo ~50 E.0 oon .50 2:2.
.ooo.v o . .o. own moooxn noovov ooo .50 5.0 oom .50 2:3.
.ooo.v omoN- own moood 3.9.3- oom N50 .50 oon .50 0.83.
v.oo.o «Nd- own moood noowd- oom .50 5.0 oom .50 233.
E A .n. 0......» . ".a— .85.... 3.65.5 33:5... team! ”Fri 835.51 .5on Ma... 3:3...—
2:32 3.53m 38‘. um .2....»on 1269... .c 80:98....5

 

 

5.3.3.50 we... 3 3.50 1.... 8.336 33.5.... 95... .5953. 3...... 5.. £53. an... no .21.... Z omega. :. 35.5.5: .3. 2......

.5on 3

145

 

_ «3...: = em... = wvm = mcocd

= card- = 25

 

 

~50 __ 5.0 4— 25 =N=o0= 2:3. _

 

_ van—d = mu..- = wen _f mcccd

= mmmvé. = cce

 

 

.50 __ :50 _— ccc «:00

 

 

 

 

2.5m. _

 

‘ 3.2.8. 3 2...;

146

The following results refer speciﬁcally to each row of results in Table 44.

Row 1. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of cans with 3 Gen 1 tag moving at 300 fpm, there a signiﬁcant statistical
difference in the number of tag reads, with the case of cans with a Gen 1 tag
moving at 300 fpm averaging 9.9 more tag reads per trial.

Row 2. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 62.7 more tag reads per trial.

Row 3. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of cans with 3 Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 49.5 more tag reads per trial.

Row 4. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 48.2 more tag reads per trial.

Row 5. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 1pm averaging 47.8 more tag reads per trial.

Row 6. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus

the case of cans with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant

147

statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 64.4 more tag reads per trial.

Row 7. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of cans with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 63.1 more tag reads per trial.

Row 8. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of cans with a Gen 1 tag moving at 300 fpm, there w_asn_ot a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 5 more tag reads per trial.

Row 9. Comparing the case of bottles with a Gen 1 tag moving at 600 1pm versus
the case of cans with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 56.5 more tag reads per trial.

Row 10. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of cans with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of cans
with a Gen 2 tag moving at 300 fpm averaging 109.4 more tag reads per trial.
Row 11. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of cans with a Gen 1 tag moving at 600 fpm, there M a
signiﬁcant statistical difference in the number of tag reads, with the case of bottles

with a Gen 1 tag moving at 600 fpm averaging 2.8 more tag reads per trial.

148

0 Row 12. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of cans with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of bottles
with a Gen 1 tag moving at 600 fpm averaging 1.5 more tag reads per trial.

0 Row 13. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of cans with a Gen 1 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of cans
with a Gen 1 tag moving at 300 fpm averaging 63.8 more tag reads per trial.

0 Row 14. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of cans with a Gen 2 tag moving at 300 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of cans
with a Gen 2 tag moving at 300 fpm averaging ] 16.6 more tag reads per trial.

0 Row 15. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of cans with a Gen 1 tag moving at 600 fpm, there w a
signiﬁcant statistical difference in the number of tag reads, with the case of cans
with a Gen 1 tag moving at 600 fpm averaging 4.4 more tag reads per trial.

0 Row 16. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of cans with a Gen 2 tag moving at 600 fpm, there was a
signiﬁcant statistical difference in the number of tag reads, with the case of cans
with a Gen 2 tag moving at 600 fpm averaging 5.7 more tag reads per trial.
Bottle-Tin Interactions
Table 45 shows the statistical analysis output of the three-way interaction between

the bottles and tins, tag type, and conveyor speed.

149

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

52.... NY"- wvn 33d 2.3.»- can «:00 a... :8 «:00 0.2.5
.aca.v 5.: wvm new...“ mnmmﬁm :3 2.00 ...... 2.0 «:00 0.2.5
.aca.v m . .m. wvn 33.». nmmﬁcv go «:00 =..—. 2.0 2.00 0.2.5
.aca.v an... wvm 2.2.6 nnmmdv :3 2.00 E... 2.0 2.00 0.2.5
coma... m9..- wvm meocd $3.... can 2.00 5... 2.0 2.00 0.3.5
2......v 3.2 53 good being. can 2.00 E... 2.0 2.00 0.3.5
.aca.v 3.3 mvm was... and... can 2.00 E... can ":00 0.3.5
.aca.v emdm wvm 2.2....” 9.. . c. 2.0 «:00 E... 8m «:00 0.3.5
.aca.v and». a: 30:6 2%... cce 2.00 E... can «:00 0....5
2......v 3.3 «3. 33d .53.. _ 0 :8 ~50 ....—. can 2.00 0.2.5
.aca.v amda men 2.2....” ego...” :8 «:00 E... can 2.00 0.3.5
.aca.v 5.0m men 30....” 2.3.3 2.0 2.00 E... can 2.00 0.2.5
2......v 0...... wvn mceod 2.3.3 3m ”:00 E... can 2.00 0.3.5
.aca.v «ﬂan wvn 30...». nmnvdw 8m 2.00 E... can 2.00 0.3.5
E A .5 0......» . “E 3.5 Emu—.05 02......5 600.51 wwki .2.—5.5.. .0025 Ma... .25....—
m....0.>. 8.035 .301. wmh¢c00nm10=uoi .0 0005.0...5

 

 

:0..a..0..00 ma... .3 .2.... .2... 0.0.2.5. 0.0.5.0.... .03... 50.505 .35. .05 .6005 us... .2. .02.... Z 0w....0>< a. 02.0.0.5: .mv 0305.

30am .3

150

 

— .aca.v = 5...... = 5.5 _—

moccﬁ

_ﬂceccdn = 2.0 =N:00 = E... __ cc...

 

 

~50 = 0.2.5

 

 

_ __ .2. __ s __

mcocd

= 25.95?” = 2.0 = 2.00: E... = 2.0

 

 

n:00= 0.2.5 _

 

2:250. mv 02.05.

151

The following results refer speciﬁcally to each row of results in Table 45.

Row 1. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 89.4 more tag reads per trial.

Row 2. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 46.5 more tag reads per trial.

Row 3. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 90.2 more tag reads per trial.

Row 4. Comparing the case of bottles with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 300 fpm averaging 86.9 more tag reads per trial.

Row 5. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 61.4 more tag reads per trial.

Row 6. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus

the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant

152

statistical difference in the number of tag reads, with the case of bottles with 3
Gen 2 tag moving at 300 fpm averaging 105.1 more tag reads per trial.

Row 7. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 101.8 more tag reads per trial.

Row 8. Comparing the case of bottles with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 300 fpm averaging 104.3 more tag reads per trial.

Row 9. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm versus
the case of tins with 3 Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 600 fpm averaging 42.8 more tag reads per trial.

Row 10. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of tins with a Gen 2 tag moving at 300 fpm, there w a
signiﬁcant statistical difference in the ntunber of tag reads, with the case of tins
with 3 Gen 2 tag moving at 300 fpm averaging 0.2 more tag reads per trial.

Row 11. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a

Gen 1 tag moving at 600 fpm averaging 43.5 more tag reads per trial.

153

0 Row 12. Comparing the case of bottles with a Gen 1 tag moving at 600 fpm
versus the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 1 tag moving at 600 fpm averaging 40.2 more tag reads per trial.

0 Row 13. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 600 fpm averaging 35.5 more tag reads per trial.

0 Row 14. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of tins with a Gen 2
tag moving at 300 fpm averaging 7.4 more tag reads per trial.

0 Row 15. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a
Gen 2 tag moving at 600 fpm averaging 36.3 more tag reads per trial.

0 Row 16. Comparing the case of bottles with a Gen 2 tag moving at 600 fpm
versus the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of bottles with a

Gen 2 tag moving at 600 fpm averaging 33 more tag reads per trial.

Can-Tin Interactions
Table 46 shows the statistical analysis output of the three-way interaction between

the cans and tins, tag type, and conveyor speed.

154

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0000.0 00.0. 000 0000.0 0000..- 000 ":00 E... 000 N000 =00
—2.0.v 00.0. 3.0 0000.0 0000.; 000 2.00 E... 000 «:00 000
.000.v 0 0.0. 000 0000.0 0000.0 000 «:00 E... 000 2.00 :00
.000.v .0.0. 000 0000.0 000 0.0.. 000 2.00 E... 000 2.00 :00
0000.0 00.0- 000 0000.0 0000.0- 000 «:00 0:. 000 . :00 :00
.000.v 00.0. 000 0000.0 0000.00 000 2.00 E... 000 2.00 :00
.000.v 00.0.. 3.0 0000.0 n . .00 . 000 2.00 E... 000 0:00 :00
.000.v 00.0.. 000 0000.0 00.0... 000 «:00 E... 000 0:00 :00
.000 .v 00.0.. 000 0000.0 00.0. 2.0 2.00 E... 000 «:00 :00
.000.v 00.00 00.0 0000.0 00.00. 000 «:00 E... 000 0:00 :00
.000.v 00.0 000 0000.0 0000.00 000 ~=00 Eh 000 2.00 :00
.000.v 00.00 090 0000.0 00.00. 000 2.00 E... 000 2.00 :00
.000.v 00.0. 0.0 0000.0 0000 .00 000 0000 Eh 000 2.00 0...0
.000.v 00.00 090 0000.0 0000.00 000 2.00 E... 000 2.00 000
E A .5 0.....> . “.0. 0...... 0.00050 0200...... 000001 03.1 20:00....I .0000 Ma... .0300...—
m=002 00.0....0 .0004 00.0000000200005 .0 0005.025

 

 

00.00.0000 M0... 2.. 2...... 0.... 0.000. 30:00.... 03... 000.505 20...... 00.. .0005 M0... .0 00.2.2.7. 0w000>< :. 0009.02.0— 00 02......

000mm .3

155

 

_ .000.v _ 00.2 __ 090 _—

0000.0

__ as... __ g.

 

 

.5 __ s __ g. T... __ .5 _

 

_ _ __ __

0000.0

= 02.0.3u = 000

 

 

_....__ =3. Ti _

 

00.2.80 0.. 2......

156

The following results refer speciﬁcally to each row of results in Table 46.

Row 1. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 99.3 more tag reads per trial.

Row 2. Comparing the case of cans with 3 Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 56.4 more tag reads per trial.

Row 3. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 100.7 more tag reads per trial.

Row 4. Comparing the case of cans with a Gen 1 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 300 fpm averaging 96.8 more tag reads per trial.

Row 5. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 109.2 more tag reads per trial.

Row 6. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus

the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant

157

statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 152.9 more tag reads per trial.

Row 7. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with 3 Gen
2 tag moving at 300 fpm averaging 149.6 more tag reads per trial.

Row 8. Comparing the case of cans with a Gen 2 tag moving at 300 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 300 fpm averaging 152.2 more tag reads per trial.

Row 9. Comparing the case of cans with a Gen 1 tag moving at 600 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 600 fpm averaging 40 more tag reads per trial.

Row 10. Comparing the case of cans with a Gen 1 tag moving at 600 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there moi a signiﬁcant
statistical difference in the number of tag reads, with the case of tins with a Gen 2
tag moving at 300 fpm averaging 3 more tag reads per trial.

