<%BANNER%>

Improving the Readability of RFID Tagged Pharmaceutical Products and Food Supplements Inside a Modified Sealed Tote

Permanent Link: http://ufdc.ufl.edu/UFE0021827/00001

Material Information

Title: Improving the Readability of RFID Tagged Pharmaceutical Products and Food Supplements Inside a Modified Sealed Tote
Physical Description: 1 online resource (95 p.)
Language: english
Creator: Dagdelen Uysal, Dilek
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: distribution, hf, metal, pharmaceutical, rfid, security, tote, uhf
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: One of the biggest advantages of radio frequency identification (RFID) technology is that it can be used to augment the security of the supply chain. A straightforward application area for which security is of utmost importance is the distribution and marketing of pharmaceutical products and food supplements. This research states that an RFID technology can be used to develop an RFID system that can reliably read all the tagged products inside a tote while trying to provide security against unauthorized readers, during distribution of these products. The challenge in using RFID to automatically, without human intervention, identify the tote, its contents, and the RFID enabled security features are due to the problems related to tag readability near and through different materials. The solutions addressing these problems are developed for sealed totes by evaluating of proper system set ups and tote designs. The products and the purchase orders (POs) in the pharmaceutical supply chain were analyzed to provide a repeatable and comparable evaluation. Each product were tagged and tested on a conveyor belt, on a turn table and inside a tote with both high frequency (HF) and ultra-high frequency (UHF) RFID systems. The results show that UHF systems have better readability performances than HF for all tested speeds levels on the conveyor belt where HF systems never achieve 100% readability. UHF systems are orientation independent (P=0.875 > 0.05=?) where HF systems are not (P=2.93E-07 < 0.05=?). Moreover UHF systems have higher read rates at reading tags inside a tote (97.61%) than do HF systems (71.70%). After concluding that UHF is more suitable frequency for the pharmaceutical application of RFID, we evaluated methods of improving tag readability within totes. Using reflective materials inside the tote to control the propagation of UHF radio waves were studied by mapping of the RF field inside totes with 0,2,3,4,5 metal sides. According to the results, 5 metal sided tote has better RF propagation than any other configurations. In the comparison of read performances of regular (97.61%), 3 metal sided (98.76%), and 5 metal sided (97.77 %) totes, the 3 metal sided tote has the highest read rate. This tote also has around 15% more probability of identifying all the tags within the tote than the other totes. The performance of each tote was also evaluated as if they were offered 2nd and 3rd time read attempts to simulate the real time process and similar results was observed. Finally it is concluded that the readability of tagged pharmaceutical products and food supplements inside a sealed tote can be improved by utilizing reflective materials inside a tote as in 3 and 5 metal sided totes which also enables tote contents of the tote undisclosed against unauthorized readers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dilek Dagdelen Uysal.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Emond, Jean-Pierre.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021827:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021827/00001

Material Information

Title: Improving the Readability of RFID Tagged Pharmaceutical Products and Food Supplements Inside a Modified Sealed Tote
Physical Description: 1 online resource (95 p.)
Language: english
Creator: Dagdelen Uysal, Dilek
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: distribution, hf, metal, pharmaceutical, rfid, security, tote, uhf
Food Science and Human Nutrition -- Dissertations, Academic -- UF
Genre: Food Science and Human Nutrition thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: One of the biggest advantages of radio frequency identification (RFID) technology is that it can be used to augment the security of the supply chain. A straightforward application area for which security is of utmost importance is the distribution and marketing of pharmaceutical products and food supplements. This research states that an RFID technology can be used to develop an RFID system that can reliably read all the tagged products inside a tote while trying to provide security against unauthorized readers, during distribution of these products. The challenge in using RFID to automatically, without human intervention, identify the tote, its contents, and the RFID enabled security features are due to the problems related to tag readability near and through different materials. The solutions addressing these problems are developed for sealed totes by evaluating of proper system set ups and tote designs. The products and the purchase orders (POs) in the pharmaceutical supply chain were analyzed to provide a repeatable and comparable evaluation. Each product were tagged and tested on a conveyor belt, on a turn table and inside a tote with both high frequency (HF) and ultra-high frequency (UHF) RFID systems. The results show that UHF systems have better readability performances than HF for all tested speeds levels on the conveyor belt where HF systems never achieve 100% readability. UHF systems are orientation independent (P=0.875 > 0.05=?) where HF systems are not (P=2.93E-07 < 0.05=?). Moreover UHF systems have higher read rates at reading tags inside a tote (97.61%) than do HF systems (71.70%). After concluding that UHF is more suitable frequency for the pharmaceutical application of RFID, we evaluated methods of improving tag readability within totes. Using reflective materials inside the tote to control the propagation of UHF radio waves were studied by mapping of the RF field inside totes with 0,2,3,4,5 metal sides. According to the results, 5 metal sided tote has better RF propagation than any other configurations. In the comparison of read performances of regular (97.61%), 3 metal sided (98.76%), and 5 metal sided (97.77 %) totes, the 3 metal sided tote has the highest read rate. This tote also has around 15% more probability of identifying all the tags within the tote than the other totes. The performance of each tote was also evaluated as if they were offered 2nd and 3rd time read attempts to simulate the real time process and similar results was observed. Finally it is concluded that the readability of tagged pharmaceutical products and food supplements inside a sealed tote can be improved by utilizing reflective materials inside a tote as in 3 and 5 metal sided totes which also enables tote contents of the tote undisclosed against unauthorized readers.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Dilek Dagdelen Uysal.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Emond, Jean-Pierre.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021827:00001


This item has the following downloads:


Full Text





IMPROVING THE READABILITY OF RFID TAGGED PHARMACEUTICAL PRODUCTS
AND FOOD SUPPLEMENTS INSIDE A MODIFIED SEALED TOTE























By

DILEK DAGDELEN UYSAL


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2007

































2007 Dilek Dagdelen Uysal


































To my family and friends









ACKNOWLEDGMENTS

I would like to thank my supervising committee chair and my advisor Dr. Emond, for his

never ending guidance, support, and most importantly for being an irreplaceable mentor who

gave me the unparalleled opportunity to work in a motivated environment with such a great

team. I thank my co-chair, Dr. Balaban for helping me not only in my research but in any aspect

of my education and also for believing in me. I thank my supervisory committee member Dr.

Engels for his valuable instructions, immense expertise in this field, and invaluable insight.

I thank Tom Reese from Motorola; Jeff Wells, Steve Dean, Sandy Dean and all the other

members of Franwell for their generous one-on-one help, practical industry advice, and

providing us up-to-date equipment.

Finally, I would never be able to thank my husband, Ismail Uysal, enough for his

patience, encouragement, and devotion along the way. Without him I could never imagine what

I have achieved; he will always have my admiration and love. I would like to extend my

gratitude towards my family, especially my sister Dilber Dagdelen and my brother-in-law Yavuz

Yagiz for their unconditional love and support. Last but not the least I would like to thank my

friends and colleagues for their never ending support.









TABLE OF CONTENTS
page

A C K N O W L E D G M E N T S ..............................................................................................................4

L IS T O F T A B L E S .............................................................................. ............... 8

LIST OF FIGURES .................................. .. ..... ..... ................. .9

A B S T R A C T ................................ ............................................................ 1 1

CHAPTER

1 INTRODUCTION ............... .......................................................... 13

2 L IT E R A TU R E R E V IE W ......................................................................... ........................ 15

Fundamentals of Radio Frequency Identification ............... .......................................15
Tag-R eader C om m unications.................................................................................... 15
F re q u en cie s ............................. ................................................................ ............... 16
L ow frequ en cy ...................................... ............................. ................ 16
High frequency .................................... ................................ ........16
U ltra-high frequency .......................................... ................... .. .. .... 17
M ic ro w av e ......................................................................................17
A antenna Polarization ............................................................. ............... 18
T ag C classification s ................................................................19
Standards (R egulations)........ ............... ............. ........ ............ ...... ..............20
Pharmaceutical and Food Supplements Distribution Chain................................................21

3 EVALUATION OF RFID PERFORMANCE FOR A PHARMACEUTICAL
D ISTR IB U TIO N C H A IN .................................................................................24

Introduction ................... .......................................................... ................. 24
M materials and M ethods .....2........... ... ..... .. .... ............ ........ .. ............. 26
Characterization of Pharmaceutical Products and Food Supplements.........................26
P hy sical ch aracterization ............................................................... .....................2 6
Radio frequency characterization................... ........ ........................... 28
Tagging the Products .......... .. .................................. ........ ...... .................. 29
C onveyor B elt T testing ............ ... .......................................................... ........... ....... 29
Turn Table Testing .................... .................... ...................30
T ote T estin g .................................................... 3 1
Results and D iscussions............................. .. ...... .............. .32
Readability of Products on the Production Line ......................................... .................32
Readability of Products in Free Space ......................................... ......... .... ............... 36
Readability of Multiple Products inside a Closed Environment................................38
S u m m ary .............. ... ..... ... ................................................... ................ 4 0









4 USING RF REFLECTIVE MATERIAL TO IMPROVE RF WAVE DISTRIBUTION
IN SID E A CLO SED EN VIRONM EN T ..................................................... .....................41

In tro d u ctio n ................... ...................4...................1..........
M materials and M ethods ............... ................................................................ .. ....... 42
Preparation of Experim ent Setup.......................................................... ............... 42


Re su ts..... .. .......... ........ ... ... ............................ ................. 46
RF Mapping of Non-Metal Sided Tote ........................................ ....................... 46
RF Mapping of 2 Metal Sided Tote ............. ................ ...............
RF M apping of 4 M etal Sided Tote...................................................... ............... 53
RF Mapping of 5 Metal Sided Tote ............ ... ......... ....... ...............55
D isc u ssio n s .............................................................................................. ...............5 7
Behavior of Reflected RF Waves ............. .... ........... ........................ 57
In su latio n ....... .. ....................................................................................5 8
Security Issues ......... ...... ................................................. ............................58
Summary ......... ............................................. ............... 59

5 EVALUATION OF READABILITY CONTENTS OF PURCHASED ORDERS ............60

In tro d u c tio n ....................................................................................................................... 6 0
M materials an d M eth o d s ..................................................................................................... 6 1
Tote D design ........................... .. ......... ...................................61
Preparation of Representative Purchase Orders ........................................ .................62
C controlled PO ................................................................................................................. 63
PO W ith One Type of Product ................................ ......................... ..............64
M medium L oad P O ................................................................65
F u ll L o a d P O .............................................................................6 5
T testing Procedure ..........................................65.............................
D ata A n a ly sis ............................................................................................................. 6 6
R results .................. ................................................................................................ 68
O ne R ead A ttem pt ..........................................68............................
Second T im e R ead A ttem pt....................................................................................... 69
T third T im e R ead A ttem pt ......................................................................................... 7 1
D isc u ssio n s ............................................. ........ .. ...... .............. ......................7 2
Placement Order of the Products in the Medium and Full Loaded Totes ..................72
Performance Differences of Modified Totes ........................................................... 73
Other Application Methods of Metal Coating ........................ ............... ............. 74
Efficiencies of 2nd and 3rd Read Attempts from the Warehouse Perspective ..................74
Sum m ary ......................................... .................. 75

6 CONCLUSION......................... ................................. 76






6









APPENDIX

A PRODUCTS USED IN THE TESTS .......................................................................78

B PR O D U C T S U H F TA G TY PE ................................................................... .....................83