Row 11. Comparing the case of cans with a Gen 1 tag moving at 600 fpm versus
the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen

1 tag moving at 600 fpm averaging 40.7 more tag reads per trial.

158

Row 12. Comparing the case of cans with a Gen 1 tag moving at 600 fpm versus
the case of tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
1 tag moving at 600 fpm averaging 37.4 more tag reads per trial.

Row13. Comparing the case of cans with a Gen 2 tag moving at 600 fpm versus
the case of tins with a Gen 1 tag moving at 300 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 600 fpm averaging 41.2 more tag reads per trial.

Row14. Comparing the case of cans with a Gen 2 tag moving at 600 fpm versus
the case of tins with a Gen 2 tag moving at 300 fpm, there wn_ot a signiﬁcant
statistical difference in the number of tag reads, with the case of tins with a Gen 2
tag moving at 300 fpm averaging 1.7 more tag reads per trial.

Row 15. Comparing the case of cans with a Gen 2 tag moving at 600 fpm versus
the case of tins with a Gen 1 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen
2 tag moving at 600 fpm averaging 42 more tag reads per trial.

Row 16. Comparing the case of cans with a Gen 2 tag moving at 600 fpm versus
the case of Tins with a Gen 2 tag moving at 600 fpm, there was a signiﬁcant
statistical difference in the number of tag reads, with the case of cans with a Gen

2 tag moving at 600 fpm averaging 38.7 more tag reads per trial.

159

CHAPTER 5 — CONCLUSIONS AND RECOMMENDATIONS

The results of the testing described in the previous chapter show that conveyor
speed had a signiﬁcant impact on the average number of tag reads per trial for RFID
transponders on packages. Additionally, product type, package type, and package shape
all had a signiﬁcant impact on the average number of tag reads per trial for RFID
transponders on packages moving on a conveyor. The results also show that the tag type,
had a signiﬁcant effect on the average number of tag reads per trial for product type
(ketchup and motor oil), and package shape (cans, bottles, and tins), but did not have a
signiﬁcant impact on average number of tag reads per trial for RFID transponders that
were placed on the package effect products (plastic tubs and metalized spiral wound
ﬁberboard containers (MSWFC)). The hypotheses outlined at the start of this research
were evaluated and determined to be true or false. The outcomes of this evaluation are
listed below. Additionally, the effect of the variable in both two-way and three-
interactions are also discussed and concluded
Hypothesis 1. Conveyor speed will not have an inﬂuence on the average amount of tag
reads per trial for RFID transponders.

This hypothesis was rejected. In the product effect, package effect, and package
shape effect tests, conveyor speed had a signiﬁcant effect on the percent readability of
transponders. The average amount of tag reads per trial with the conveyor operating at
300 feet per minute (fpm) was always signiﬁcantly greater than the average amount of

tag reads per trial with the conveyor operating at 600 fpm (Table 47).

160

Table 47 : Average Advantage in Tag Reads at 300 fpm Over 600 fpm

 

Average Advantage in Tag Reads at 300 fpm Over 600 fpm, Speed Least Squares Means

 

 

 

 

 

Effect Test Estimate Standard DF t Value Pr > |t|
Error
Mean Difference (MD) -109.l7 2.4358 232 44.82 <.0001
between 300fpm -600 fpm
(Product Effect)
MD between -1 12.16 2.0078 232 55.86 <.0001
300 fpm- 600 fpm
(Package Effect)
MD between -54.4833 1.2494 348 43.61 <.0001
300 fpm-600 fpm
(Package Shape Effect)
= — _

 

 

 

 

 

In all two-way and three-way interactions involving conveyor speed, speed had a

 

 

signiﬁcant effect on the average number of tag reads per trial, acting as the dominant

variable in the interactions for the product effect products, the package effect products,

and the product shape effect products (not involving tins).

The Effect of Conveyor Speed on the Product Effect Test, Ketchup and Motor Oil

Two-Way Interactions

Product by Speed

In studying Table 6, which shows the difference in the average number of tag

reads per trial between product by speed, some overall conclusions can be formed.

When speed was the only variable, 300 fpm always averaged more tag reads per

trial than 600 fpm (Table 48, Row 1 and 2).
When product and speed were the two variables, speed was the dominant

variable, making either product average more tag reads per trial when moving at

300 fpm compared to the product moving at 600 fpm (I‘ able 48, Row 3 and 4).

161

Table 48: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Caused by Speed
Differences of Product*Speed Least Squares Means
Product Speed _Product _Speed Estimate Standard DF t Value Pr > |t|
Error
Ketchup 300 Ketchup 600 74.6167 3 .4447 232 21.66 <.0001
Oil 300 Oil 600 143.72 3.4447 232 41.72 <.0001
Ketchup 300 Oil 600 44.5000 3 .4447 232 12.92 <.0001
Ketchup 600 Oil 300 - l 73.83 3 .4447 232 -50.46 <.0001
Tag Type by Speed

In studying Table 10, which shows the difference in the average number of tag

reads per trial between tag type by speed, some overall conclusions can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 49, Row 1 and 2).

0 When speed and tag type were the two variables, speed was the dominant
variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 49, Row 3 and 4).

Table 49: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag

 

 

 

 

 

 

Reads Caused by Speed
Differences of Speed*Tag Least Squares Means
Tag Speed _Tag _Speed Estimate Standard DF t Value Pr > |t|
Error
Genl 300 Genl 600 90.1667 3 .4447- 232 26.18 <.0001
Gen2 300 Gen2 600 128.17 3.4447 232 37.21 <.0001
Genl 300 Gen2 600 100.38 3.4447 232 29.14 <.0001
Gen2 300 Genl 600 l 17.95 3.4447 232 34.24 <.0001

 

 

 

 

 

 

 

 

 

 

 

162

Three-Way Interactions
Ketchup-Ketchup Interactions

In studying Table 12, which shows the difference in the average number of tag
reads per trial between product (ketchup) by tag type by speed, some overall conclusions
can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 50, Row 1 and 2).

0 When speed and tag type were the two variables, speed was the dominant
variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 ﬁam (T able 50, Row 3 and 4).

163

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

58... 0...: mm... m _ a... 8%... .00 .50 .223. 00... .50 02.38.
38.,» 3... mm m _ :3 mm... 6 8. .50 02.38. 80 .50 02.38.
327 8.2 m... m . :2. 8%.... 00. ~50 9:03. 000 NE... 9:05.
58v 0...: Rm 0 _ :3 28...... a... .50 9:03. 000 .50 .238.
00......
E A 0.. 0......» . 0... 00000.30 0.000.... 00000I ushl 00:000.] 00000 00... .0000...—
00002 0000000 .30.. 00.00000930000. .0 000000020.

 

 

002.0 .3 000000 .003. mu... .0 000.00 Z 0. 0000000.... 00.000000. 20.50000... 000003.-..00003. .00 0.00...

164

Ketchup-Motor Oil Interactions
In studying Table 13, which shows the difference in the average number of tag reads
per trial between product (ketchup and motor oil) by tag type by speed, some overall
conclusions can be formed.

0 When the product and speed were the two variables, speed was the dominant
variable, providing more average reads per trial for the product moving at 300
fpm compared to the product moving at 600 fpm (Table 51, Row 14).

0 When product, tag, and speed were all variables, speed was the dominant
variable, providing more average reads per trial for the product and tag moving at
300 fpm, compared to the product and tag moving at 600 fpm (Table 51, Row 5-
8).

165

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

527 no.3 «mm 25.00 $3.? 80 E00 :0 can «:00 9580!
5.5V 8.2- NM” 336 mc.mo—- 03. 7.00 :0 000 «:00 95800-
chv 3.5.- «mm m:.w.v 3.37 gm «:00 =0 80 2.00 9530!
38V 3”.» «MN 25... 39.3 80 «=0U =O can 1.00 02.80!
33V Eda- «mm 25... 2..wa can ~50 =0 :3 «:00 95300—
527 3.2 «mm m _ 5... 300.3 80 «:00 :0 can «:09 950.0!
38V 3.3. nnn 23... .8me can E00 =0 :00 2.09 9530!
327 and NMN 25... A532: :00 2.00 :6 can 2.00 530.0!
Eta
E A 5 0=_0> a mm 0.305% 03:50”.— 000mm| uahl 00=0chml 00am nah «0:00.:—
0:002 3.5—Em .0004 w0h0000mm20=005 .3 80:000....5

 

 

000mm 3 002.00 0003— wah .«0 038:2 5 00:20.5: .00002005 za>>00EH 0O 0202-.....300M 3m 030,—.

166

Motor Oil-Motor Oil Interactions
In studying Table 14, which shows the difference in the average number of tag reads
per trial between product (motor oil) by tag type by speed, some overall conclusions can
be formed.
0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 52, Row 1 and 2).
0 When speed and tag type were the two variables, speed was the dominant

variable, making either generation of tag average more tag reads per trial when
traveling at 300 {pm versus 600 fpm (Table 52, Row 3 and 4).

167

 

 

 

 

 

 

 

:xEV $3M «mu 95... 3.0: 80 E00 :0 can ”:00 :0

38V 3.0" «mm m—Sé 8.02 80 «:00 :0 can 2.00 :0

8.5V :3». «mm 95.0 mnde— 2.0 «:00 :0 can N50 :0

53V 8.3 an" 25..“ Enem— 30 E00 :0 can 2.00 :0
.805

E A .5 0:03 0 “3 0:00:05 00050.5 0025.. wahl 00:00.5I 000mm nah 00:00.:—

 

 

 

 

 

 

 

 

 

 

 

 

00002 83:3 0000:— w0h0000¢m000=08m .8 80:00.5:

 

000mm .3 00300 0003— wah .«0 038:2 5 00:03.05 53.00000:— >0>>00EH :0 .88—2-00 .8002 "Nm 0Eah

 

168

The Effect of Conveyor Speed on the Package Effect Test, Chips in Plastic Tubs and
Chips in MSWFC:

Two-Way Interactions
Product by Speed
In studying Table 20, which shows the difference in the average number of tag
reads per trial between product by speed, some overall conclusions can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 53, Row 1 and 2).

0 When product and speed were the two variables, speed was the dominant
variable, making either product average more tag reads per trial when moving at
300 fpm compared to the product moving at 600 fpm (Table 53, Row 3 and 4).

169

Table 53: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Caused by Speed
Differences of Product*Speed Least Squares Means
Product Speed _Product _Speed Estimate Standard DF t Value Pr > |t|
Error
Metal 300 Metal 600 60.7333 2.8395 232 21.39 <.0001
Plastic 300 Plastic 600 163.58 2.8395 232 57.61 <.0001
Metal 300 Plastic 600 30.2333 2.8395 232 10.65 <.0001
Metal 600 Plastic 300 -l94.08 2.8395 232 -68.35 <.0001
Tag Type by Speed

In studying Table 24, which shows the difference in the average number of tag
reads per trial between tag type by speed, some overall conclusion can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 54, Row 1 and 2).

0 When speed and tag type were the two variables, speed was the dominant
variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 54, Row 1 and 2).

Table 54: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag

 

 

 

 

 

 

Reads Caused by Speed
Differences of Speed*Tag Least Squares Means

Tag Speed _Tag _Speed Estimate Standard DF t Value Pr > |t|

Error
Genl 300 Genl 600 l09.70 2.8395 232 38.63 <.0001
Gen2 300 Gen2 600 l 14.62 2.8395 232 40.37 <.0001
Genl 300 Gen2 600 I 14.17 2.8395 232 40.21 <.0001
Gen2 300 Genl 600 l 10.15 2.8395 232 38.79 <.0001

 

 

 

 

 

 

 

 

 

 

 

170

Three-Way Interactions
MSWFC-MSWFC Interactions
In studying Table 26, which shows the difference in the average number of tag
reads per trial between product (MS WFC) by tag type by speed, some overall conclusions
can be formed.
0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 55, Row 1 and 2).
0 When speed and tag type were the two variables, speed was the dominant

variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 55, Row 3 and 4).

171

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Soc? NOE mmm 020.1 000%? 000 _000 .0002 can "000 .302
500;» 3.2 mmm 0m. _ 0.0 $00.3 000 «:00 .0002 00m. _000 .502
0000....» 002 gm 0m 5..., $018 000 N000 .502 can N000 .502
500V 00.3 mmm 0m 0 0.0 0000.3. 000 2.00 .502 can _000 .502
00000
E A 0A— 0:_0> 0 05 0000005 0.00—0.0mm 080ml will 00:00.5! 000mm wah 00:000.—
00002 000003 00004 u0h00000mt00000m .00 0000000005

 

 

0025 r3 000:00 0000M M08 .00 000.0: Z .0 000000.05 003000005 033-000.; 0h3m2-0m>>m2 "mm 030,—.

172

MSWFC-Plastic Tub Interactions

In studying Table 27, which shows the difference in the average number of tag
reads per trial between product (plastic tubs and MSWFC) by tag type by speed, some
overall conclusions can be formed.

0 When the product and speed were the two variables, speed was the dominant
variable, providing more average reads per trial for the product moving at 300
fpm compared to the product moving at 600 fpm (Table 56, Row 1-4).

0 When product, tag, and speed were all variables, speed was the dominant
variable, providing more average reads per trial for the products and tags moving
at 300 fpm compared to the products and tags moving at 600 fpm (Table 56, Row
5-8).

173

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.aca.v mm... "mm ems... cams" cce .:0U 0:3... 8m «:06 .302
.aca.v mod? NnN em 3... Sum: . - 3m .50 0:20.. 8e «:00 .502
.aca.v 8.x? Nan om . .3. 3.3 . - can «:00 0.025. 2.0 .56 .302
32.!» 3.x «mm 35... 53.3 80 «:06 0.7.0.“. can E00 .502
.aca.v mad... «an em 3... Eng . - can «:06 0.037. 26 «:06 .502
.aca.v cm... «nm em . oi beet—m com «:00 0.00:... can «:00 .502
.aca.v .9.»..- «an cm 3.9 3.3 . - can .:0U 0:37. 25 .50 .502
.aca.v «N.» unn 3...... 2.3.9.” .50 .50 0.0%.: can 200 .502
30.5.
E A .5 0:.:> . “a 33:05 2.2:..0”. .000le wahl 3309...! 600.5 ME, .2.—5....
0:02 00:35 00.00:. wak+c00am103953 30:03.59 .

 

 

.0025 3 @0300 06:0”. nah .0 03:57. :. 00:20.59 53.00.52: b03003... 3.... 0.0%...-Um3m2 “em 0.5.3..

174

Plastic Tub-Plastic Tub Interactions
In studying Table 28, which shows the difference in the average number of tag
reads per trial between product (plastic tubs) by tag generation by speed, some overall
conclusions can be formed.
0 When speed was the only variable, 300 fpm always averaged more tag reads per
tn'al than 600 fpm (Table 57, Row 1 and 2).
0 When speed and tag type were the two variables, speed was the dominant

variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 57, Row 3 and 4).

175

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.aca.v vmév NmN $36 3.9— :8 .:00 0.00.0... can ~:00 0.00:...
.aca.v 36v nnm cm 5... 3.3. cce ~:00 0.000.”. can .:00 0.090...
Eccv mm..v «mm om . c... $.90. 25 ~50 0.0007. can «:00 0.03.9.
.aca.v 206m Nmn 02:... 3.3. 80 :00 0.00.0.0. com .:00 0.00:7.
0000...
E A 0.. 0:.:> 0 0.: BEESm 0:05.00... c0005.. wahl 0000.000ml 03am ms... 00:00....
3:02 0000:.0m 030‘. uﬁTU00amL0=Uo0m .o 80:000....5

 

 

.0025 r. 6000.000 0.000% nah no 000.5: Z .0. 00:000....5 50.00.0000:— z:3-0000.,., 35,—. 0.00.03-30; 0.00:... Km 0.0.9.,

176

The Effect of Conveyor Speed on the Package Shape Effect Test, Metal Bottles,
Metal Cans, and Metal Tins

Two-Way Interactions
Product by Speed
In studying Table 35, which shows the difference in the average number of tag
reads per trial between product by speed, some overall conclusions can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 58, Row 1-3).

0 When product and speed were the two variables, speed was the dominant
variable between only bottles and cans. The product moving at 300 fpm always
averaged more tag reads per trial than the product moving at 600 fpm (Table 58,
Row 4-9).

177

Table 58: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Caused by Speed
Differences of Product*Speed Least Squares Means
Product Speed _Product _Speed Estimate Standard DF t Value Pr > |t|
Error
Bottle 300 Bottle 600 57.7333 2.1641 348 26.68 <.0001
Can 300 Can 600 85.1333 2.1641 348 39.34 <.0001
Tin 300 Tin 600 20.5833 2.1641 348 9.51 <.0001
Bottle 300 Can 600 56.2833 2.1641 348 26.01 <.0001
Bottle 600 Can 300 -86.5833 2.1641 348 -40.01 <.0001
Bottle 300 Tin 600 96.0000 2.1641 348 44.36 <.0001
Bottle 600 Tin 300 17.6833 2.1641 348 8.17 <.0001
Can 300 Tin 600 124.85 2.1641 348 57.69 <.0001
Can 600 Tin 300 19.1333 2.1641 348 8.84 <.0001
Tag Type by Speed

 

In studying Table 39, which shows the difference in the average number of tag

reads per trial between tag type by speed, some overall conclusions can be formed.

When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 59, Row 1 and 2).

When speed and tag type were the two variables, speed was the dominant
variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 59, Row 3 and 4).

178

Table 59: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag
Reads Caused by Speed

 

Differences of Speed*Tag Least Squares Means

 

Tag Speed _Tag _Speed Estimate Standard DF tValue Pr>|t|
Error

 

Genl 300 Genl 600 35.5889 1.7670 348 20.14 <.0001

 

Gen2 300 Gen2 600 73.3778 1.7670 348 41.53 <.0001

 

Genl 300 Gen2 600 36.4778 1.7670 348 20.64 <.0001

 

Gen2 300 Genl 600 72.4889 1.7670 348 41.02 <.0001

 

 

 

 

 

 

 

 

 

 

 

Three-Way Interaction
Bottle-Bottle Interaction
In studying Table 41, which shows the difference in the average number of tag
reads per trial between product (bottles) by tag type by speed, some overall conclusion
can be formed.

0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 60, Row 1 and 2).

0 When speed and tag type were the two variables, speed was the dominant
variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 60, Row 3 and 4).

Can-Can Interaction
In studying Table 42, which shows the difference in the average number of tag
reads per trial between product (cans) by tag type by speed, some overall conclusions can
be formed.
0 When speed was the only variable, 300 fpm always averaged more tag reads per
trial than 600 fpm (Table 61, Row 1 and 2).
0 When speed and tag type were the two variables, speed was the dominant

variable, making either generation of tag average more tag reads per trial when
traveling at 300 fpm versus 600 fpm (Table 61, Row 3 and 4).

179

 

 

.aca.v 00.00. avn 300d cud: 80 2000 :00 can N000 :00

 

 

 

.aca.v bad— nvn moved hooedm 80 N000 :00 can 1.00 :00
500V mndm 3n 32.6 mad. . .50 «:00 :00 can ":00 :00
.aca.v and. «um 309m mmmmdm 80 .:00 :00 can .:00 :00

 

EA...”— 0:.0>0 0.: 0000.0 00005.5 000E.0nm 0025.. “0... 00:000.... 080m ma... 00:00.5

 

 

 

 

 

 

 

 

 

 

 

0:002 00000.5 300‘. 00.....000am000:000n. ..0 80:000....5

 

 

 

000mm .3 000.000 0003. m0... .«0 000.0007. .0. 0000000.: 50.000000:— z03-0000..—. 500-000 30 0.0.0...

 

300w. N. .om 30m mooo. m 000m. . 0 .50 2000 0.30m 8m «:00 0.30m

 

.ooo.v .08. gm 300. m coco. mm. 000 «:00 0.0.05 can 2.00 0.80..

 

.ooo.v mvdm w; wooed 253.0 :00 N50 0.23. ea N000 0.30m

 

.ooo.v mm. m. 3% moco. m $000.. .50 .000 0.30.. can :00 0.025

 

0000”,.
_0_A0A. 0:.0>. .2. 0000505 000000.00... 000aml M0...

00:00.51 003m 00.. «0:00....

 

 

 

 

 

 

 

 

 

 

 

0:002 0000va 0000‘. “0.0100005 10:000.. ..0 80:000....5

 

 

 

000mm .3 000000 0003. w0h .00 000.807. .0. 0000000.: 50.000002: 005-0000; 05090.33. 50 0.0.05.

180

Tin-Tin Interaction
In studying Table 43, which shows the difference in the average number of tag
reads per trial between product (tins) by tag type by speed, some overall conclusions can
be formed.

0 When speed was the only variable:

0 With the case of tins traveling 300 fpm and the case of tins traveling 600
fpm equipped with a Gen 1 tag, there was no statistical difference in the
average number of tag reads per trial (Table 62, Row 1).

0 With the case of tins traveling 300 fpm and the case of tins traveling 600
fpm equipped with a Gen 2 tag, speed was the dominant variable, with the
case of tins moving at 300 fpm averaging 40.4 more tag reads per trial.
(Table 62, Row 2).

0 When speed and tag type were the two variables:

0 With the case of tins with a Gen 1 tag was moving 300 fpm, there was not
a signiﬁcant statistical difference in the average number of tag reads per
trial compared to the case of tins with 3 Gen 2 tag moving 600 fpm (Table
62, Row 3).

o The case of tins with a Gen 2 tag moving 300 fpm, averaged more tag
reads per trial than the case of tins with a Gen 1 tag moving 600 fpm.