C PREPARATION OF PURCHASE ORDERS.............................................84

L IST O F R E F E R E N C E S ......... .. ............... ................. ............................................................9 1

B IO G R A PH IC A L SK E T C H .............................................................................. .....................95












































7









LIST OF TABLES


Table page

1-1. R elated ISO R F ID standards ....................................................................... ....................2 1

3-1. Physical characterization of pharmaceutical products depending on their product type
an d p ack ag e m aterial.......... ........................................................................ ....... ............... 2 8

3-2. Radio frequency categorization of products..................................................28

3-3. Means of read rates out of 100 read cycles of individual products on a turntable in
different orientation .................................................... ................... .. ............37

3-4. Means of read rates of 100 totes and probability of getting 100% read rate with HF and
U H F system s ........... ..... ................... ......... ........ ........ ........... 38

5-1. Statistical analysis of read rates of products on turn table* ............................................63

5-2. Example of a calculation of read performance of a tote if the totes offered 2nd time read
a tte m p t* ................... ............................ ............................ ................ 6 7

5-3. Example of a calculation of read performance of a tote if the totes offered 3rd time read
atte m p t* ................................ ................ ....................... ................ 6 7

5-4. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal sided
tote offered 1 read attem pt* ...................................................................... ...................68

5-5. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal sided
tote offered 2nd tim e read attem pt* ............................................. ........................... 70

5-6. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal sided
tote offered 3rd tim e read attem pt* .............................. ........................ ............... 71









LIST OF FIGURES


Figure page

1-1. Linear polarized antenna radiation ........................................................................ 18

1-2. Circularly polarized antenna radiation ............................................................................ 18

3-1. Picture of a turn table testing set-up ........................ ................................. ...............31

3-2. Probability of reading and not reading products in low speed with a) HF system and b)
U H F system ..............................................................................33

3-3. Probability of reading and not reading products in medium speed with a) HF system
and b) U H F system .......................... ........................... .......................34

3-4. Probability of reading and not reading products in high speed with a) HF system and b)
U H F system ..............................................................................35

3-5. Change in read rates of product types according to RF frequency and the orientation of
the product. ............................................................................... 37

3-6. Distribution of the means of read rates of 100 totes with using HF and UHF systems. .......39

4-1. 2 m etal sided tote ..............................................................................42

4-3. 4 m etal sided tote ..............................................................................43

4-4. 5 m etal sided tote ..............................................................................43

4-5. RF mapping inside a tote with different levels of metal coatings set-up. ..........................44

4-6. L ayer arrange ents of m apping tests ...................................................................................45

4-7. Color scale for dBm color coding for all of the signal strength graphs ..............................46

4-8. Mapping results of 0 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d)
L ay er 4 ...........................................................................................4 8

4-9. Mapping results of 2 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d)
L ay er 4 ...........................................................................................50

4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d)
L ay er 4 ...........................................................................................52

4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d)
L ay er 4 ...........................................................................................54









4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d)
L ay er 4 ......................................................... ..................................56

5-1. 5 metal sided tote with a grid pattern texture. ............................................ ............... 62

5-2. Content of the product, equate effervescent pain relief, used in the preparation of one
type of product POs. .......... ..... ..... ...... .......... .............. .. ............. 64

5-3. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 1 read attem pt. ....................................................................... ...................69

5-4. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 2nd tim e read attem pt. ............................................ ............................. 70

5-5. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 3rd tim e read attem pt ...................... ..................................... .............. 72









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IMPROVING THE READABILITY OF RFID TAGGED PHARMACEUTICAL PRODUCTS
AND FOOD SUPPLEMENTS INSIDE A MODIFIED SEALED TOTE

By

Dilek Dagdelen Uysal

December 2007

Chair: Jean Pierre Emond
Cochair: Murat O. Balaban
Major: Food Science and Human Nutrition

One of the biggest advantages of radio frequency identification (RFID) technology is that

it can be used to augment the security of the supply chain. A straightforward application area for

which security is of utmost importance is the distribution and marketing of pharmaceutical

products and food supplements. This research states that an RFID technology can be used to

develop an RFID system that can reliably read all the tagged products inside a tote while trying

to provide security against unauthorized readers, during distribution of these products.

The challenge in using RFID to automatically, without human intervention, identify the

tote, its contents, and the RFID enabled security features are due to the problems related to tag

readability near and through different materials. The solutions addressing these problems are

developed for sealed totes by evaluating of proper system set-ups and tote designs.

The products and the purchase orders (PO) in the pharmaceutical supply chain were

analyzed to provide a repeatable and comparable evaluation. Each product were tagged and

tested on a conveyor belt, on a turn table and inside a tote with both high frequency (HF) and

ultra-high frequency (UHF) RFID systems. The results show that UHF systems have better

readability performances than HF for all tested speeds levels on the conveyor belt where HF









systems never achieve 100% readability. UHF systems are orientation independent (P=0.875 >

0.05=a) where HF systems are not (P=2.93E-07 < 0.05=a). Moreover UHF systems have higher

read rates at reading tags inside a tote (97.61%) than do HF systems (71.70%). After concluding

that UHF is more suitable frequency for the pharmaceutical application of RFID, we evaluated

methods of improving tag readability within totes. Using reflective materials inside the tote to

control the propagation of UHF radio waves were studied by mapping of the RF field inside totes

with 0,2,3,4,5 metal sides. According to the results, 5 metal sided tote has better RF propagation

than any other configurations. In the comparison of read performances of regular (97.61%), 3

metal sided (98.76%), and 5 metal sided (97.77 %) totes, the 3 metal sided tote has the highest

read rate. This tote also has around 15% more probability of identifying all the tags within the

tote than the other totes. The performance of each tote was also evaluated as if they were offered

2nd and 3rd time read attempts to simulate the real time process and similar results was observed.

Finally it is concluded that the readability of tagged pharmaceutical products and food

supplements inside a sealed tote can be improved by utilizing reflective materials inside a tote as

in 3 and 5 metal sided totes which also enables tote contents of the tote undisclosed against

unauthorized readers.









CHAPTER 1
INTRODUCTION

Radio frequency identification (RFID) technologies are the latest branch of information

technologies to begin changing our everyday lives. Several pharmaceutical manufacturers and

distributors are implementing RFID technologies following the US Food and Drug

Administration's (FDA's) 2004 recommendation. The FDA encouraged the use of RFID

technologies in the pharmaceutical supply chains to improve security and prevent counterfeit

products from reaching consumers.1

Over the counter medicine, as well as prescription drugs and food supplements need to be

fully controlled due to the high risk of counterfeiting. According to World Heath Organization

(WHO) between 5% and 8% of pharmaceutical products are counterfeit worldwide,2 which costs

an average of $40 billion lost every year.3 When the drugs reach to the wholesaler they are often

broken down from cases and pallets and shipped in reusable plastic totes where totes contain the

items of a purchase order (PO) placed by a retailers, hospitals or another wholesalers. One of

the best ways to keep a tight control during their transportation from the distribution center is to

place every order in a sealed tote where the products inside are labeled with an RFID tag and the

tote is scanned without opening it.

With the popularity of blister packs and liquid products in the pharmaceutical industry,

the probability of not reading an RFID tag is significant when the tagged products are placed

randomly within a tote due to the attenuation of the RF signal.4 Furthermore, a mixed purchase

order with just a few blister packages may create enough interference to reduce the read rate (the

percentage of RFID tags read) of a tote significantly during the scanning of that tote.

This research states that an RFID technology can be used to develop an RFID system that

can reliably read all the tagged products inside a tote while trying to provide security against









unauthorized readers, during distribution of these products. Therefore, the objective of this

research is to improve security and readability of the RFID tagged pharmaceutical products and

food supplements inside a sealed tote by proper designs of the tote and portal. The problems in

the readability of different types of POs are addressed by development of a novel tote design as

well as a system encompassing all the aspects from the product picking to scanning process.

The findings of this thesis have significant contributions not only to the pharmaceutical

industry but also have direct implications for researchers developing new RFID technologies and

will also help with the implementation of RFID to real life systems. It enables a secure

distribution of pharmaceutical products by suggesting a reliable RFID system set-up for the item

level pharmaceutical application which can also be applied to the other item level applications.

Moreover, it serves as a guide for the beneficial use of reflective materials to improve readability

performances of UHF systems.

In the following chapters, first background information about both the RFID technology

and the pharmaceutical distribution chain are provided. Then, the performance of RFID

technology was evaluated for a pharmaceutical distribution chain. In this chapter the most

suitable frequency, UHF or HF, for the item level pharmaceutical application was analyzed. In

chapter 4, the usage of reflective material to improve RF wave distribution inside a tote was

studied by mapping of RF power inside the totes with different levels of metal coatings. The

chapter 5 evaluates the readability performances of the totes that are found to have better RF

power distribution compared to a regular tote.









CHAPTER 2
LITERATURE REVIEW

Fundamentals of Radio Frequency Identification

Radio frequency identification (RFID) technologies are automatic identification

technologies that use tags attached to objects to wirelessly communicate information about that

object to a reader. The primary building blocks of RFID systems consist of a tag, a reader, and a

host system that processes the data collected from tags by the reader.

The reader sends radio frequency (RF) waves through its antenna and collects the RF

waves emitted or reflected from the tag. The information, such as a unique serial number for

every tag, is carried with these RF waves. RF waves can be absorbed or reflected by certain

materials, which impacts the readability of the tags by the reader.

Passive RFID systems are beginning to be widely deployed in retail and pharmaceutical

supply chains to improve inventory visibility, asset management, supply chain efficiencies, and

supply chain security.5

Even though passive RFID systems are being deployed and piloted for improved supply

chain management and security, the technologies and systems being implemented are rapidly

evolving and their use is not well understood. Significant technical, operational, and social

problems remain to be solved to achieve the maximum possible benefits from the use of RFID

systems. Many of these problems are intimately connected to the tags, the frequencies, and the

standards being used.

Tag-Reader Communications

There are two methods of communications between the passive tag and the reader: near

field inductive coupling and far field backscatter communications. In inductive coupling, the

reader antenna coil generates an alternating magnetic field that when part of it is enclosed by the









tag antenna a current is induced in the tag antenna providing a power supply to the tag. The

information in the tag is communicated to the reader by having the tag change how much power

it draws from the reader's magnetic field. The modulation in the tag load is detectable by the

reader. Far field backscatter communications use reflected, or backscattered, reader emitted

signals to communicate from the tag to the reader. While inductive coupling relies on the fact

that the reader emitted signal is still electrically connected to the antenna, backscatter

communication works when the emitted signal is not connected to the reader antenna. The tag is

powered up by harvesting electrical and magnetic energy from the electromagnetic field created

by the reader.6

Frequencies

There are 4 primary ranges of frequencies within which passive RFID systems operate:

low frequency (LF), high frequency (HF), ultra-high frequency (UHF), and microwave.

Low frequency

LF tags operate in the frequency range less than 135 kHz by using inductive coupling.

Passive tags using LF frequencies have demonstrated read-ranges up to 2.0 m.7 Although low

frequencies enable RF waves to go through metal and water, they have longer read times, higher

costs, and larger tag sizes. The maximum rate of tags that existing LF system can read is limited

to 20-25 tags a second.8 Their application today is mostly limited to animal ID, access control,

and vehicle immobilizers.