I Since the dominant tag type and dominant speed prevailed in this
particular scenario, it is difﬁcult to say that speed or tag type was
the dominant variable in this three-way interaction (Table 62, Row
4)

181

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.000.v w«.v. men 300.». 000...: 000 .:00 5... 00m «:00 5...
0000.0 3.0- awn 2.00..” nmmm.«- 000 «:00 5... 00m .:00 E...
.000.v 0«.m . mum 300.». 0009.00 000 «:00 0.... 00m «:00 5...
300.0 m«.0 wen m000.m $00... 000 —:00 5... 00m .000 E...
0000“,.
_0_ A 00. 0:.0> 0 0.: 000055 000E33— 000nml w0.—.I 0000000.! 0025 w0... 00:00.:—
m:002 0000:.0m 0.000... 000.00000mi00000m 00 80:000....5

 

 

000mm .3 0000.00 0003. w0... .00 000.8: Z :. 000000.05 60.00008:— h0>>000ah 0.20:. «0 0.00M.

182

Bottle-Can Interactions
In studying Table 44, which shows the difference in the average number of tag
reads per trial between two products (bottles and cans) by tag generation by speed, some
overall conclusions can be formed.

0 When product and speed were the two variables, speed was the dominant
variable, providing more average reads per trial for the product moving at 300
fpm compared to the product moving at 600 fpm (Table 63, Row 1-4).

0 When product, tag, and speed were all variables, speed was the dominant
variable, providing more average reads per trial for the product and tag moving at
300 fpm compared to the products and tags moving at 600 fpm (Table 63, Row 5-
8).

183

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.000.v 00.0” 000 0000.0 5005.00- 000 _:00 :00 000 «:00 0.38—
~000.v 00. — « 000 0000.0 $00.00 000 _ :00 :00 000 «:00 0.30m
S00.v 3.00. 000 0000.0 0v.00~- 000 «:00 :00 000 _:00 0.330
_000.v 050— 000 0000.0 00000.. 000 «:00 :00 000 _:00 0.3.5
_000.v _ _00- 000 0000.0 00.0- T 000 «:00 :00 000 «:00 0.00m
~000.v «0.0« 000 0000.0 000 _ .00 000 «:00 :00 000 «:00 0:08—
_000.v 5.0—- 090 0000.0 0000.00- 000 2.00 :00 000 E00 0:25
~000.v 0_ .0— 00.0 0000.0 309.3. 000 —:00 :00 000 _:00 0.30m
.5: u
E A .5 0=_a> 0 no 0:00:80 03:53. 00000! nah! 00:09:... 00000 ME. 00:00....—
m:a0_z 00.5100 .000: waki000000.0:00..n_ 3 80:20.05

 

 

000mm .3 000.500 “000% wan. 00 .50—5.2 5 00:20.05 £000.23:— 5030030. 50.0025 "00 030,—.

184

Bottle-Tin Interactions and Can-Tin Interaction
In both bottle-tin and can-tin interactions, product type, not speed, was not the
dominant variable. This is discussed in more detail in the bottle-tin and can-tin sections

under the package shape effect section within the discussion of hypotheses 4.

Hypothesis 2. Product will not have an inﬂuence on the average amount of tag reads per
trial for RFID transponders on packages moving on a conveyor.

This hypothesis was rejected. The case of motor oil averaged 64.7 more tag reads
per trial than the case of ketchup (Table 64).

Table 64: Average Number of Tag Reads Per Trial and Difference Between Means,

 

 

 

 

 

 

by Product Type
Product Least Squares Means
Product Estimate Standard DF t Value Pr > |t|
Error
Ketchup 74.4583 1.7223 232 43 .23 <.0001
Oil 139.13 1.7223 232 80.78 <.0001
MD -64.6667 2.4358 232 -26.55 <.0001

 

 

 

 

 

 

 

 

 

Product also had an effect in two-way and three-way interactions. When product
was the only variable, motor oil always performed better than ketchup. In two-way and
three-way interactions, in which product and tag type were the variables, the dominant
variable was product (motor oil), making either generation of tag attached to the case of

motor oil average more tag reads per trial.

185

The Product Effect Test, Ketchup and Motor Oil

Two-Way Interactions
Product by Speed
In studying Table 6, which shows the difference in the average number of tag
reads per trial between product by speed, some overall conclusions can be formed.

0 When product was the only variable, motor oil always averaged more tag reads
per trial than ketchup (Table 65).

Table 65: Difference in Average Number of Tag Reads Per Trial Between Product
by Speed ‘

 

Differences of Product*Speed Least Squares Means

 

 

 

Product Speed _Product _Speed Estimate Standard DF t Value Pr > |t|
Error

Ketchup 300 Oil 300 -99.2167 3.4447 232 -28.80 <.0001

Ketchup 600 Oil 600 -30.1 167 3.4447 232 -8.74 <.0001

 

 

 

 

 

 

 

 

 

 

 

Product by Tag Generation
In studying Table 8, which shows the difference in the average number of tag
reads per trial between product by tag type, some overall conclusions can be formed.

0 When product was the only variable, motor oil always averaged more tag reads
per trial than ketchup (Table 66, Row 1 and 2).

0 When product and tag type were the two variables, product was the dominant
variable, making either generation of tag average more tag reads per trial when on
the case of motor oil versus the case of ketchup (Table 66, Row 3 and 4).

186

Table 66: Difference in Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

by Tag Generation
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error
Ketchup Genl Oil Genl -68.7667 3.4447 232 -l9.96 <.0001
Ketchup Gen2 Oil Gen2 -60.5667 3.4447 232 -17.58 <.0001
Ketchup Genl Oil Gen2 -73.4500 3.4447 232 -21.32 <.0001
Ketchup ﬁenz Oil Genl -55.8833 3.4447 232 -l6.22 <.0001
a-ﬁ- ==H====nnad== —= E

 

 

 

Three-Way Interactions
Ketchup-Motor Oil Interactions
In studying Table 13, which shows the difference in the average number of tag
reads per trial between product Gretchup and motor oil) by tag type by speed, some

overall conclusions can be formed.

0 When the product was the only variable (tag type and speed were the same)
motor oil always averaged more reads per trial than ketchup (Table 67).

0 When the product and tag type were the two variables, product was the
dominant variable, providing more average reads per trial for either generation of

tag attached to the case of motor oil compared to the generation of tag on the case
of ketchup (Table 68).

187

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 
 

 

0:80:00. 00000 3 5.50050

  
 

:53. .50- «an 22.... ~22... 80 .50 .5 08 «80 9:58.

.aca.v 0.....- an” 25... 88.0.- 80 ~50 .5 80 .50 9.5.8.

.aca.v 00.2- N2 5...... 03.....- 8m .50 .5 08 ~50 9.5.8.

.027 8...". «ma 2:... 8.2.- 8... ~50 .5 8m .:00 02.38.

z A .0 2.3, 0 .5 3.2.... 0.023 305...... 08%- »ﬁ- .262..- 0800 was .025...
0:00.). 00.5.50 .50.. «00.0000000050000 5 80:05.05

00... 3 0.0 .55.). 0:0 0.50.3.0 00:00.... 50500 5...... 00.. 00000 0....- .5 000:... Z 00000.5. :0 00:05.05 "00 0.00.0.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.000.v 00.0- «0« . 0.5.0 - 0000.00...- 000 «,:00 . -. :0 000 «:00- -02—000:-

.000.v 00.0.- «0« 0.5.0 000«.00- 000 .:00 :0 000 .:00 0.50.0!

.000.v 00.0.- «0« 0 .50 000 .00- 000 «:00 :0 000 «:00 0.50.0:

.000.v 0«..«- «0« 0.0.0.0 0000.- 000 .:00 :0 000 .:00 02.0.0!

E A 00 0:03 0 a... 5.9.0 0.50530 05500.0 00000.: 00....- .0:00..0I 00000 00... 00:00....
0:00.). 00.3.0.0 .304 00.000000000000000 5 80:05:...

 

9:30:00. 00000 3 3:50:00. 5.50050

00... 3 0.0 .55.). 0:0 0.30.3.0 00:09.0 50.500 5...... 000 0.0000 00... .5 000:... Z 000003. :. 00:05.05 "0.0 0.00...

 

188

Hypothesis 3. Packaging materials will not have an inﬂuence on the average amount of
tag reads per trial for RF ID transponders on packages while moving on a conveyor.

This hypothesis was rejected. The case of potato chips in plastic tubs averaged
81.9 more tag reads per trial than the case of potato chips in metalized spiral wound

ﬁberboard containers (MSWFC) (Table 69).

Table 69: Average Number of Tag Reads Per Trial and Difference Between
Means, by Product Type

 

Product Estimate Standard DF tValue Pr>|t|
Error

 

Metal 60.1833 1.4197 232 42.39 <.0001

 

Plastic 142.11 1.4197 232 100.09 <.0001

 

MD -81.9250 2.0078 232 -40.80 <.0001

 

 

 

 

 

 

 

 

Package materials also had an effect in two-way and three-way interactions.
When package was the only variable, plastic tubs always performed better than ketchup.
In two-way and three-way interactions, in which product and tag type were the variables,
the dominant variable was package type (plastic tubs), making either generation of tag

attached to the case of plastic tubs average more tag reads per trial.

Evaluation of Package Effect Test, Chips in Plastic Tubs and Chips in MSWFC
Two-Way Interactions

Product by Speed

189

In studying Table 20, which shows the difference in the average number of tag
reads per trial between product by speed, some overall conclusions can be formed.

0 When product was the only variable, plastic tubs always averaged more tag

reads per trial than MSWFC (Table 70).

Table 70: Difference in Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

by Speed
Differences of Product*Speed Least Squares Means
Product Speed _Product _Speed Estimate Standard DF tValue Pr>|t|
Error
Metal 300 Plastic 300 -l33.35 2.8395 232 -46.96 <.0001
Metal 600 Plastic 600 -30.5000 2.8395 232 —lO.74 <.0001
Product by Tag Type

In studying Table 22, which shows the difference in the average number of tag
reads per trial between product by tag type, some overall conclusion can be formed.

0 When product was the only variable, plastic tubs always averaged more tag reads
per trial than MSWFC (Table 71, Row 1 and 2).

0 When product and tag type were the two variables, package was the dominant
variable, making either generation of tag average more tag reads per trial when on
the case of plastic tubs versus the case of MSWF C (Table 71, Row 3 and 4).

190

Table 71: Difference In Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

by Tag Type
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > It]
Error
Metal Genl Plastic Genl -80.7000 2.8395 232 -28.42 <.0001
Metal Gen2 Plastic Gen2 -83.1500 2.8395 232 -29.28 <.0001
Metal Genl Plastic Gen2 -79.9 I67 2.8395 232 -28. I4 <.0001
Metal Gen2 Plastic Genl -83.9333 2.8395 232 -29.56 <.0001

 

 

 

 

 

 

 

 

 

 

 

Three-Way Interactions
MSWFC-Plastic Tub Interactions
In studying Table 27, which shows the difference in the average number of tag
reads per trial between product (plastic tubs and MSWFC) by tag type by speed, some

overall conclusions can be formed.

0 When the product was the only variable (tag type and speed were the same)
plastic tubs always averaged more average reads per trial than MSWFC (Table
72).

0 When the product and tag were the two variables, product was the dominant
variable, providing more average reads per trial for either generation of tag
attached to the case of plastic tubs compared to the generation of tag on the case
of MSWFC (Table 73)

191

 

.000.V :0. «MN 00 .0.0 0.000.00- 000 .:0U 0.50..— 000 «:09 .302

 

 

 

.000.V 00.0- NMN 00.0.0 0000.0"- 000 «:00 0:00... 000 .:00 5.0—).
.000.v 0.60- «an 00.0.0 00.~n.- 000 .:00 0.50.0 000 «:00 5.0.).
.000.V unﬁn- NmN 00.0.0 00.00.- 000 «:00 0.50...— 000 .:00 5.0—).

 

EA... 0:5>. 0A. 5.... 0.00:0.0 052...... 00000 00...: 8:00.01 00000 00... 5:00.:—

 

 

 

 

 

 

 

 

 

 

 

m:00.). 00.0:00 .0004 00%.0000045520 .5 30:20.02.

 

 

 

$5.500. 00000 3 5.00.056
M0... 3 6.030.). 0:0 3:... 0000.... 5:020 503.3. 5...... .00 0000.. ma... .5 .2557. 000.03.. :. 00:20.5: "0.. 0.5...

 

 

 

 

 

.000.v Nb.»- Nnn 00 .0... 0000. . m- 000 500 0:00... 000 «:00 5.0.).
.000.v >0...- umu 00.0.0 0000.00- 000 506 0.2.0.0 000 .:00 5.02
.000.v 00.00- «mm 00 5..u 00.00 _ - 00m «:00 2.00.0 00m «:00 5.0.).
.000.v and..- nmu 00 . 0.0 00. _ m . - 000 506 0:00.... 000 .:00 5.0.).
E A .0 0:5 > . 0n. 5.... 0505.0 0.0.5.... 000001 00...] .0:0:.0I 00000 00... 5:020

 

 

 

 

 

 

 

 

 

 

 

0:00.). 0.0.0.000 .0001. 000.009.000.500... .5 80:20:...

 

 

 

50.0.5.0. 00000 3 9:02:00.
00%... M0... 3 6.550.). 0:0 3:... 0.50.0. 5:00... 50.5.8. 5:... .00 0000”. 00... .5 .2557. 000.03. :. 00:05.0... "N0. 030...

192

Hypothesis 4. Package shape will not have an inﬂuence on the average amount of tag
reads per trial for RFID transponders on packages moving on a conveyor.
This hypothesis was rejected. Cans averaged more reads per trial than bottles or

tins. Bottles averaged more reads per trial than tins (Table 74).

Table 74: Difference Average Number of Reads Per Trial by Product Type

 

Differences of Product Least Squares Means

 

Product _Product Estimate Standard DF tValue Pr>|t|
Error

 

Bottle Can -15.1500 1.5302 348 -9.90 <.0001

 

Bottle Tin 56.8417 1.5302 348 37.15 <.0001

 

Can Tin 71.9917 1.5302 348 47.05 <.0001

 

 

 

 

 

 

 

 

 

In two-way and three-way interactions, product did not always act as the
dominant variable.

Cans and Bottles

In two-way interactions in which product was the only variable (speed was held
constant), cans averaged more tag reads per trial than bottles at 300 fpm. However at 600
fpm, cans and bottles averaged a statistically similar amount of reads per trial.

In two-way interactions in which product was the only variable (tag type was held
constant), when cans and bottles were equipped with a Gen 1 tag, they averaged a
statistically similar amount of readers per trial. However, when equipped with a Gen 2
tag, cans averaged more reads per trial than bottles.

In three-way interactions in which product was the only variable (tag type and

speed were held constant) cans always averaged more reads per trial than bottles, except

193

on one occasion. When cans with a Gen 1 tag were moving 600 fpm, they performed
equal to bottles with a Gen 1 tag moving 600 fpm.

Bottles and Tins

In two-way interactions involving bottles versus tins, bottles always averaged
more reads per trial than tins.

In three-way interactions bottles always averaged more tag reads than tins except
on two occasions. When product and speed were variables (tag type held constant) tins
with 3 Gen 2 tag moving 300 fpm averaged more tag reads per trial than bottles with a
Gen 2 tag moving 600 fpm. When product, tag type and speed were all variables, tins
with a Gen 2 tag moving 300 fpm averaged a statistically similar amount to bottles with a
Gen 1 tag moving 600 fpm.

Cans and Tins

In two-way interactions involving cans versus tins, cans always averaged more
reads per trial than tins.

In three-way interactions cans always averaged more tag reads than tins except on
two occasions. When product and speed were variables (tag type held constant) tins with
a Gen 2 tag moving 300 fpm averaged a statistically similar amount to cans with a Gen 2
tag moving 600 fpm. When product, tag type and speed were all variables, tins with a
Gen 2 tag moving 300 fpm averaged a statistically similar amount to cans with a Gen 1

tag moving 600 fpm.

194

Evaluation of Package Shape Effect Test, Metal Bottles, Metal Cans and Metal Tins

Two-Way Interactions

Product by Speed

In studying Table 35, which shows the difference in the average number of tag

reads per trial between product by speed, some overall conclusions can be formed.

When product was the only variable:

0

O

O

Cans averaged more tag reads per trial than bottles when moving at 300
fpm (Table 75, Row 1).

Cans did not have a statistically different average number of tag reads per
trial than bottles when moving at 600 fpm (Table 75, Row 2).

Cans always averaged more tag reads per trial than tins independent of
conveyor speed (Table 75, Row 3 and 4).

Bottles always averaged more tag reads per trial than tins, independent of
conveyor speed (Table 75, Row 5 and 6).

195

Table 75: Difference in Average Number of Tag Reads Per Trial Between Product

by Speed

 

Differences of Product*Speed Least Squares Means

 

 

 

 

 

 

 

 

Product Speed _Product _Speed Estimate Standard Error DF t Value Pr > |t|
Bottle 300 Can 300 -28.8500 2.1641 348 -13.33 <.0001
Bottle 600 Can 600 - l .4500 2.1641 348 —0.67 0.5033

Can 300 Tin 300 104.27 2.1641 348 48.18 <.0001
Can 600 Tin 600 39.7167 2.1641 348 18.35 <.0001
Bottle 300 Tin 300 75.4167 2.1641 348 34.85 <.0001
Bottle 600 Tin 600 38.2667 2.1641 348 17.68 <.0001

 

 

 

 

 

 

 

 

 

196

 

Product by Tag Type
In studying Table 37, which shows the difference in the average number of tag
reads per trial between product by tag type, some overall conclusions can be formed.

0 When the product was the only variable:

0 Cans were not statistically different from bottles in the average ntunber of
tag reads per trial when each had a Gen 1 tag attached to the case (Table
76, Row 1).

o Cans averaged more tag reads per trial than bottles when each had a Gen 2
tag attached to the case (Table 76, Row 2).

0 Both bottles and cans always averaged more tag reads per trial than tins,
independent of tag type (Table 76, Row 3-6).

197

Table 76: Difference in Average Number of Tag Reads Per Trial Between Product

 

 

 

 

 

 

 

 

by Tag Type
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error

Bottle Genl Can Genl -3.5333 2.1641 348 -1.63 0.1034
Bottle Gen2 Can Gen2 -26.7667 2.1641 348 -12.37 <.0001
Bottle Genl Tin Genl 66.4833 2.1641 348 30.72 <.0001
Bottle Gen2 Tin Gen2 47.2000 2.1641 348 21.81 <.0001
Can Genl Tin Genl 70.0167 2.1641 348 32.35 <.0001
Can Gen2 Tin Gen2 73.9667 2.1641 348 34.18 <.0001

 

 

 

 

 

 

 

 

 

 

 

Three-Way Interaction
Bottle-Can Interactions

In studying Table 44, which shows the difference in the average number of tag

reads per trial between two products (bottles and cans) by tag type by speed, some overall

conclusions can be formed.

When the product was the only variable, cans always averaged more tag reads
per trial than bottles, except when each product was equipped with a Gen 1 tag
moving at 600 fpm, in which the two products were statistically similar in there
average amount of tag reads per trial (Table 77, Row 1).

When product and tag were the two variables, there was no statistical difference
in the average number of tag reads per trial between bottles and cans, except with
the case of bottles with a Gen 1 tag moving at 300 fpm compared to the case of
cans with a Gen 2 moving at 300 fpm, in which case cans with a Gen 2 tag
averaged more tag reads per trial than bottles with a Gen 1 tag (Table 77, Row 2).

198

 

359V em...” wvm £59m manhme- can «:00 .50 can E00 0.38—

 

gemé 3.: wvm £59m gcmﬁ cce :30 .30 cce :30 2:3—

 

EAE 2:5: m: Leta FEE—Sm San—3mm 33ml we... 333.51 team man. 825...—

 

 

 

 

 

 

 

 

 

 

 

»...32 825$ .33 mahsuocamtuavem he 89:20.:5

 

 

 

933:5 2558.5 2: gel 33 .269:— 0355 .6925
,3 5:22.00 ash. .3 9:; 2:. 8:35 835.:— =3>$om 3:. 3m 33% wan. we non—=52 09:05.. E 3:955: Kb 035,—.

199

Bottle-Tin Interactions

In studying Table 45, which shows the difference in the average number of tag

reads per trial between two products (bottles and tins) by tag type by speed, some overall

conclusions can be formed.

When the product was the only variable (tag type and speed were the same)
bottles always averaged more tag reads per trial than tins.
When the product and tag were the two variables, product was the dominant
variable, with bottles always averaging more tag reads per trial than tins,
regardless of tag type on the case.
When the product and speed were the two variables, product was the dominant
variable in 3 of the 4 comparisons, with the bottles always averaging more reads
per trial than tins.
o The exception was when the case of bottles had a Gen 2 tag and was
moving at 600 fpm compared to the case of tins with a Gen 2 tag moving
300 fpm, with the latter averaging 7.4 more tag reads per trial (Table 78,
Row 2). In this case, speed (300 fpm) was more dominant than product
(bottles)
When product, tag, and speed were all variables, product was the dominant
variable in 3 of the 4 comparisons, with the bottles always averaging more reads
per trial than tins.
o The exception was when the case of bottles had a Gen 1 tag and was
moving at 600 [pm compared to the case of tins with a Gen 2 tag moving
300 fpm, there was not a signiﬁcant statistical difference in the average
number of tag reads (Table 78, Row 2). In this case, the dominant tag
(Gen 2) and dominant speed (300 fpm) were more dominant than the
dominant product type (bottles).

200

 

 

 

 

 

 

 

 

 

 

 

 

 

83.: mod- wvn 336 $6 _ d- can ~50 =5. 80 :30 2:3.
336 «EN- wvn 336 259.»- can N50 :2. 25 N50 0.33—
E A .5 03.5 a ”a Stu Eat—_Sm 2aEEM tooaml 3...! 335.51 coon—w nah. .2.—.9.."—

 

 

mcaoE 8,—2.5 3qu wahigoamiguem .3 mun—53:5

 

 

,3 accenonow man. .3 3:5. can 853: .2695 season _atH hem meanw— wah no Loan—=2 03.53,. E 09:29th "an 035—.

933:5 2358.5 2: 82 as?» 833.:— 9355 60on

201

Can-Tin Interactions
In studying Table 46, which shows the difference in the average number of tag
reads per trial between two products (cans and tins) by tag type by speed, some overall

conclusions can be formed.