High frequency

HF tags operate in the ISM frequency band centered at 13.56 MHz and are similar to LF

tags in the sense that they use inductive coupling. Passive tags using HF waves may have read-

ranges up to 1.5 m depending upon antenna sizes.9 RF waves with this frequency can penetrate

through water; however, metal objects still constitute a problem. HF tags have faster read rates









than do LF tags where up to 800 tags/s can be read with the newer standard protocols.9 Their

primary applications today are in smart cards, access control systems, vehicle immobilization

systems, subway cards, and the Mobile Speedpass system.

Ultra-high frequency

UHF tags operate in the frequency range from 860 MHz to 960 MHz. Unlike LF and HF

tags, they use electromagnetic coupling communications, i.e., backscatter communication in the

far field of the RF radiation. UHF passive tags have exhibited a maximum read-range of 20m.10

The new Gen 2 standards provide a read-rate up to 1000 tags/s for the passive tags.1 Unlike HF

and LF, UHF is more affected by materials; it is highly absorbed by liquids and reflected by

metals, especially for far field tags. However, recent introduction of near field tags operating in

the UHF band provides similar advantages that the lower frequencies have, such as being less

effected by water.12 UHF tags are being deployed in supply chain applications at the case and

pallet level and passive and semi-passive UHF tags are used for toll collection applications.

Microwave

Microwave tags operate in frequency bands centered at either 2.45 GHz or 5.8 GHz.

They use far field communications as a result of which electromagnetic coupling is employed.

They have the fastest read-rates, although their read ranges are significantly less than those

achievable with UHF passive tags.

In summary, all frequencies have their respective advantages and disadvantages. As the

carrier frequency gets higher, tag sizes get smaller in general, the read range and read rate

increases but the impact of environmental materials increases.13 These properties should be

considered carefully for different system designs and applications in order to utilize the most

appropriate RFID technology.








Antenna Polarization

Polarization is defined as the orientation of an electric field of the wave. Most antennas

radiate via either linear or circular polarization. In the linear polarization, the radiation pattern of

an antenna stays in the same plane at all times. Depending on the propagation direction with

respect to the earth surface; it can be vertically or horizontally polarized (Figure 1-1).14

//



iI, _f .r:.







Figure 1-1. Linear polarized antenna radiation.14

A circularly polarized antenna radiates an electric field which is the sum of two fields

that have equal amplitude and magnitude with a 900 phase difference; thus it rotates in a circular

motion completing a full 3600 within one wavelength (Figure 1-2)14.15 Circularly polarized

antennas can either be right-handed (clockwise rotation) or left-handed (counter clockwise

rotation).,


Figure 1-2. Circularly polarized antenna radiation.14









Tag Classifications

While we are concerned with passive tags, other types of RFID tags exist. The tags may

be classified according to their source of operating and communicating power and

communication method. There are 4 basic types of tag: active, semi-active, semi-passive, and

passive. Active tags have their own power source (typically a battery) and are able to actively

generate and transmit an electromagnetic signal. They typically have higher performance, larger

memories, and longer read-ranges than do the other types of tags.16 However, their application is

limited primarily due to their higher costs, larger form factors, and limited power sources which

limits their lifetime.

Semi-active tags, like active tags, have an internal power source and use active

communication; however, they do not continuously communicate like active tags. Instead, they

are often activated manually, such as the keyless entry system in a vehicle or a garage door

opener. The use of active communication significantly increases the communication robustness

in the presence of noise, allowing for more reliable communications at short ranges and overall

longer possible communication ranges as compared to passive communications.

Semi-passive tags, like active and semi-active tags, have an on-tag power source, but

unlike these two types of tags, semi-passive tags use passive communication, either load

modulation or backscatter, to communicate with a reader. The most common type of semi-

passive tag today is a battery assisted passive tag operating in the UHF band. By using their on-

tag power to operate the logic and functionality of the tag, semi-passive tags are able to

communicate at roughly an order of magnitude further distance than can passive tags, and they

communicate at roughly an order of magnitude less distance than do active and semi-active tags.

Passive tags contain no on-tag power source. Instead they harvest all of their operational

energy from the reader's communication signal and use passive communication techniques,









either load modulation or backscatter, to communicate with the reader. The need to harvest all of

their operational energy from the reader's communication field limits the communication range

of passive tags to no more than a few meters at best.

The biggest potential for the mass deployment of RFID systems lies within these tags due

to their low cost and small sizes. However, the performance of passive tags is more sensitive to

orientation and environmental materials than are semi-passive, semi-active, and active tags.

Their placement with respect to the polarization of the antenna and product package and content

has a significant effect on readability.17 18

The tags have been classified by the Auto-ID center according to their operation

characteristics and functionality as Class 0 through Class 5.19

> Class 0 Passive read only tags.
> Class 1 Passive write once, read many tags.
> Class 2 Passive rewritable tags
> Class 3 Semi-passive tags
> Class 4 Active tags power up itself
> Class 5 Active tags ability to power up other tags as well.

Standards (Regulations)

Regulations related to RFID systems are created to ensure safe operation with other

devices that use RF such as nearby radio and television stations, mobile radio services (police,

security services, industry) and cellular phone to prevent any interference. There are many

standards governing RFID systems, but The International Standards Organization (ISO) and

EPCglobal are the major organizations that set standards for RFID systems on data, air interface,

conformance and its applications. Table 1-1 presents ISO standards related to this research.20

EPCglobal is another organization which regulates the electronic product code (EPC) that

has a unique number for every item in the supply chain. The standards mainly focus on different

classes of UHF tags as well as readers.21









Table 1-1. Related ISO RFID standards
ISO Standards Title

ISO 18000 -1 Generic parameters for the air interface for globally accepted
frequencies
ISO 18000 -2 Parameters for air interface communications below 135 kHz

ISO 18000 -3 Parameters for air interface communications at 13.56 MHz

ISO 18000 -4 Parameters for air interface communications at 2.45 GHz

ISO 18000 -5 Parameters for air interface communications at 5.8 GHz

ISO 18000 -6 Parameters for air interface communications at 860-930 MHz

ISO 18000 -7 Parameters for air interface communications at 433 MHz

ISO 15693 Identification cards- contactless integrated circuit(s) cards-
vicinity cards for 13.56 MHz

ISO 10374 Freight containers automatic identification

ISO 17363 Supply chain application for RFID Freight containers

ISO 18185 Freight containers radio frequency communication protocol
for electronic seal


Pharmaceutical and Food Supplements Distribution Chain

Prescriptions drugs and over-the-counter (OTC) products such as food supplements and

personal health care products constitute the goods distributed via the pharmaceutical supply

chain. Prescription drugs are the ones that affect the distribution system most with their highly

restricted regulations.

Before the products reach retail pharmacies or hospitals they pass through at least 4

different stages. The active ingredient of the product is produced in a chemical plant and carried

to the manufacturers for packaging after which the wholesalers distribute the product to the final

seller (retailers or hospitals).









In the United States 90% of the distribution of prescription drugs are made by only five

major companies. These companies are primary, authorized distributors who have an ongoing

relationship with the manufacturer and purchase a large majority of their drugs directly from

them. Secondary wholesalers purchase their products generally from authorized wholesalers

with a rare purchase directly from the manufacturer or from other sources. Secondary

wholesalers are used by pharmacies to obtain rare products or to purchase drugs when the

pharmacy is in an area that none of the primary distributors have sufficient service. According to

FDA's report to the Congress, it is estimated that there are more than 6500 wholesalers in US.

Of these, 83% are small (fewer than 20 employees), 11% are medium-sized (with 20 to 99

employees), and 6% are large (with more than 100 employees).22

When it comes to the safety of pharmaceutical products throughout the chain, the

wholesalers play a very important role because most of the counterfeited products are introduced

to the market during the distribution process. A counterfeited drug is defined by WHO as

follows: "A medicine that is deliberately and fraudulently mislabeled with respect to identity

and/or source. Counterfeit medicines may include products with the correct ingredients but fake

packaging, with the wrong ingredients, without active ingredients or with insufficient active

ingredients."23

To prevent counterfeiting, the FDA has only recently indicated that they intend to enforce

the pedigree provision of Prescription Drug Marketing Act (PDMA) in the regulations in 21 CFR

part 203 which concentrates on the regulation of secondary wholesalers.24 PDMA of 1988

requires a pedigree, documentation on a product history, for each product to track them

throughout the supply chain to ensure the safety and the effectiveness of the products. Hence,

for a complete pedigree a tracking and tracing of every product especially during their









transportation by wholesalers are required. The adoption of RFID technology offers efficient

preparation of e-pedigree which will enable the localization of every product in the market. This

will expedite the distribution progress while adding more security features and saving companies

both time and money.









CHAPTER 3
EVALUATION OF RFID PERFORMANCE FOR A PHARMACEUTICAL DISTRIBUTION
CHAIN

Introduction

Implementation of RFID technology to pharmaceutical supply chain is highly favorable

mainly due to the safety and security concerns in this field. By providing better tracking and

tracing with a complete electronic pedigree, RFID will not only deter counterfeiting products but

also provide complete supply chain visibility which precludes tampered products as well as

products with unacceptable status.25 Other significant advantages of RFID include

improvements in supply chain efficiencies and asset management like in other supply chains.26' 27

After the recognition of more benefits and the recommendation of FDA in 2005, RFID

started to attract a greater interest in the pharmaceutical industry. However, only ten

pharmaceutical products are currently being planned to be tagged by the end of 2007 which

indicates a slow progress towards its realization.28 One of the main reasons is that the companies

are still waiting for the conflict to be resolved among the two major RFID systems, high

frequency (HF) and ultra-high frequency (UHF), before investing in this technology.

The three main reports in the literature which study both systems for a side-by-side

comparison for the pharmaceutical industry are prepared by Texas Instruments29, ODIN

technologies & Unisys Corp30 and leading UHF vendors11. The first two papers suggest that HF

has more advantages over UHF at the item level. Their common conclusion is that even though

UHF is not speed dependent and has a greater read range; HF should still be used for the item

level pharmaceutical packaging due to its superior performance with liquid products.

However, according to the paper by UHF vendors, the recent introduction of UHF Gen 2

"near field tags" changed the performance of UHF systems significantly. These tags have

antenna that are more efficient at harvesting the energy in the magnetic field than the more









commonly used UHF tags. This enables better performance through liquids providing similar

performance to HF tags. The same UHF tag can also have both near and far field features which

offer the advantages of both systems.12 31 Moreover, Wal-Mart, the world's largest retailer, and

the US Department of Defense (DoD) are leading the adoption of passive RFID systems by

requiring passive UHF tags to be placed on all pallets and cases delivered to their supply

chains.20, 32,33 Hence, if a company chooses to use an HF system, it might be forced to tag its

products with both HF and UHF tags, which might increase the cost of overall implementation.

Both systems still have problems to be solved before practical implementation

worldwide. The author believes that the most important issue that needs to be addressed is a

thorough frequency performance analysis for these two systems. After all, the real advantage of

using passive RFID is the ability to utilize complete automation which requires fail-proof

performance.

For better evaluation of the systems, each step in the pharmaceutical distribution chain

through which the products flow should be considered and simulated. The usage of RFID starts

right after the packages are filled during the manufacturing process. In this step, each tagged

product is scanned individually on the conveyor belt and then cased or palletized to be sent to the

wholesalers. The wholesalers separate the products according to their ID and placed them onto

the shelves. The orders of the retailers are put into the sealed totes which are scanned and

confirmed before the distribution. Then, the retail pharmacy rescans each tote without opening

it. If the content is confirmed the seal is opened and the products are placed onto shelves for

sale. If it is not the tote is sent back to the wholesaler.