0 When the product was the only variable (tag type and speed were the same) cans
always averaged more tag reads per trial than tins.

- When the product and tag were the two variables, product was the dominant
variable, with cans always averaging more tag reads per trial than tins, regardless
of tag type on the case.

0 When the product and speed were the two variables, product was the dominant
variable in 3 of the 4 comparisons, with the cans always averaging more reads per
trial than tins.

o The exception was when the case of cans had a Gen 2 tag and was moving
at 600 fpm compared to the case of tins with a Gen 2 tag moving 300 fpm,
there was not a signiﬁcant statistical difference in the average number of
tag reads per trial (Table 79, Row 1). In this case, speed (300 fpm) was
more dominant than product (cans).

c When product, tag, and speed were all variables, product was the dominant
variable in 3 of the 4 comparisons, with the cans always averaging more reads per
trial than tins.

o The exception was when the case of cans had a Gen 1 tag and was moving
at 600 fpm compared to the case of tins with a Gen 2 tag moving 300 fpm,
there was not a signiﬁcant statistical difference in the average number of
tag reads per trial (Table 79, Row 2). In this case, the dominant tag (Gen
2) and dominant speed (300 fpm) were more dominant than the dominant
product type (cans).

202

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cmmmé 36. men 38% 53.? can ~50 Eh. cce :50 .50
ambmé ems- mvn waved 83.7 8m «50 En. Se «:90 55
E A .:— o=_a> « “.5 Sta FEE—Sm Suﬁzmm 63am! wahl 3.59:1 team an... 833.5

 

 

2:32 833m 3.31— wahauouamtoauoi .3 82.93.59

 

 

,3 3:22.00 wag. .3 A33. 35 £38 835...— 5950m— Eth hum memo: war—t he .33: Z 09295. :_ 3:95.59 "an 2:3.

033:5 2555.5 2: 32 33 335.:— 0355 60on

203

Hypothesis 5. Tag type will not have an inﬂuence on the average amount of tag reads
per trial for packages moving on a conveyor.

This hypothesis was neither rejected nor failed to be rejected. In the product
effect test and the package shape effect test, tag type had a signiﬁcant effect on the
average number of tag reads per trial, therefore rejecting the hypothesis. In the package
effect test, tag type did not have a signiﬁcant effect on the average number of tag reads
per trial, therefore failing to reject the hypothesis (Table 80).

Table 80: Average Difference Between Number of Tag Reads Per Trail for Gen 1

 

 

 

 

 

Tag versus Gen 2 Tag
Average Difference Between Number of Tag Reads Per Trial, Tag Least Square Means
Effect Test Estimate Standard DF t Value Pr > It]
Error
MD Between Gen 1 vs Gen 2 -8.7833 2.4358 232 -3.61 0.0004

(Product Effect)

 

MD between Gen 1 vs Gen 2 2.0083 2.0078 232 1.00 .3182
(Package Effect)

 

MD between Gen lvs Gen 2 -18.0056 1.2594 348 -14.41 <.0001
(Shape Effect)

 

 

 

 

 

 

 

 

In both two-way and three-way interactions in which the products (all 7 cases)
were moving speed of 300 fpm , the Gen 2 tag either averaged a greater or equal
statistical amount of tag reads per trial compared to the Gen 1 tag.

In both two-way and three-way interactions in which the products (6 out of 7
cases) were moving speed of 600 fpm , the Gen 2 tag either averaged a lesser or equal
statistical amount of tag reads per trial compared to the Gen 1 tag (except for the case of
Tins, Gen 2 averaged more reads per trial statistically).

In two-way interactions involving ketchup, with tag type being the only variable,

Gen 2 averaged more reads per trial than Gen 1.

204

In all two-way interactions involving motor oil, with the tag type being the only
variable, Gen 1 averaged more reads per trial than Gen 2.

In all two-way interactions involving MSWFC, with the tag type being the only
variable, Gen 1 and Gen 2 averaged an equal amount of reads per trial.

In all two-way interactions involving plastic tubs, with the tag type being the only
variable, Gen 1 and Gen 2 averaged an equal amount of reads per trial.

In all two-way interactions involving bottles, with the tag type being the only
variable, Gen 1 and Gen 2 averaged an equal amount of reads per trial.

In all two-way interactions involving cans, with the tag type being the only
variable, Gen 2 averaged more reads per trial than Gen 1.

In all two-way interactions involving tins, with the tag type being the only

variable, Gen 2 averaged more reads per trial than Gen 1.

The Effect of Tag Type in the Product Effect Test
Two Way Interactions
Product by Tag Type, Tag Type Only Variable
In studying Table 8, which shows the difference in the average number of tag

reads per trial between product by tag type, some overall conclusions can be formed.

0 When tag type was the only variable, tag type had a signiﬁcant effect on the
average number of tag reads per trial on the case of ketchup with the Gen 2 tag
averaging 12.9 more reads per trial than the Gen 1 tag (Table 81, Row 1).

o In contrast, tag type did not have a signiﬁcant effect on the average number of tag
reads per trial for the case of motor oil, making the Gen 1 tag and Gen 2 tag
perform as equals (1‘ able 81, Row 2).

205

Table 81: Product-Tag Type Two-Way Interaction, Difference in Number of Tag
Reads Caused by Tag Type

 

Differences of Product*Tag Least Squares Means

 

Product Tag _Product _Tag Estimate Standard DF tValue Pr>|t|
Error

 

Ketchup Gen Ketchup Gen -12.8833 3.4447 232 -3.74 0.0002

 

 

l 2
Oil Gen Oil Gen 4.6833 3.4447 232 -1.36 0.1753
1 2

 

 

 

 

 

 

 

 

 

 

Tag Type by Speed, Tag Type Only Variable
In studying Table 10, which shows the difference in the average number of tag
reads per trial between tag generation by speed, some overall conclusions can be formed.

0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, in which the Gen 2 tag averaged 27.8 more reads per trial than the Gen 1
tag (Table 82, Row 1).

0 However, at 600 fpm the Gen 1 tag performed better than the Gen 2 tag by an
average of 10.22 more reads per trail (Table 82, Row 2).

Table 82: Product-Tag Type Two-Way Interaction, Difference in Number of Tag
Reads Caused by Tag Type

 

Differences of Speed*Tag Least Squares Means

 

Tag Speed _Tag _Speed Estimate Standard DF tValue Pr>|t|

 

 

Error
Genl 300 Gen2 300 -27.7833 3.4447 232 -8.07 <.0001
Genl 600 Gen2 600 10.2167 3.4447 232 2.97 0.0033

 

 

 

 

 

 

 

 

 

 

 

206

Three-Way Interactions
Ketchup-Ketchup, Tag Type Only Variable
In studying Table 12, which shows the difference in the average number of tag
reads per trial between product (ketchup) by tag type by speed, some overall conclusions
can be formed.

0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, as the Gen 2 tag averaged 31.9 more reads per trial than the Gen 1 tag
(Table 83, Row 1).

0 However, at 600 fpm there was no statistical difference between the Gen 1 and
Gen 2 tag (Table 83, Row 2).

Motor Oil-Motor Oil, Tag Type Only Variable
In studying Table 14, which shows the difference in the average number of tag reads
per trial between product (motor oil) by tag generation by speed, some overall
conclusions can be formed.
0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, in which the Gen 2 tag averaged 23.7 more tag reads per trial than Gen 1
tag (T able 84, Row 1).

0 However, at 600 fpm the Gen 1 tag averaged 14.3 more tag reads per trial than the
Gen 2 tag (Table 84, Row 2).

207

 

omocd voﬁ an" 23.? mama's..— 80 N50 =O 25 :50 :0

 

5.5V 3.... NMN m _ $6 coardm- can ~50 :0 can :30 :0

 

335.51 nova—m Man. .259..—

 

EAE 2:5». ma St”.— ESEEm 82.5mm coon—ml nah

 

 

 

 

 

 

 

 

 

 

2:32 8:55 .23..— wahsuoonmtgcoi .5 82.2959

 

 

 

25,—. wan. .3 com—EU 2:3”— wah no hog—=52 E 35.—0&5 52.228:— hakréouah. =0 .882-=O .832 new 033—.

 

med mm; mmm .23.? ooo—.0 95 ~50 9:3ro— cco :30 9.5.83—

 

_ooo.v «We- Nmm 22:. $3.3. can ~50 9533. can :50 9533—

 

EALL 02.5. km Stu Ewe—_Sm 8.35%. E251 nah! 335.51 6925 ME. 3355

 

 

 

 

 

 

 

 

 

 

 

mES—z 8:35 .234 wahauooawtusuci .8 3233::—

 

 

 

oaza. ma. .3 com—EU 38.: Ma. no .538: Z a. 35.—ohm: 63.228:— zaBéPEH a:a3«v—-m=:80v— "mm 035—.

208

The Effect of Tag Type in the Package Effect Test
Two Way Interactions
Product by Tag Type, Tag Type Only Variable
In studying Table 22, which shows the difference in the average number of tag
reads per trial between product by tag type, some overall conclusion can be formed.
0 When tag type was the only variable, tag type did not have a signiﬁcant effect on
the average number of reads for either the case of plastic tubs or MSWFC,

showing the Gen 1 tag and Gen 2 tag performed statistically similar (Table 85).

Table 85: Product-Tag Type Two-Way Interaction, Difference in Number of Tag
Reads Caused by Tag Type

 

Differences of Product*Tag Least Squares Means

 

Product Tag _Product _Tag Estimate Standard DF tValue Pr>|t|
Error

 

Metal Genl Metal Gen2 3.2333 2.8395 232 1.14 0.2560

 

Plastic Genl Plastic Gen2 0.7833 2.8395 232 0.28 0.7829

 

 

 

 

 

 

 

 

 

 

 

Tag Type by Speed, Tag Type Only Variable
In studying Table 24, which shows the difference in the average number of tag
reads per trial between tag generation by speed, some overall conclusion can be formed.
0 When tag type was the only variable, the tag type did not have a signiﬁcant effect

on the average number of reads for either speed of 300 fpm or 600 fpm, showing
the Gen 1 tag and the Gen 2 tag performed statistically similar (Table 86).

209

Table 86: Product-Speed Two-Way Interaction, Difference in Number of Tag Reads
Caused by Tag Type

 

Differences of Speed*Tag Least Squares Means

 

Tag Speed _Tag _Speed Estimate Standard DF tValue Pr>|t|
Error

 

Genl 300 Gen2 300 -O.4500 2.8395 232 -0. I 6 0.8742

 

Genl 600 Gen2 600 4.4667 2.8395 232 1.57 0.1171

 

 

 

 

 

 

 

 

 

 

 

Three-Way Interactions

MSWFC-MSWFC, Tag Type Only Variable

In studying Table 26, which shows the difference in the average number of tag
reads per trial between product (MSWFC) by tag generation by speed, some overall
conclusions can be formed.

0 When tag type was the only variable, the tag’s generation did not matter as there
was no statistical difference in the average number of tag reads per trial between

the Gen 1 tag and the Gen 2 tag at either 300 fpm or 600 fpm (Table 87).