The objective of this chapter is to analyze the advantages and disadvantages of using HF

and UHF systems in pharmaceutical supply chains. State-of-the-art technology is used for the









evaluation of both systems to prevent any bias. While simulating each step in the product flow

through the distribution chain following criteria are used for comparison:

Scanning speed by testing products in the production line,

Orientation and material sensitivity by turn table test,

Performance of reading several products at the same time in a closed environment by tote
testing.

Materials and Methods

Characterization of Pharmaceutical Products and Food Supplements

The pharmaceutical industry has a wide variety of products in terms of the physical

properties of both the products and their packages. In order to analytically gauge the

performances of both RFID systems and to provide a repeatable and comparable evaluation,

characterization of the products needed. Two primary types of characterization were performed:

Physical characterization: Characterize the products according to products types and the
types of materials used in their packaging

Radio frequency characterization: Characterize the products according to their effects on
RF waves.

Physical characterization

The physical state of pharmaceutical products is usually categorized into 4 groups; solid,

semi liquid, liquid and gas.34 Examples of products that fall in these categories are as follows:

Solid products: Tablets capsules, powders, granules, patches etc.

Semi-liquid products: Creams, gels, foams etc.

Liquid products: Syrups, oral liquids, droplets, solutions etc.

Gas products: Liquefied or pressurized gasses

After a market research it was found that this list is not comprehensive enough to identify

all the products distributed via pharmaceutical supply chain. As it is indicated in chapter 2, OTC









products are also considered as pharmaceuticals with a bigger market share than the prescription

drugs.35 Some of the very common products in this group are health care products that are not

consumed as drugs per se but used during a treatment. Such products can also be prescribed

along with actual drugs. Hence, to comprise these products in the list another category is created

as follows:

Sub-products: Needles, syringes, health care products, etc.

For the packaging materials, the products are categorized based on only their primary

containers such as in plastic, in metal or in glass. Paper was not included in this list since it is

not a common material that functions as the main containment of the physical product itself.

Recent developments in plastic enabled it to be the most widely used material in the

pharmaceutical industry where it could be used in 99% of all products within 5-10 years.34 The

common applications of plastic are bottles, closures, blister packs and film layers. Unlike

plastic, metal usage as a whole container has declined in recent years, but since the blister packs

are being used more frequently (approximately 85% of solid drugs in Europe are packed in

blisters36) metal is still a very popular material for pharmaceutical products. Moreover the

applications of metal are more widespread than any other materials: such as lids, aerosol

containers, blister packs, closer, and sachets. Glass is the least common material for

pharmaceutical products. However, it is still very valuable especially for the liquid ones. Bottles

for the liquids, ampoules and vials are different applications of glass packaging in

pharmaceutical industry.

Based on the above information and discussions, the Table 3-1 is prepared to summarize

how the pharmaceutical products were categorized according to their physical properties for this

research.









Table 3-1. Physical characterization of pharmaceutical products depending on their product type
and package material


Product Type Package Material

a) In plastic b)In metal c) In glass

1. Solid X X N/A

2. Semi-liquid X X N/A

3. Liquid X X X

4. Gas N/A X X

5. Sub-products X X N/A


Radio frequency characterization

The physically characterized products are grouped into 3 categories, RF lucent, RF

absorbent and RF opaque, according to their containments/materials interaction with RF waves.

RF lucent products are defined as the products that do not affect RF waves. RF absorbent

products cause attenuation on the RF waves by absorbing their energy and similarly RF opaque

products do not let RF waves pass through by reflecting them.37 In Table 3-2 the products that

contain substances with high dielectric constant, such as water, were placed in the RF absorbent

category since it absorbs electromagnetic RF waves and products with conductors, such as

metals, are placed in the RF opaque group.

Table 3-2. Radio frequency categorization of products

Package Categories
1. RF lucent 1.a, 2.a, 4.c, 5.a

2. RF absorbent 3.a, 3.b, 3.c

3. RF opaque 1.b, 2.b, 3.b, 4.b, 5.b









Tagging the Products

For tagging the products the most suitable tags and the best locations to place them to get

the highest read rates during the production line, individual read, and inside the tote were chosen

according to following criteria:

Package content

Package surface availability for the size of the tag to prevent performance loss due to
antenna curving38'39

Application of tags onto products during manufacturing and labeling40, 41

However, the products were not only evaluated independently but also in a group of

different other products to minimize tag collision problems which may occur in the tote

application.

48 products (6 RF lucent, 25 RF absorbent, 17 RF opaque) were first tagged with UHF

Symbol (Rockville, MD) RFX6000 lxl far field and UHF Avery Dennison (Clinton, SC) Gen 2

Satellite near/far field inlays respectively. According to the preliminary tests results of conveyor

belt and turn table tests the tag which has the highest performance depend on the paired-t test

with a significant level of 0.05 was chosen for each product for the UHF system. The

determination of UHF tag type can be found in Appendix B.

For the HF system, HF Rafsec (Fletcher, NC) ISO 15693 14 x 31 mm inlays were used

during the conveyor belt and turn table tests. However, these tags were replaced by Omron

(Schaumburg, IL) V720-D52P04 inlays due to having higher read distances for the tote testing.

Conveyor Belt Testing

The tagged product are scanned on the conveyor belt 5 times with 3 different speed

levels; 0.17 m/s as the low speed, 0.34 m/s as the medium speed, and 0.51 m/s as the high speed.

A line was created on the conveyor belt to keep products in the same position throughout the









belt. At the end of the line the antennas were placed to be on the top or the side of the product

based on the tag position for each package to simulate the manufacturing line. The antenna used

for UHF tags was a Symbol linear conveyor belt antenna with a 30 x 50 mm dimensions and for

HF, a 190 mm loop antenna with Feig Electronics (Hessen, Germany ) tuning board which was

tuned properly right after the set-up was used. A Symbol XR 400 fixed reader and a Feig

Electronics OBID i-scan ID ISC.LR200 reader were used for UHF system and HF system,

respectively. To compare the two RF systems, the probability of reading a product was

calculated for each product type and speed level.

Turn Table Testing

As shown in Figure 3-1, the already tagged products for the conveyor belt testing were

put onto a turntable which was perpendicular to the antenna with its center 100 mm away. The

antenna position was decided to be parallel to the ground which is a common placement in the

smart shelves. This placement will also provide a better setup for the tote testing discussed in the

next section. The read rate of a product out of 100 read cycles was recorded for 5 different

angles which were 00, 450, 900, 1350and 1800 with 00 indicating its upright position.

The same readers as in the conveyor belt testing were used for both HF and UHF

systems. However, the Symbol AN200 circularly polarized antenna for the UHF system and a

250x 250 mm loop antenna which was tuned with Feig Electronics tuning board for the HF

system was employed.

For the analysis, the products were considered within their groups as in the previous test.

However, instead of a full probability comparison, the means (first order statistics) of read rates

were evaluated. Paired-t test was conducted between both systems within the same orientations

and ANOVA (analysis of variance) was performed between the orientations to decide whether

the system was orientation dependent or not.


































Figure 3-1. Picture of a turn table testing set-up. Circularly polarized antenna was placed under
the products attached to a turn table.

Tote Testing

A total of 100 different totes, 50 full loaded (with 48 products), 30 medium loaded (with

24 products), 10 with one type of product (with around 32 products) and 10 with controlled

products (with 13 products), were scanned on a conveyor belt with a low speed using both UHF

and HF systems. Each tote was scanned 3 times (see Chapter 5 for more information about the

preparation of totes).

In the UHF set-up the totes was scanned with using a Symbol XR440 reader and 1

Symbol AN200 circularly polarized antenna which was placed under the conveyor belt. Based

on the previous test results all the products were tagged with UHF near/far field inlays for better

performance. Moreover, besides Avery Denison Gen 2 Satellite tags, Omron Gen 2 V750









"Ninja" near/far field tags which are found to have higher read range were placed on some

products containing liquid.

Unlike UHF, in HF setup a tunnel with 4 antennas tuned with Feig tuning boards was

created on the conveyor belt such that the tags were able to be scanned in any orientation inside

the tunnel. The same reader used for conveyor belt and turn table testing was employed along

with Feig ID ISC.ANT.MUX 8 Channel multiplexer.

The data is analyzed in two different ways. First the read rates of totes, i.e., the read

percentage of the content of the tote, are evaluated. Then, the percentages of getting a 100%

read rate from a tote out of 100 totes are compared.

Results and Discussions

Readability of Products on the Production Line

Figures 3-2, 3-3 and 3-4 show the contingency analysis of response by products in

different speed levels with HF and UHF systems. In the figures, "0" (red colored areas)

represents not reading a product and "1" (green colored areas) represents reading a product at

least one time.

According to Figure 3-2 in the low speed all the products that are in the RF absorbent and

RF opaque groups were read with HF system. However, 12% of the RF lucent products was not

detected. Note that even though some products were containing small portion of conductive

materials such as aluminum lids or needles, they were still categorized as RF lucent products (see

Appendix A for the products categories). Hence, not reading some products in this group can

still be attributed to sensitivity of the system either to orientation or material. The overall

probability of reading products in low speed with HF system was 97.50%. In contrast, UHF

system had 100% probability of reading products by detecting all the products on the line for the

same speed level.
















U-rr ii


RF absorbent


RF lucent RF opaque
RF lucent RF opaque


products


a)


RF absorbent RF lucent RF opaque
RF absorbent RF lucent RF opaque


products


b)


Figure 3-2. Probability of reading and not reading products in low speed with a) HF system and
b) UHF system.


0




0 .-



0.uu-


1.00-



0.75-

C-
o0.50-



0.25-



0.00-


I












Figure 3-3 illustrates the probability of reading and not reading products in medium


speed. Where 23.64% of RF absorbent, 32.00% of RF lucent, 17.50% of RF opaque products


were not read by the HF system, the UHF system read all except 2.50% of RF opaque products.


The performance of the HF system for medium speed has decreased 21.36% compared to low


speed by having a total of 76.67% probability of reading all products. However UHF has a very


negligible decrease which is less than 1% by reading 99.17% of all products.


0.75-

(,
o
-


0-
(U)


0.25-



0.00-







1.00-


0 5-


(,

c,
00 5,0-
0)


0 .005-



0.00-


RF absorbent RF lucent RF opaque


products



17__ _


1 I I i
RF absorbent RF lucent RF opaque


products


Figure 3-3. Probability of reading and not reading products in medium speed with a) HF system
and b) UHF system.


1111










Finally, Figure 3-4 demonstrates the readability of products in high speed. With the HF

system, almost half of the products in RF absorbent and RF opaque groups and 68.00% of the

products in RF lucent group could not be read. On the contrary with UHF only 3.64% of RF

absorbent and 5.00% of RF opaque groups could not be read. The overall HF performance

showed even a higher decrease of 39.13% compared to medium speed by having 46.67% of

readability of all products whereas UHF had only a slight performance loss (2.50%) compared to

medium speed by reading 96.67% of all products.


a) 1.00












0.00
RF absorbent RF lucent RF opaque

products


1.00-
b)

0.75-
o.

0 .50- 1
0)


0.25-
cii




0.00---
RF absorbent RF lucent RF opaque

products

Figure 3-4. Probability of reading and not reading products in high speed with a) HF system and
b) UHF system.









We conclude that HF is a speed dependent system with a gradual decrease in the

probability of reading products on the conveyor belt with increasing speed. More importantly, it

could never achieve a 100% read rate even at the low speed level. Even though the UHF system

is also affected by conveyor belt speed, it is not as significant as HF. Moreover, the performance

of the system is sufficient to get a 100% read at the low speed level. Finally, it is worth

mentioning that the speed level for the UHF system can be increased in small steps because the

performance difference is marginal between low and medium speeds. Depending on the product

type the manufacturer can choose to use higher speeds without any significant adjustments to the

system setup.

Readability of Products in Free Space

The statistical means of read rates out of 100 read cycles of individual products on a

turntable in different orientations are shown in Table 3-3. According to pair-t test results HF has

a statistically better performance than UHF only at reading RF absorbent products at 900. In all

other product-orientation combinations UHF has either significantly better results (reading all

types of products at 00 and 450) or has a similar performance with no statistical differences.

Moreover, if the read rates of orientations averaged for every product and system are compared,

UHF is found to be significantly better (P=0.00135 < 0.05=a) at reading different types of

products.

Moreover, Figure 3-5 demonstrates that even though there are no statistical differences

between 5 different orientations (00, 45, 900, 1350, 1800) in UHF (P=0.875 > 0.05=a), HF is

observed to be orientation dependent with a significant difference between orientations

(P=2.93E-07 < 0.05=a) where 900, 1350, 1800 are similar and better than 00 and 450 at scanning

a tag on different products.










Table 3-3. Means of read rates out of 100 read cycles of individual products on a turntable in
different orientation *


0 Degree 45 Degree 90 Degree 135 Degree 180 Degree

HF UHF HF UHF HF UHF HF UHF HF UHF

RF 45.45 85.45 36.36 73.45 90.90 63.63 81.72 72.72 63.27 72.72
absorbent B A B A A B A A A A

RF lucent 40.00 100 40.00 100 100 100 100 100 100 100
B A B A A A A A A A

RF opaque 25.00 94.12 61.25 100 87.50 100 98.50 100 87.50 100
B A B A A A A A A A

*Same letters represent no difference.





110

100-- --
90 -


-A- rf absorbent hf
-- rf absorbent uhf
rf lucent hf
rf lucent uhf
-- rf opaque hf
-- rf opaque uhf


Angle


Figure 3-5. Change in read rates of product types according to RF frequency and the orientation
of the product.









To conclude, HF reading RF absorbent products better than UHF at one degree proves

the theory that HF is less material sensitive than UHF.9 However, while HF might be superior

when reading liquid products, it can be claimed that it is still negatively affected by metal like

UHF. Furthermore, a multiple antenna setup is required for HF to minimize its reliance on

orientation whereas a single antenna is sufficient for UHF which increases its cost efficiency.

Readability of Multiple Products inside a Closed Environment

Table 3-4 summarizes the experimental results for 100 different product mixes

corresponding to distinct Purchase Orders (POs). The POs were scanned with both HF and UHF

systems. The mean distributions for different types of POs were also added to the results for a

detailed overview. The mean of total read rates with the UHF system is 97.61+0.52, [97.09,

98.12] which is substantially higher than the mean read rates for the HF system that is

71.701.90, [69.80, 73.56] with a 95% confidence interval. Figure 3-6 provides a quick

visualization of the performance differences and data distribution of the two RFID systems. As

can be observed from this figure, the UHF system is a more reliable system by having less

variance in performance while reading different totes when compared to the HF system.

Table 3-4. Means of read rates of 100 totes and probability of getting 100% read rate with HF
and UHF systems

PO Types Number of Probability of getting
Samples Read Rates 100% read rate

UHF HF UHF HF

Control 10 98.46 74.62 8.00 0.33

One type 10 97.18 65.42 6.00 0

Medium tote 30 96.67 80.28 12.00 0

Full tote 50 98.08 67.22 21.00 0
97.61 71.70
Total 100 [97.09, 98.12] [69.80, 73.56] 46.00 0.33






















a)


70-



60-



50-


HF UHF

system

Figure 3-6. Distribution of the means of read rates of 100 totes with using HF and UHF systems.

Another significant statistic is the occurrence of 100% read rates. Table 3-4 shows that

the UHF system has outperformed the HF system in this category, as well. Out of 100 totes, the

contents of 46 totes were fully scanned with the UHF system where as not even 1 tote could be

read entirely with the HF system. The goal of using RFID tagged products is to obtain 100%

read rates for all totes.

To sum up, UHF has a superior performance than HF at reading products in a closed

environment such as in a tote. Getting such a lower reads from HF can be attributed to not only

HF being speed and orientation depended as proved in previous tests but also having a lower

performance at scanning multiple products together. The individual results for different totes

showed that, in general, the read rate decreased as the amount of products included in the totes

increased.









Summary

Recent innovations in RFID technology promise many advantages for pharmaceutical

supply chains. However, there are many details that need to be considered regarding the system

setup, some of which were highlighted in this chapter.

Three different tests were performed to analyze various properties of HF and UHF

systems. Even though both systems have their pluses and minuses, UHF pulls forward when

pharmaceutical applications are considered in particular. The main reason is that UHF systems

have less overall dependency on system setup for optimal performance. Other reasons include

the fact that they are less affected by speed and orientation and finally becoming less sensitive to

materials as well, with the recent introduction of near/far field UHF tags.

To conclude, this chapter of the research recommends the use of UHF systems in the item

level pharmaceutical applications. Even though a better HF system set-up (reader/antenna

configuration) could be used to improve the individual product read rate, UHF systems have a

greater opportunity to be used in a physically secure tote due to the angular dependence of HF

systems.









CHAPTER 4
USING RF REFLECTIVE MATERIAL TO IMPROVE RF WAVE DISTRIBUTION INSIDE A
CLOSED ENVIRONMENT

Introduction

This chapter states that RF propagation can be guided to a better distribution of RF power

by the reflective materials inside a closed environment. The motivation is that the metal reflects

UHF radiation; therefore, metal coatings on the sides and within the totes can be used to increase

the received power of the tag inside a tote which might result in the increase of overall

readability of tote content. The metal can also be utilized as a physical security barrier to prevent

unauthorized readers scanning the tags inside a metal cage which provides another security

aspect usually concerned in the RFID systems.

As the principle idea, electromagnetic waves reflect off metals at the same angle they

encounter the surface with a phase reversal (1800 phase shift).42' 43 During this reflection

insignificant power loss is observed since metals have lower index of refraction.44

The wave energy is also affected by interference of waves at the same frequency. This

interference can either be constructive or destructive depending on the phase, amplitude and the

polarization of waves. For instance 0 to 1200 and 2400 to 3600 phase differences result in

constructive interference whereas 1200 to 2400 phase differences cause destructive interference

in the waves with same polarizations and same frequencies.45

In summary, the objective of this part of the research is to determine the RF power

distribution within totes that have different levels of metal coatings. In order to accomplish this

goal, mappings of RF wave power inside totes with 0 (control), 2, 3, 4, and 5 sides coated with

reflective materials were experimentally determined.









Materials and Methods


Preparation of Experiment Setup

5 different totes with different levels of aluminum coating as a reflective material were

prepared. The totes that are tested are listed as follows:

Non-metal sided tote as a control unit
2 metal sided tote has metal coating on the 2 adjacent sides (Fig. 4-1)
3 metal sided has metal coating on the 2 adjacent sides and the top (Fig. 4-2)
4 metal sided tote has metal coatings on 4 adjacent sides (Fig. 4-3)
5 metal sided tote has metal coating on all sides except the bottom (Fig.4-4)

The totes were placed 50 mm away from the top of the UHF circularly polarized antenna

with a 6 dBi gain, Symbol AN200, which is connected to a Symbol XR400 reader.


Figure 4-2. 3 metal sided tote.


I


Figure 4-1. 2 metal sided tote.


































Figure 4-3. 4 metal sided tote.


Figure 4-4. 5 metal sided tote.


43









Taking the Measurements

The RFID system was operated with the highest output power of 30 dBm. The

distribution of RF wave power inside the totes was measured with the Agilent/HP 8595E RF

spectrum analyzer. The frequency of the spectrum analyzer was centered at 915 VMHz with a 200

MHz span. As shown in Figure 4-5, a 38 mm omni directional antenna with a 1 dBi gain was

placed at the center of each cell on a grid platform with 4x4 cells and the signal strength was

recorded at every cell. This process is repeated throughout the tote where the platform was

moved to the centers of 7 intervals that divide the tote into equal spaces.
























Figure 4-5. RF mapping inside a tote with different levels of metal coatings set-up.

Analysis of Data

During the analysis the data is compressed into 4 layers instead of 7 for better

visualization of the power distribution. Layer 1 is the closest layer to the front and layer 4 is the

furthest as illustrated in Figure 4-6. The data on each layer is shown with a surface chart where









each corer except the ones on the edges represents a value taken from the measurements. Since

it was impractical to take data on the sides of the totes, a suitable value was assigned to the

corners at the edges in order to map the entire tote. This value is determined in such a way that

for the data points at the edges if there is no metal coating, the values of closest row or column

are averaged and if there is a metal coating -10 dBm was assigned as a minimum value to power

up the tag46 to show that the tags can not be read on these surfaces.





























Figure 4-6. Layer arrangements of mapping tests. Layer 1 is the closest layer to the front and
layer 4 is the furthest.

In the results and the discussion sessions the totes with metal coatings are evaluated as if

there were 10 mm insulators after the metal surfaces. The insulator borders are presented in the

surface charts as dashed lines.









The scale for the RF wave power in dBm was color coded from red to blue as signal

strength increase, the color changes from blue to red as shown in Figure 4-7. Light blue colors

show the null spots where a tag cannot be read.

0 35-40
m 30-35
m 25-30
m 20-25
S15-20
S10-15
o 5-10
0-5
S-5-0
m-10--5
*-15--10

Figure 4-7. Color scale for dBm color coding for all of the signal strength graphs.

Results

RF Mapping of Non-Metal Sided Tote

Figure 4-8 demonstrates the RF power distribution inside a regular tote with no metal

coating. Even though a null spot is not detected anywhere, the signal strength at the bottom of

the tote drops to as low as -10 to -5 dBm. These and their surrounding zones lower than 0 dBm

can be considered as critical areas which have a greater possibility of creating a zone where

reading a tag may not occur when the tote is filled with products. The low power at the bottom

of the tote, the places close to the antenna, can be ascribed to the antenna radiation pattern. In a

typical UHF antenna, the waves are radiated in a cone like shape which has narrow radiation

beam close to the antenna causing more null spots or low power levels in these areas as it is also

verified by the results of non-metal sided tote.

The maximum power that is observed in all layers is between 10 to 15 dBm and the

dominant power level is between 5 to 10 dBm.

















m 35-40
A 30-35
m 25-30
m 20-25
0 15-20
m 10-15
E 5-10
S0-5
0 -5-0

D m-10--5
E -15--10


bottom
7 6 5 4 3 2 1



a) Layer 1



top


_-A m3540
30-35
m 25-30
-B 20-25
0 15-20
S10-15
C 0 5-10
D 0-5
S-5-0
-D
D -10-5

m -15-10


7 6 5 4 3 2 1


b) Layer 2


Figure 4-8. Mapping results of 0 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.

