Plastic Tubs-Plastic Tubs, Tag Type Only Variable

In studying Table 28, which shows the difference in the average number of tag
reads per trial between product (plastic tubs) by tag generation by speed, some overall
conclusions can be formed.

0 When tag type was the only variable, the tag’s generation did not matter as there

was no statistical difference in the average number of tag reads per trial between
the Gen 1 tag and the Gen 2 tag at either 300 fpm or 600 fpm (Table 88).

210

 

 

 

 

 

 

 

 

 

 

 

 

 

«.33. a... um" em _ 3. Sea...” 25 ~50 3.3: So .50 9:3:
3mm... 86. «mm on _ 3. 83$ 8n ~50 9:3:— can :30 3.3.“—
E A .:— o=_u> a he .:Em Eaunﬁm oaaﬁzmm 33ml 9:... 635.51 .5on us... 823....

 

 

2.32 manna—um :3:— wahguonmazseem .B moo—5.5::—

 

 

omhb man. .3 @0230 mecca wan. me 3:552 E 35.—ob:— .:o:oa._3=~ $3-825. 3:8 urns—mé55h. owns—.— "an 03.2.

 

 

 

 

 

 

 

 

 

 

 

 

 

w: No To... mmm om ﬂ o6 Scoé :3 «EU .802 cce :50 .802
30:. Rd mmm cm _ ed 800. can ~50 .502 can .:ow .592
E A .5 o=_a> . he .Stm “:2.ng Saﬁzmm coemml unhl .2.—Sin .5on wan. 335.5

 

 

2502 menu—5m unwou— uwhi—anmtuzuoi .3 39:05.:5

 

 

09C. uah .3 63.30 230m ugh. no Lona-.2 E 35.—eta £33.28:— >a>>noohah Umgmzuohamz "ha 033—.

211

The Effect of Tag Type in the Package Shape Effect Test
Two Way Interactions
Product by Tag Type, Tag Type Only Variable
In studying Table 37, which shows the difference in the average number of tag

reads per trial between product by tag type, some overall conclusions can be formed.

0 When tag type was the only variable, Gen 2 averaged more tag reads per trial
than Gen 1 for cans and tins (Table 89, Row 1 and 2).

o The exception arose when the Gen 1 tag on the case of bottles was
compared to the Gen 2 tag on the case of bottles, in which their was no
statistical difference between the average number of tag reads with the
Gen 1 tag versus the Gen 2 tag (Table 89, Row 3).

0 When tag and product were both variables, product was always the dominant
variable (Cans>Bottles>Tins) with one exception.

0 Bottles with a Gen 2 tag and cans with a Gen 1 tag averaged a statistically
similar amount of tag reads per trial. This shows a signiﬁcant advantage
for the Gen 2 tag in this situation, with the tag being able to become the
dominant variable in the tag by product interaction (Table 90).

212

Table 89: Product-Tag Type Two-Way Interaction, Difference in Number of Tag

 

 

 

 

 

 

 

 

 

 

 

 

 

Reads Caused by Tag Type
Differences of Product*Tag Least Squares Means
Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error
Can Genl Can Gen2 -27.0667 2.!64l 348 42.51 <.0001
Tin Genl Tin Gen2 -23.I I67 2le 348 -IO.68 <.0001
Bottle Gen l Bottle Gen2 -3 .8333 2. I 641 348 -I .77 0.0774

 

 

 

Table 90: Two-Way Interaction Between Case of Bottles With Gen 2 Tag versus
Case of Cans With Gen 1 Tag

 

Differences of Product*Tag Least Squares Means

 

 

 

 

 

 

 

 

 

 

Product Tag _Product _Tag Estimate Standard DF t Value Pr > |t|
Error
Bottle Gen2 Can Genl 0.3000 2.164] 348 0.14 0.8898

 

 

 

Tag Type by Speed, Tag Type Only Variable
In studying Table 39, which shows the difference in the average number of tag
reads per trial between tag type by speed, some overall conclusions can be formed.

0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, in which Gen 2 performed better than Gen 1 (Table 91, Row 1).

0 However, at 600 fpm the tag ‘8 generation didn’t matter as there was no statistical
difference between the Gen 1 tag and the Gen 2 tag at 600 fpm (Table 91, Row
2).

Table 91: Tag Type-Speed Two-Way Interaction, Difference in Number of Tag
Reads Caused by Tag Type

 

Differences of Speed*Tag Least Squares Means

 

 

 

 

 

 

 

Tag Speed _Tag _Speed Estimate Standard Error DF t Value Pr > |t|
Genl 300 Gen2 300 -36.9000 1.7670 348 -20.88 <.0001
Genl 600 Gen2 600 0.8889 1.7670 348 0.50 0.6l52

 

 

 

 

 

 

 

213

Three-Way Interactions
Bottle-Bottle , Tag Type Only Variable
In studying Table 41, which shows the difference in the average number of tag
reads per trial between product (bottles) by tag generation by speed, some overall
conclusion can be formed

0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, in which Gen 2 performed better than Gen 1 (Table 92, Row 1).

0 However, at 600 fpm the tag’s generation didn’t matter as there was no statistical
difference between the Gen 1 tag and the Gen 2 tag at 600 fpm (Table 92, Row
2).

Can-Can Interaction, Tag Type Only Variable
In studying Table 42 which shows the difference in the average number of tag
reads per trial between product (cans) by tag generation by speed, some overall
conclusions can be formed.
0 When tag type was the only variable, the tag’s Generation made a difference at
300 fpm, in which Gen 2 performed better than Gen 1 (T able 93, Row 1).
0 However, at 600 fpm the tag’s Generation didn’t matter as there was no statistical
difference between the Gen 1 tag and the Gen 2 tag at 600 fpm (Table 93, Row
2).
Tin-Tin Interaction, Tag Type Only Variable
In studying Table 43, which shows the difference in the average number of tag
reads per trial between product (tins) by tag generation by speed, some overall
conclusions can be formed.
0 When tag type was the only variable, the tag’s generation made a difference at
300 fpm, in which Gen 2 performed better than Gen 1 (T able 94, Row 1).
0 However, at 600 fpm the tag’s generation didn’t matter as there was no statistical

difference between the Gen 1 tag and the Gen 2 tag at 600 fpm (Table 94, Row
2).

214

 

09C. man. .3 e830 unwou— wah he Loan—=2 E 35.—aha 6.5223:— magbonah Ewan—E. 6a 935.

0

 

 

:2... 8;- 8n Seed :82. Se «Eu .5 Se .50 5
88¢ 8.3- a...“ 335 88."... Sn «Eu 5 Sn :8 5
E A E 3....» . .5 Sta 2.2.3 3.53m 88m: use: .2.—5.5- 88m use 8.68..

 

 

 

 

 

 

 

 

 

 

 

 

2:32 ﬁrearm «was wahsgwtuavohm _o $23.65:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«2...; =3- wvn 336 been..- so «new .30 :3 Few 55
327 pm.» T awn 33d Sewﬁm- 8m «:00 55 can :30 55
E A .:— o=_a> . ma Stu 9:655 Sag—rum Beam] nah; .2.—veil 30mm nah. 835.5

 

2:82 moans—cm «man‘— wahagonmtogoi he 82.95::—

 

 

 

mh. web .3 tom—=5 33m wma. we Luna—=2 E 35.-9E9 £33.28:— haaéosa. 5.0-50 "ma 035.

 

 

 

 

 

 

 

 

 

 

 

 

 

hm _ o.o om...“ we». wooed mmmms .50 N50 28°: :8 :50 2:3—
Soom. 3:... mam mecca 080.3- can ~50 2:3— ccm :56 2.25
E A .5 33> . "a Stu Emu—ii 25:me .3251 web! 335E! 3on new. «us—5...—

 

 

mESE mocazcm «may; wakruuomwtusuohm “a 82.28::

 

 

09¢. mam. .3 eon—EU £33m nah. we Loan—=2 E 85.5%: 533.28:— .3.$-3.E.H cacao—tom "3 29¢.

215

Thoughts and Conclusions

As Dr. Robb Clarke, Associate Professor at Michigan State University, referred to
in his article in RFID Product News, “RFID does not work well in all situations and,
realistically, only works really well in very controlled environments,” due to what Clarke
calls, “some inherent complications given the choice of ultra-high frequency (UHF) as a
global frequency”.[59]

It is apparent that the variables studied in this research (conveyor speed, product
type, package type, and package shape) participated in creating complications, upon
determining that each variable had an effect on the average number of RFID tag reads per
trial. Additionally, tag type, when placed on the case of ketchup, motor oil, metal tins,
metal cans, and metal bottles had an effect on the average number of RF ID tag reads per
trial, but did not effect the average number of tag reads per trial while on the case of
potato chips in plastic tubs, and the case of potato chips in MSWFC. Additionally, the
documentation of the two-way and three-way interactions that occur between
product/package, conveyor speed, and tag type, provide resounding evidence that there is
no one ﬁx-all solution for optimal RFID tag performance.

In concentrating on the effect of conveyor speed, when comparing 300 fpm to 600
fpm across all 3 tests, there was an average of almost two times the amount of tag reads
per trial at 300 fpm than there was at 600 fpm (91.9 more tag reads per trial at 300 fpm
than 600 fpm)

Throughout the entire testing procedure, 6 out of 7 seven products tested were
detected in every trial, at each conveyor speed, equipped with either generation of tag.

The only product that occasionally went undetected through the portal was the case of

216

metal tins. It is believed that this was due to the fact that there was not one responsive
tag location, in either the horizontal or vertical tag orientation, on the case of tins. To
possibly improve the readability of the case of tins, previous testing has shown for
difﬁcult to read products, the use of a corrugated spacer greatly improved read rates.

It is important to note that in three-way interactions involving all product types,
the Gen 2 tag performed greater than or equal to the Gen 1 tag at 300 fpm. Additionally,
in three-way interactions involving all product types, the Gen 1 tag performed greater
than or equal to the Gen 2 tag when moving at 600 fpm (except for three way interactions
involving the case of tins moving at 600 fpm, in which the Gen 2 tag performed better
than the Gen 1 tag). The implications of this are clear; companies mandating the use of
Gen 2 tags may limit the beneﬁts of the technology (better operation around liquids and
metals), when operating conveyors at 600 fpm.

It is also important to note that tag dominance varied by product. The Gen 2 tag
outperformed the Gen 1 tag on the case of ketchup, metal cans and metal tins. The Gen 1
tag outperformed the Gen 2 tag on the case of motor oil. The Gen 1 tag performed equal
to the Gen 2 tag on the case of metalized spiral wound ﬁberboard containers, plastic tubs,
and metal bottles. In summary, the Gen 2 tag tested provided an advantage, or at least
performed equally to Gen 1, for those products or packages which contained high
contents of metal or water (metal cans, tins, bottles, and ketchup).

With the knowledge that conveyor speeds have a drastic effect on RF ID tag
readability, it is crucial that suppliers meeting RFID mandates communicate with their
retailers regarding their distribution center and conveyor speed operations. Armed with

this information, suppliers can improve product detection at retailer RF ID checkpoints,

217

ensuring payment for their product. With retailers having knowledge of the tagged
product’s location, they are less likely to incur a situation in which a product is out of
stock, and through this, retailers are able to increase product sales.

Recommendations for future research involve determining if there is an effect on
the average number of reads per trial by both number and antennae location. Previous
research performed in this area studied the effect of the number of, and location of,
antennae, which was performed using a pallet-load of goods (as compared to the case-
loads tested in the conveyor test), placed on a stretch-wrapping machine, all of which was
surrounded by an RFID antennae portal. The test, which was performed by Crawforth of
Michigan State University, showed that number and location of antennae did have a
signiﬁcant effect on the percent readability of transpondersml It would be of beneﬁt to
determine if a similar effect occurs with antennae used on conveyor lines, to eliminate

unnecessary costs while failing to limit tag readability.

218

IO.

11.

12.

13.

14.

REFERENCES
RFIDJoumal, "Glossary of Terms", 2006,

EPCglobal, "Frequently Asked Questions",
http://www.epcglobalinc.org/about/faqshtml, 2003, (1/4/2005)

 

MIT, "Technology Guide",
http:// archiveepc globalinc.org,r’ncw_mediw’broch urcs/Technolo gy_G uidc. pd f,
2002, (12/12/2004)

 

EPCglobal, "EPCglobal Ratiﬁes Royalty-Free UHF Generation 2 Standard",
http://www.epcglobalinc.org/news/pr_detail_epcinc.cfm?release_id=l 83, 2005,
(1/4/2005)

Collins, J ., "What's Next for Gen 2?"
haw/216.121 .131 .l29/article/ar’ticleprint/l 30]/-l/l/, 2004 A, (1/5/2005)

 

Matrics, "Understanding The Wal-Mart Initiative",
http://rfidblucstarinc.com/rcsources/Understanding_The Wal-
Martilnitiativepdf#search='1\»latrics9/02C‘%2()Undcrstanding‘Vo20WalIV1art9/0201nit
iative’, 2004, (4/1/06)

 

 

SmartCode, http://smartcodccorpcom/newsroom/ l 6-12-04.a§p, 2004, (1/5/2005)

 

Ferguson, R.B., New Wave for RFID; EPC'S UHF Gen 2 works in dense reader
envioronments., in eWeek. 2005. p. 30.

Diorio, C., "Gen 2 - The Tough Questions",
http://www.rﬁdi0urnal.com/article/articleview/l 783/1/82/, 2005, (6/9/2006)

 

Collins, J ., "Hiccups Expected for Gen 2",
http://216.l2l.l 3 l.129/article/articleprint/l 289/-1/l/, 2004 B, (1/4/2005)

BusinessWire, "Intermec Applauds Ratiﬁcation of EPCglobal UHF Generation 2
Air Interface Protocol", http://tmcnet.com/usubmit/2004/Dec/1 102596.htm, 2004,
(1/4/2005)

 

Kleist, R., RFID Labeling. 1 ed. 2004, Irvine: Printronix. 211.
SymbolTech, Class 0 Tag Information, J. Falls, Editor. 2006.
AlienTechnology, "Alien Squiggle Family of EPC RFID Tags",

http://www.alientechnology.com/products/documems/alien squiggle‘familypdf,
2006, (06/08/06)

 

219

 

15.

16.

17.

18.

19.

20.

2].

22.

23.

24.

25.

26.

27.

28.

29.

RFidGazette, "RFID regulations by FCC",
http://www.rﬁdgazette.0rg/2005/ 12/rﬁd_regulation.html, 2005, (5/22/06)

Quirk Jr., R., "Don't get Behind the Regulatory Eight Ball",
http://wwwrﬁd'muma].com/article/articleview/l484/1/133/, 2005, (5/22/06)

Wikipedia, "Wal-Mart", http://en.wikipedia.0rg/wiki/Wal-Mart, 2006, (2/18/06)

 

IDTechEx, "RFID progress at Wal-Mart",
http://www.idtechex.com/products/en/articles/OOOOO161asp, 2005, (3/31/06)

 

Roberti, M., "Wal-Mart Begins RFID Rollout",
http://www.rﬁdjoumal.com/article/articIeview/926/1/1, 2004, (3/25/2006)

Johnson, J ., "What RFID did for Wal-Mart",
http://www.dcvelocity.com/articles/20060101/rﬁdwatch.cfm, 1/7/06, (2/20/06)
O'Connor, M., "Wal-Mart Speciﬁes Gen 1 Sunset, Forklift Pilot",
http://wwwrﬁdioumal.com/article/articleview/2271/1/1/, 2006, (4/15/06)

 

 

Crawforth, T., The Eﬂect 0f Antennae Conﬁguration, Product and Tag Type on
Readability of Passive UHF RF ID Transponders, in The School of Packaging.
2005, Michigan State University: East Lansing. p. 60.

Intermec, "Nut grower Paramount Farms speeds load processing time by 60
percent with Intermec RFID",
http://\L'ah00.knowledgestorm.com/shared/write/collateral/CST/12482i] 7163 613

 

27_Param0untFarmsics_wcb.pdf, 2004, (1/20/2005)

Williams, D., "The Strategic Implications of Wal-Mart‘s Mandate",
http://www.directionsmagcom/article.php?article_id=629&trv=1 , 2004,
(3/18/2006)

Defense, D.o., United States Department of Defense Suppliers' Passive RF ID
Information Guide (Version 8.0). 2005. p. 27.

MSNBC, "Thompson resigns with grim warning",
http://www.msnbc.msn.com/id/6644980, 2004, (4/4/06)

 

DCVelocity, "FDA turns its gaze on the food chain",
http://www.dcve10city.com/articlesﬁan2005/news.cfm, 2005, (1/24/2005)

Wessel, R., "Orion Pharma Tests Item-Level Tagging",
http://wwwrﬁcjoumal.com/article/articleview/Z179/1/1/, 2006, (3/19/2006)

Wasserman, E., "Purdue Pharma to Run Pedigree Pilot",
http://www.rﬁdj0umal.com/article/articleview/l 626/1/1, 2005, (3/19/2006)

220

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

O'Connor, M., "Pﬁzer Using RFID to Fight Fake Viagra",
http://wwwrﬁdiournal.com/article/articleview/2075/1/1/, 2006, (3/19/2006)

Roberti, M., "RFID in Europe: A Special Case",
mtp://www.rﬁdi0urnal.com/article/articleview/1688/1/2/, 2005, (05/05/2006)

Collins, J ., "Tesco CTO Describes Europe's Hurdles",
http://v~ww.rﬁdj0urnal.com/article/articleview/l 502’1/20/, 2005, (4/23/06)

 

Collins, J ., "Tesco Revises RF ID Plans",
http://www.rﬁdjoumal.com/articIe/artic1cview/2243/1/9/, 2006, (4/23/06)

Collins, J ., "Marks and Spencer to Extend Trial to 53 Stores",
http://\x'\~'x~'.rfidj0umal.com/article/articleview/14 1 2/1/20/, 2005, (04/24/06)

Collins, J ., "Metro Group Reaps Gains From RFID",
http://\~'ww.rﬁdjournal.com/article/articleview/l 355/1/20/, 2005, (5/1/2006)

 

Roberti, M., "What Metro Knows About RFID",
htg)://www.rfidjoumal.com/article/articleview/2200/1/2/, 2006, (5/1/2006)

O'Connor, M., "NEC to Sell Gen 2 Readers in Japan",
hup://www.rﬁdioumal.com/article/articleview/2287/1/1/, 2006, (5/1/2006)

Konomi, S.i., "RFID in Japan: RFID Jeans in Future Department Store",
http://ubiks.net/local/blog/jmt/archives3/004744.html, 2006, (5/10/2006)

 

EPCglobal, "EPC global Korea Provides Support to the Flourishing RFID
Market",
http://mvw.epcglobalinc.org/thcsourec/2005i04/hlml/mcmber_showcase.html,
2005,

 

O'Connor, M., "Sun to Build Korean RF ID Test Center",
http://www.rﬁdj0umal.com/article/articleview/l473/1/26/, 2005, (05/10/06)

RFIDjoumal, "RFID News Roundup- Singapore Expands RFID UHF Spectrum",
http://wwwrﬁdjoumal.com/article/articleview/2241/2/1/, 2006,

ChinaRFIDNews, "RFID Standards Battle in China",
http://www.chinarﬁdnews.c0m/index.php/2006/04/l 7/chinese-rﬁd-dcvekmi11g;
with-government-guidance/, 2006, (5/10/2006)

 

 

ChinaRFIDNews, "Chinese RFID: Developing With Government Guidance",
http://www.chinarﬁdncws.c0m/index.php/2006/04/17/chinese-rﬁd-developin};
with-gm’ernmcnt-Jquidancc/, 2006, (5/10/2006)

 

 

221

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

ChinaRFIDNews, "China Retailer Pilots RFID Program",
http://www.chinarﬁdnews.c0m/indcx.php/2006/04/28/chinesc-rctailer-pilots-rﬁd-
program/, 2006, (5/10/2006)

 

OnlineDictionary, "Conveyor",
htLpz/Mictionary.reference.com/search‘?q=Conveyor, 2006, (3/23/06)

 

Apple, J ., Material Handling Systems Design. 1972, John Wiley & Sons: New
York. p. 656.

Roberti, M., "Wal-Mart Supplier Discuss RFID",
http://wwwrﬁdiournal.com/article/articleview/956/1/1, 2004, (3/24/2006)

 

RFIDAlIianceLab, "How the Lab Tests Were Conducted",
http://www.rﬁdjournal.com/article/articleview/1256, 2005, (6/9/2006)

Narayanan, R.a., Performance benchmarks for passive UHF RF ID tags. 2005,
The University of Kansas. p. 100.

Yanes, L., Investigation of passive transponder for radio frequency identiﬁcation.
2000, University of Louisville. p. 95.

Fletcher, R., Low-cost electromagnetic tagging: Design and implementation.
2002, Massachusetts Institute of Technology.

Jonson, M., Determination of Class 0 radio ﬁequency identiﬁcation tag failure
modes. 2005, Michigan State University. p. 87.

Nanj undaiah, M., Enhanced class 0 RFID anticollision protocol. 2005, Wayne
State University.

Gaukler, G., RFID in supply chain management. 2005, Stanford University. p. 96.

Ryan, M., A model for the implementation of a radio frequency identification
system into a warehouse environment. 2002, Michigan State University. p. 117.

De, P., An RFID based ubiquitous ﬁamework for mobile object tracking. 2004,
The University of Texas at Arlington. p. 70.

Onderko, J ., Radio ﬁequency identification transponder peformance on
refrigerated and ﬁozen beef loin muscle packages. 2004, Michigan State
University. p. 44.

Tazelaar, J., The eﬂect of tag orientation and package content on the readability
of radio frequency identiﬁcation (RF ID) transponders. 2004, Michigan State
University. p. 166.

222

59. Clarke, Robert. "Assessing Readability Problems with RFID Systems",
http://www.rﬁdproductnews.c0m/issues/2005.01/fcature/readability.p@, 2005,
6/9/2006

 

223

rlTimmlﬁlﬂlllﬁﬂmsﬂjﬁl181W