S6 5 4 m 35-40
m 30-35
m 25-30
B 20-25
m 15-20
m 10-15
C 5-10
E 0-5
N-5-0

D 0 -10-5
m -15-10


bottom
7 6 5 4 3 2 1



c) Layer 3




top




F 3540
0 30-35
25-30
B 20-25
0 15-20
0 10-15
C O 5-10
0 0-5
S-5-0

D -10-5
0 -15-10


bottom
7 6 5 4 3 2 1



d) Layer 4


Figure 4-8. Mapping results of 0 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d)
Layer 4. The antenna was placed 50 mm away from the bottom of the tote.












RF Mapping of 2 Metal Sided Tote

According to Figure 4-9 in the tote with 2 metal coated sides, the same data points have


the lowest powers but higher values within a range of -5 to 0. This clearly shows that the


distribution of RF waves is changed by the waves reflected from the metal sides. However there


is no significant difference in the average overall power by having a layer, layer 4, which has a


maximum power between 5-10 dBm.

Top
0 3540
0 30-35
0 25-30
0 20-25
0 15-20
B 10-15
O 5-10
O O -5
-C
I -5-0
0 -10-5
D -15-10
Insulator
border
bottom
7 6 5 4 3 2 1


a) Layer 1


I 35-40
0 30-35
A 25-30
0 20-25
S15-20
IB 10-15
O 5-10
o00-5
-C
0 -5-0
0 -10-5
D 0-15-10

Insulator
border
bottom
7 6 5 4 3 2 1

b) Layer 2


Figure 4-9. Mapping results of 2 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.

















S35-40
SI /3A
E A .30-35
25-30
20-25
I 15-20
*K- -* -- -B
0 10-15
0 5-10
r 0-5
__--C --5-0
-10-5
0 -15-10
-D ----Insulator
border


bottom
7 6 5 4 3 2 1

c) Layer 3




top

S3540
--__ A 330-35
S25-30
20-25
15-20






0 -15-10

D __ Insulator
border


bottom
7 6 5 4 3 2 1



d) Layer 4



Figure 4-9. Mapping results of 2 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d)
Layer 4. The antenna was placed 50 mm away from the bottom of the tote.












RF Mapping of 3 Metal Sided Tote

The results of the tote with 3 metal coated sides are shown in Figure 4-10. Notice that the


average power has significantly increased. Inside the free space of the tote, the lowest points are


located only in a small area, the right top covers of layer 3 and layer 4, with values between -10


dBm and -5 dBm. Moreover the maximum power is higher than the previous totes with a value


between 15-20 dBm instead of 10-15 dBm.

top
F 35-40
1 30-35
A 25-30
: 20-25
m 15-20
B 010-15
o 5-10
0 0-5
C m-5-
0 -10-5
0 -15-10

-Insulator
border

bottom
7 6 5 4 3 2 1

a) Layer 1


top

3540
A m30-35
0 25-30
0 20-25
B 015-20
m 10-15
O 5-10
C 0-5
0-5-0
m -10-5
D -15-10

-Insulator
border
bottom
7 6 5 4 3 2 1


b) Layer 2


Figure 4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.





































7 6 5 4 3 2 1


c) Layer 3



top

..... -5-40
A 30-35
0 25-30
M 20-25
B 15-20
0 10-15
O 5-10
E 0-5
m-5-0
-10-5
0 -15-10
D
-Insulator
border

bottom
7 6 5 4 3 2 1


d) Layer 4


Figure 4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.












RF Mapping of 4 Metal Sided Tote


As illustrated in Figure 4-11, the tote with 4 metal coated sides has similar power levels


on average with the 3 metal sided tote. However for the first time a null spot is observed inside a


tote at the left bottom corner of layer 4. As in the non-coated and 2 metal sided tote, the lowest


power is distributed at the bottom of the tote inside the free space of the tote. This indicates that


the sides across each other created a multi path where the waves canceled each other. Due to


having a slight angle on the sides this effect is not observed for the top areas.


top
0 35-40
30-35
A L 25-r30
0 20-25
15-20
B 10-15
O 5-10
0 O-5
C 0-50
S-10--5
D -15--10
-D
F r -Insulator
border
bottom
7 6 5 4 3 2 1

a) Layer 1

top

S35-40
A 30-35
M 25-30
m 20-25
M 15-20
B m 10-15
o 5-10
E 0-5
C M-5-0
m -10-5
M -15-10
D
-Insulator
border
bottom
7 6 5 4 3 2 1

b) Layer 2


Figure 4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.




















S35-40
A 30-35
: 25-30
0 20-25
15-20
B .10-15
D 5-10
S0-5
C -5-0
-10-5
0 -15-10


-Insulator
border

bottom
7 6 5 4 3 2 1

c) Layer 3




top
m3540
m 30-35
A m 25-30
m 20-25
S15-20
B 10-15
E 510
E 0-5
C m -5-0
m-10-5
m -15-10

D --Insulator
border


bottom
7 6 5 4 3 2 1


d) Layer 4



Figure 4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.












RF Mapping of 5 Metal Sided Tote


Figure 4-12 shows that the tote with 5 metal coated sides has the best RF power levels of


all the totes. It has more uniform wave distribution along the edges as well as a higher average


power level. The lowest signal strength which is only observed at one point is between 0-5 dBm


inside the free space of the tote. Moreover the highest power is between 35-40 dBm which is


almost 3 times more than the highest value measured in the other configurations.


7 6 5 4 3 2 1

a) Layer 1


S35-40
A 30-35
S25-30
S20-25
B *15-20
10-15
O 5-10
0 O0-5
-5-0
S-10-5
S-15-10
-D
-Insulator
border
bottom





top

S35-40
A 30-35
A
25-30
20-25
S15-20
B
10-15
S5-10
0O-5
S *-5-0
S-10-5
S-15-10
D
-Insulator
border


bottom


7 6 5 4 3 2 1

b) Layer 2


Figure 4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.


M-MONM










































7 6 5 4 3 2 1


c) Layer 3


7 6 5 4 3 2 1


top

S35-40
A 30-35
25-30
20-25
B 15-20
-B
10-15
0 5-10
S0-5
c -5-0
-10-5
-15-10
D
-Insulator
border

bottom


d) Layer 4


Figure 4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and
d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.


m 3540
m 30-35
A
m 25-30
S20-25
S15-20
B *10-15
S5-10
0-5
C -5-0
-10-5
S-15-10
D
-Insulator
border

bottom









Discussions


Behavior of Reflected RF Waves

This part of the project face with two very important problems that surround the

propagation of RF waves inside closed environments. First, the occurrence of destructive

interference of the waves by metal surfaces, which can be observed from the results of tote # 4,

explains how RF waves cancel each other while having multiple paths.

Second, the path loss occurrence causes an attenuation of the RF waves. As it is known,

the energy of the waves decreases proportionally to the inverse square of the distance6' 47, 48 that

they travel. Hence the power loss will be greater inside metal coated totes since the waves will

travel more due to the reflection in the path of RF waves.

However in this chapter, it was shown that both these problems can be addressed by

properly guiding the RF waves with reflective materials. For example, in the example of path

loss, which is simple unavoidable, the average RF wave power level in a region can still be kept

the same by the joined force of multiple waves that are reflected from different angles instead of

a single wave. This is observed in all the totes with metal surface, that the average RF wave

power level was even increased compared to the control tote. Moreover the lowest power

regions remain similar in every tote, but depending on the metal surface configuration they have

different signal strengths, which is another example of the joined force.

In the tote designs, it is found that the top metal surface plays a greater role in increasing

RF power when the results of tote #3 and #5, which has the highest measure power respectively,

compared to the others.

Furthermore the selection of antenna type based on polarization also plays an important

role in the distribution of RF waves. Notice that the RF power distribution is not symmetric in

any totes due to the use of circularly polarized antenna. This type of polarization enables to









decrease the probability of destructive interference occurrence of the waves reflected from the

metal surfaces across each other in the totes #4 and #5.

Insulation

The other interesting observation from the results of these tests is the power measured

near the metal surface in the case of layer # 4 tote # 5. Here, the high power zone affect and

strengthen the field around the metal surface adjacent to these regions. More power level may

turn on the tag and backscatter the signal which will enable the tag to be read even when it is in

close proximity to a metal surface. This may lead us to use thinner insulator which can be

studied as a future work.

Security Issues

One of the main drives for this research is to keep the order safe by sealing the tote to

prevent loss in any step of the distribution. The other security issue can be considered as keeping

the order, i.e., the contents of the tote, undisclosed. One of the drawbacks of an RFID system is

that it raises security and privacy concerns by making "unprotected tags vulnerable to

eavesdropping, traffic analysis, spoofing or denial of service" as defined by Weis S.A. and

others.49 Among them spoofing is a real threat to the supply chain systems where an

unauthorized reader can scan, change or copy the information of the tag. For example in the case

of a sealed pharmaceutical tote, someone can single out and counterfeit only the totes that

contain valuable or special products which the counterfeiter is interested in.

In order to preclude the disclosure of the totes' contents, the readability of the tags inside

the tote by unauthorized readers should be prevented. The results of this chapter showed that the

metal can also be used as a shield against unauthorized readers by preventing RF waves passing

through it. For instance, if 5 metal sided totes are stacked in a pallet on a metal plate, a complete

protection against unauthorized readers is achieved. Same effect can be ensured for 4 metal









sided totes when top and bottom metal plates are used on a pallet. For 3 metal sided totes, a

special configuration of stacking is needed where the metal sides of half of the totes are reversed

and the totes are put in an order to cover all possible intrusion sides.

Summary

In this chapter the metal effect on the distribution of electromagnetic RF waves inside a

closed environment was evaluated by RF mapping of totes with different levels of reflective

material coatings. It is always claimed that the metal causes attenuation in the UHF signal by

creating multi paths and interference of the waves; however this phenomenon can be controlled

and even be used as an opportunity to increase RF power level by proper use of metal. Tote

number 5 is a good example, where all the sides except the bottom are coated with metal. The

highest measured RF power in this tote was almost 3 times higher than the control tote (non-

metal sided tote).

Metal can also be utilized as a barrier against RF waves. When tote #4 and #5 palletized

a full protection against unauthorized readers can easily be achieved. This prevents disclosure of

the totes' contents providing more security.

Finally, better distribution of RF waves via use of reflective material may increase the

probability of reading tags, which will be explored and evaluated in the next chapter.









CHAPTER 5
EVALUATION OF READABILITY CONTENTS OF PURCHASED ORDERS

Introduction

The contents of the totes used for distribution of pharmaceutical products vary in size and

material. The random orientations and mixing of products within the totes make it probable that

any liquids and metals that are part of one product or its packaging will block the path of RF

waves thus negatively impacting the readability of the totes' contents. The results of Chapter 4

showed that the metal can be utilized to increase the RF power inside a tote. This chapter will

investigate how the use of metal in a tote design affects the readability of the tote.

The negative effects of metal on the tag antenna, such as detuning, are well known. For

instance, when the tag is placed too close to the metal, the tag might not be read due to the

changes in the antenna's radiation pattern, input impedance, radiation efficiency or resonant

frequency.50 To minimize these negative impacts the tag should be separated from the metal. A

good example is the degradation of antenna radiation in metal products being much higher than

the ones with liquid or any other materials, which requires the tag to have better insulation from

the product itself.5 Moreover the size and the shape of the metallic object also affect the

radiation pattern of the antenna.52

However, metal can also be used in ways to actually improve the performance of an

RFID system. Applications, such as using a metallic surface as a reflector or the ground plane of

an antenna to harvest more energy for the tag, are becoming more common.53' 54, 55 Researchers

have recently begun investigating the possible use of the metal package material itself as an

antenna.56

This chapter explores a novel constructive use of metal to generate multiple paths via

reflected waves, potentially creating a viable communication path to a tag that would otherwise









be shielded by the contents of the tote. This is accomplished by developing a modified tote to

improve the readability of its contents with a UHF system by using reflective materials inside.

Different modified totes were evaluated with a set of representative POs, where each PO is

offered 1, 2 and 3 times of read attempts.

Materials and Methods

Tote Design

The tote is designed based on the work of the previous chapter that suggests the metal can

be utilized to improve readability if better distribution of reflected RF waves is achieved. The

first design was to use a full metal cover on the 5 sides leaving the bottom open to enable the

waves entering through there since during the RF mapping analysis this tote showed the best RF

wave propagation inside.

However, inside 5 metal sided tote the tags placed on the 10 mm insulator, used in the

previous chapter, from the metal surface could not be read due to metal feature of detuning the

antenna. The thickness of the insulator was increase by 5 mm starting from 10 mm and it was

found out that a minimum insulation of 25 mm between the metal side and the tag was required

to read the tag when the tote is closed. According to researches the bigger the size of the metal

the more separation is required between the tag antenna and the metal surface.45 To decrease the

insulator thickness for more available free space for the products, the concentration of the metal

was decreased by using metal grid texture on the surfaces that has 10 mm gap between each 10

mm metal stripes instead of a full metal cover shown in Figure 5-1. Different thickness levels

for insulators starting from 25 mm were tried and the minimum required thickness was found to

be 10 mm. This texture also provided a similar effect with the full cover by still reflecting the

RF waves without letting them go through the surfaces.









Along with the 5 metal sided tote, the 3 metal sided tote with metal grid texture was also

tested since it had the second best RF propagation to evaluate and better understand the usage of

reflective material at improving the read rate.
























Figure 5-1. 5 metal sided tote with a grid pattern texture. The bottom side is not covered with
metal surface.

Preparation of Representative Purchase Orders

The POs which only represents the worse case scenarios were prepared to lower the

sample number that needs to be tested and to observe if there is a significant difference between

the modified totes. A total of 100 POs was categorized into 4 groups which were controlled PO,

PO with only one type of product, medium load PO and full load PO. The distribution of 100

POs based on their categories was; 10 controlled, 10 with one type of product, 30 with medium

load and 50 with full load POs.

The detailed information about the products and pictures are available in the appendix

section.









Controlled PO

To prepare the controlled POs first the products that had the worst read rates in turn table

tests in Chapter 3 were chosen. Then, for every product both HF and UHF results were assessed

together within orientations and t test was conducted to find the differences between the

orientations. If there were differences, t test with a significant level (a=0.05) is performed to

rank the read rates. The results of statistical analysis for every product that are used in this type

of PO are shown in Table 5-1.

Table 5-1. Statistical analysis of read rates of products on turn table*
0 45 90 135 180
Equate Flu Relief 76.50 100 100 100 100
B A A A A
Magnesium Citrate Oral 0 7.00 50.00 50.00 0
Solution B B A A B
Pedialyte Freezer Pops 0 50 100 100 100
C B A A A
OCuSoft 46.18 47.50 66.17 100 100
A A A
B B B
MGP Bromaxefed DM 100 48.00 50.00 100 100
RF Syrup A B B A A
MGP Bromaxefed DM 16.33 0 0 0 0
RF Syrup 1 Pint A B B B B
Lovenox 50.00 50.00 100 100 100
B B A A A
NuvaRing 100 95.00 100 100 50.00
A A A A B
Procanbid 50.00 50.00 100 100 100
B B A A A
Xylocaine-MPF 50.00 50.00 100 100 100
B B A A A
Equate Triacting Cold 50.00 50.00 100 100 100
& Cough B B A A A
Hurricaine Tropical 50.00 50.00 94.00 100 100
Anesthetic Spray B B A A A
Tinactin Antifungal 50.00 50.00 100 100 100
Spray Can B B A A A

*Same letter represents no difference between the orientations.









Finally the worse orientations for each product were selected. If there was more than

one, each one was assigned a number which was then selected randomly. 10 different PO was

created out of 13 different products that were in their worse orientations. A detailed worksheet

for this study can be found in Appendix C.

PO With One Type of Product

The tote was filled completely with equate effervescent pain relief packages which

contain the tablets inside aluminum patches. This product is chosen instead of any blister

packages to create more challenge during scanning. As shown in Figure 5-2 there is no air gap

between each aluminum patches which do not leave any space for the waves to pass through it.

The only way to reach a tag in the middle is by utilizing the air gap between each product which

can be blocked easily by the patches with even a small movement in the tote.

The packages were arranged inside the tote such that they were all positioned in the same

orientation as if they were in a case. 10 different POs were created out of 28 or 32 products

depending on the case size that fits inside the tote.


















Figure 5-2. Content of the product, equate effervescent pain relief, used in the preparation of one
type of product POs.









Medium Load PO

Medium loaded tote contained 24 products that have 50% RF opaque and 50% RF

absorbent products. The samples were prepared by taking all of the products outside and placing

them again randomly inside the tote and each sample scanned 3 times without changing any

placement of the products. However the products with liquid were first put inside before other

products. This type of placement is also a common practice in real world applications as it is

confirmed by McKesson, one of the largest wholesalers.

30 different POs were tested in this category.

Full Load PO

The totes with full load POs had 48 products, 75% of which were RF opaque (18) and RF

absorbent (18). More samples (50) were prepared for this type simply due to having a higher

population number by the permutation of random orientations for more products. Similar to

medium load POs, for every sample, all the products were taken outside and placed back inside

the tote randomly. The placement order of the products is first the RF absorbent, then the RF

lucent and finally the RF opaque products. The products with blister packs were put in such a

way that the metal surface would be facing to top.

Testing Procedure

The same types of tags that are used in chapter 3 for the UHF setup were applied onto

products before the POs were prepared. Each PO was tested 3 times for the validity of statistical

analysis by using the same speed level (0.17 m/s) on a conveyor belt within a regular tote (non

metal sided tote), 3 metal sided tote, and 5 metal sided tote. The content of the totes were

scanned with one Symbol UHF circularly polarized antenna placed under the conveyor belt and

Symbol XR440 reader.









Data Analysis

During the analysis two-way analysis of variance (AOV) and t test statistical methods

with a 5% significance level were performed for the evaluation of percentages of read rates. In

addition, the probability of getting 100% read rate with regular tote and the modified totes were

compared. To simulate the real time process the data was also analyzed as if the totes that were

not 100% read were given a 2nd and 3rd time of read attempt. Hence the performances of

different tote designs at reading 100 totes were evaluated into 3 groups:

a) The totes only offered 1 read attempt
b) The totes that were not read entirely offered a 2nd time read attempt
c) The totes that were not read entirely offered a 3rd time read attempt

For the 2nd and 3rd trials, instead of scanning each PO 9 times (3 times for each 1st, 2nd

and 3rd trials) 3 times of reading were conducted without writing over each other. Then,

according to trial version, the 6 combinations out of 3 scans were evaluated as if the data were

written over each other. The read rates were averaged and the probability of getting 100% read

rate calculated along with the time spent on reading that tote. For example; a full tote analysis in

a group b was performed as in Table 5-2. Similar calculations were conducted for group c totes;

however the 6 combinations were abc, acb, bac, bca, cab, cba (Table 5-3.)

Moreover, the evaluation of the other trials provided us with a means to compare the

average times of reading 100 POs within different types of totes.









Table 5-2. Example of a calculation of read performance of a tote if the totes offered 2nd time
read attempt*

Read rates of samples out of 48 products for a PO: a= 47 (13) b= 47 (20) c=48

Sample order Read rate out of 48 Time spent on probability of getting
scanning 100% read rate

ab 47 2 0
ac 48 2 100
ba 47 2 0
bc 48 2 100
ca 48 1 100
cb 48 1 100
Average 47.66 = 99.31% 10/6 = 1.66 66.66%

*The read rates of same tote samples are represented by a, b and c. The numbers next to them
are the numbers of products that are read inside the tote and the numbers in the parentheses are
the product ID's.


Table 5-3. Example of a calculation of read performance of a tote if the totes offered 3rd time
read attempt*
Read rates of samples out of 48 products for a PO: a= 46 (18,27) b= 47 (18)
c= 47 (6)

Sample order Read rate out of 48 Time spent on probability of getting
scanning 100% read rate

abc 48 3 100
acb 48 2 100
bac 48 3 100
bca 48 2 100
cab 48 2 100
cba 48 2 100
Average 48 = 100% 14/6 = 2.33 100%

*The read rates of same tote samples are represented by a, b and c. The numbers next to them
are the numbers of products that are read inside the tote and the numbers in the parentheses are
the product ID's.









Results


One Read Attempt

Table 5-4 indicates the data of reading 100 POs without offering another read attempt.

According to the results the average read rates of a total of 100 different POs within totes have

statistical differences (P= 3.25E-05 <0.05). However the regular tote and the 5 metal sided tote

have similar values of 97.610.52 and 97.770.45 respectively with no significant difference.

Even though the 5 metal sided tote has better performances than the regular tote in most of the

PO types, its performance degrades significantly for full totes. On the contrary, the 3 metal sided

tote has an almost 1% better read rate than the 5 metal sided tote and more than 1% better read

rate than the regular tote with a value of 98.760.34. It has the highest performance not only in

the average read rate but also in every PO type with a smaller variance (Figure 5-3).

Furthermore with the 3 metal sided tote 18% more totes than the regular tote and 15%

more totes than the 5 metal sided tote were read entirely which serves as the real indicator of the

dramatic improvement in performance.

Table 5-4. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal
sided tote offered 1 read attempt*
PO Type # of Read Rates Probability of getting
Samples 100% read rate
3 metal 5 metal 3 metal 5 metal
regular sided sided regular sided sided
control 10 98.46 99.49 99.74 8.00 9.33 9.66

one type 10 97.19 97.81 97.81 6.00 6.33 6.66
medium
loaded 30 96.66 99.17 97.18 12.00 24.00 13.33
full
loaded 50 98.08 98.56 97.72 21.00 24.33 19.33
Total 97.61 98.76 97.77
Results 100 B A B 46.00 64.00 49.00
Margin of
Error 0.52 0.34 0.45
*Same letter represents no difference between the totes.










101-

99- -

97-
0)
.2 95-

| 93-

91-

89-

87
3 side 1 5 sided 1 regular 1 Each Pair
Student's t
Column 1 0.05

Figure 5-3. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 1 read attempt. Circles separated each other represents significant
differences between the totes.

Second Time Read Attempt

Table 5-5 shows that there is significant differences between totes at reading 100 POs

that are offered a 2nd time read attempt if 100% read is not achieved (P= 6.2E-04<0.05). Even

tough there is an increase in the performance of 5 metal sided tote by having a better read rate at

reading full tote, no significant differences was observed with a regular tote. However the 3

metal sided tote still has the highest read rate on average, 99.26+0.26, with the least variance as

illustrated in Figure 5-4. For the probability of getting 100% read rate, the 3 metal sided tote

shows a superior performance with reading 75 POs completely out of 100 whereas the 5 metal

sided and regular totes reads 66 POs and 58 POs, respectively (Table 5-5). Moreover the 3 metal

sided tote requires less time than the other totes, as much as 11.08% less than the regular tote and

9.71% less than the 5 metal sided tote.










Table 5-5. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal
sided tote offered 2nd time read attempt*

PO # ofn Read Rates Probability of getting Time of Read
Type 100% read rate

3 5 3 5
metal metal metal metal regular/ 3 metal 5 metal
regular sided sided regular sided sided 6 sided/6 sided/6
control 10 99.23 99.74 100 90.00 96.66 100 12.00 11.00 10.33
one
type 10 98.33 98.85 99.38 70.00 73.33 83.33 14.00 13.66 13.00
mediu
m
loaded 30 97.80 99.49 98.24 55.56 87.78 63.33 48.33 36.00 47.00
full
loaded 50 98.57 99.11 98.58 50.66 63.33 57.33 79.00 75.66 80.66
Total 98.38 99.26 98.70
Results 100 B A A 58.00 75.00 66.00 153.33 136.33 151.00
Margin
of Error 0.45 0.26 0.37

*Same letter represents no difference between the totes.


101
100-
99- -
98-
97-
96-
95-
94-
93-
92-


3 side 1


5 sided 1


regular 1


Column 1


Each Pair
Student's t
0.05


Figure 5-4. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 2nd time read attempt. Circles separated each other represents significant
differences between the totes.









Third Time Read Attempt

The results of read rates if the totes are offered a 3rd time read attempt is demonstrated in

Table 5-6. The average read rate with the 5 metal sided tote has continued to increase to reach a

level that eliminates the significant difference between the 3 metal sided tote, whereas the

regular tote continues to have the lowest read rate (P=1.5E-03<0.05). Figure 5-5 shows the

distribution of read rates according to tote type. When it comes to reading all the products, the 3

metal sided tote has 80% POs with 100% read rate whereas the 5 metal sided tote has 75% POs

and the regular tote has only 62% POs. More importantly the 3 metal sided tote required 13.77%

and 18.99% less time than the 5 metal sided and the regular tote, respectively, to achieve this

performance.

Table 5-6. Statistical analysis of read performances of 100 POs with regular, 3, and 5 metal
sided tote offered 3rd time read attempt*

PO Type # of Read Rates Probability of getting Time of Read
n 100% read rate

3 5 3 5
metal metal metal metal regular/ 3 metal 5 metal
regular sided sided regular sided sided 6 sided/6 sided/6

control 10 99.23 100 100 90.00 100 100 13.00 11.00 10.33

one type 10 98.75 99.06 99.69 80.00 80.00 90.00 16.33 98.00 15.00
medium
loaded 30 98.19 99.58 98.89 60.00 90.00 73.33 62.66 39.33 58.66
full
loaded 50 98.75 99.29 98.86 54.00 70.00 67.35 105.66 94.00 102.33
Total
Results 100 98.63 99.43 99.37 62.00 80.00 74.67 198.33 160.66 186.33
B A A
Margin
of Error 0.43 0.25 0.34

*Same letter represents no difference between the totes.










101
100 I
99-
98-
97-
96-
95-
-o 9
a 94-
93-
92-
91-
90-
89-
3 side 1 5 sided 1 regular 1 Each Pair
Student's t
Column 1 0.05

Figure 5-5. Distribution of the means of read rates of 100 POs with regular, 3, and 5 metal sided
tote offered 3rd time read attempt. Circles separated each other represents significant
differences between the totes.

Discussions

Placement Order of the Products in the Medium and Full Loaded Totes

As described in the methods and materials sections the products are placed in a specific

order inside the medium and full loaded totes. Main reasons for such a placement include both

physical constraints and the aim to improve the read rates. As an example of a physical

constraint, the liquid products are usually heavier; thus can damage other products if they are

placed on top. Moreover, glass is a common package material for the liquid products and

placing them at the bottom makes them more stable and offers more protection from breakage.

Regarding issues with read rates, liquid products tagged with near field tags should be placed

closer to the antenna as they have a short read range. Placing them at the bottom ensures that

they are closer to the antenna.

In the case of the blister packs, they are placed last with their metal surfaces facing the

top. This placement prevents adjacent product tags from getting detuned by the metal of the









blister packs since the tags below will have enough air gaps and in almost all the blister packs the

tags are located on the side. Finally, please note that this particular placement described above is

easily possible with automation during product pick-up in the warehouse.

Performance Differences of Modified Totes

The results of the mapping tests, Chapter 4, are validated in this section. By using an

aluminum coating on the sides of the tote, the RF wave propagation was guided in such a way

that it reaches everywhere inside the tote by creating more waves resulting in more RF power

coming from different directions. Hence, the 5 metal sided tote performed better at reading

controlled and one type product POs than all the other totes. However, when the tote is filled

completely with the products, its performance has dropped substantially. One of the reasons for

this observation is due to the fact that in the regular tote the waves emitted outside of the tote is

also utilized to read the products placed closer to the edges on the top areas. Conversely, in

metal coated totes, these waves are blocked by the metal surfaces and the sides' convex angel

limits the waves coming from inside to those areas. Via the same reasoning, the 3 metal sided

tote has a better performance by utilizing both the outside waves and the reflected waves inside.

On top of everything it should always be considered that all the results are affected by the

tote movement on the conveyor belt, as the wave pattern inside the tote changes along with this

motion. Even a slight change within each read attempt, such as the location of the tote on the

conveyer belt, cause completely different wave patterns to occur. 5 metal sided tote is affected

most by this phenomenon simply because it has more metal surfaces and thus more reflections

within the tote. Therefore it displays a high variance in overall performance due to these random

wave patterns, which result in a non-linear (x2) increase in its performance, whereas other totes

display a linear increase.









To improve the readability performance of the 5 metal sided tote, there are multiple

solutions such as using a wider bottom surface with more perpendicular sides. Moreover,

coating only half the side surfaces with metal instead of the whole surface might result in better

performance as well. This is mainly because, in the case of a full tote, the top half of the side

metal surfaces are not being used effectively (the reasons described above) and in the case of a

small loaded tote waves reflecting from there do not contribute in reading as much.

Other Application Methods of Metal Coating

The metal coating can be applied to the totes in different ways as they can either be part

of the insulator or the tote. For example, metal can be embedded to the plastic of the tote during

injection molding or metal spray paint can be applied onto the plastic after the production of the

tote. Hence, metal layer can be put inside or outside of the tote material; however the layer will

be more protected if it is placed inside.

Efficiencies of 2nd and 3d Read Attempts from the Warehouse Perspective

The results of this chapter can be used by the distributor on how to decide the number of

read attempts that should be conducted for the totes. According to the results for the 3 metal

sided tote, 30.60% and 20.00% of the remaining totes can be read entirely in the 2nd and 3rd

attempt, respectively. For the 5 metal sided tote these numbers change to 33.30% in the 2nd

attempt and 25.50% in the 3rd attempt. However in the regular tote fewer totes can be read

22.20% in the 2nd attempt and only 9.50%in the 3rd attempt.

When time consumption (given in the results section of this chapter) and the percentages

above are both taken into consideration, it is observed that using the conveyor belt to scan the

tote for the 3rd attempt is not very efficient for the warehouses. Depending on the order size, it is

better to do the scanning manually for the 2nd or 3rd attempts. For example, according to the









results, in the case of the 3 metal sided tote, only 1 or 2 products are missed when the read rate is

less than 100%, which can be easily read in a manual scan.

Summary

In this chapter, a regular tote, a 3 metal sided tote and a 5 metal sided tote were compared

at scanning 100 different POs with UHF system on a conveyor belt. Most challenging POs with

different sizes were created. The metal coating of the tote was textured with a grid like pattern,

which enabled to use thinner insulator. The data is analyzed as if the totes were offered up to 3

read attempts.

As a conclusion there were no significant differences in the average read rates between

the regular tote and the 5 metal sided tote at all attempts. However, 3 metal sided tote had

always the best performance in all totes by having highest read rates and reading more POs

entirely within a better time range. Finally the distributor is suggested to attempt up to 2nd read

attempt to the totes not read 100% on the conveyor belt for more time efficiency.









CHAPTER 6
CONCLUSION

Pharmaceutical distribution chain is very vulnerable to counterfeiting which is observed

very often in this industry. The main objective of this research is to improve the security of and

readability of the RFID tagged pharmaceutical products and food supplements inside a sealed

tote by proper designs of the tote and portal.

Several tests were performed to analyze different aspects of an RFID system as well as to

gauge the improvement introduced by novel contributions to the tote and portal design. The first

test compared the performances of UHF and HF systems and showed that the UHF RFID system

is more suitable to be used in item level of pharmaceutical applications than the HF system with

its higher performance of reading products on a high speed operating conveyor belt, on a turn

table and inside a tote.

The second phase of the research found out that the use of reflective materials as part of

the sealed tote provides a control over the RF propagation which leads to a better RF power

distribution by RF mapping of 0, 2, 3, 4, and 5 metal sided totes. The 5 metal sided tote had the

highest RF power level which was almost 3 times higher than the regular (0 metal sided) tote.

This chapter also indicates that metal can be used to enhance the security of the tote by

preventing disclosure of the RFID tagged products inside a tote against unauthorized readers.

The 3, 4 and 5 metal sided totes can be palletized in a special configuration which prevents RF

waves reaching to the totes' contents.

The final test includes the scanning of different types of POs with modified totes along

with a regular tote (control unit). The results show that the 3 metal sided tote with a grid texture

on the metal surfaces has improved readability of products inside the sealed tote by having

higher average read rates and more 100% read POs with the least time consumption.









Even though the 3 metal sided tote has the highest read rates, its shielding effect against

unauthorized readers is achievable but not as efficient as the 5 metal sided tote. However, if the

5 metal sided tote is to be used due to offering more protection, the read rates can be improved

further by using wider bottoms or sides with different angles.

To sum up, this thesis proves that a security and a readability of tagged pharmaceutical

products and food supplements inside a sealed tote can be improved by utilizing reflective

materials in the tote design with a proper system choice and setup.









APPENDIX A
PRODUCTS USED IN THE TESTS

Lovenox


Manufacturer: Aventis Pharma Specialties, Maisons-Alfort, France
Tag place: Top, perpendicular
Dimensions: height (h)=168 mm, length (1)=140 mm, width (w)=65 mm
Product physical type: 5. a) Sub-products (syringes) in a plastic
Product RF type: RF lucent

NuvaRing


Manufacturer: N.V. Organon, Oss, The Netherlands
Tag place: Side, perpendicular
Dimensions: h=110 mm, 1=120 mm, w=19 mm
Product physical type: 5. b) Sub-products in a metal
Product RF type: RF opaque

BD Ultra-Fine III


Manufacturer: BD, Franklin Lakes, NJ, USA
Tag place: Top, perpendicular
Dimensions: h=85 mm, 1=130 mm, w=72 mm
Product physical type: 5. a) Sub-products in a plastic
Product RF type: RF lucent


Equate Chest Rub


Manufacturer: The Mentholatum Co., Inc., Orchard Park, NY, USA
Tag place: Top, parallel
Dimensions: h=75 mm, 1=55 mm, w=55 mm
Product physical type: 2.a) Cream in a plastic
Product RF type: RF lucent

Cardizem LA

Manufacturer: Biovail Corporation, Mississauga, ON, Canada
Tag place: Side, perpendicular
Dimensions: h=95 mm, 1=35 mm, w=35 mm
Product physical type: 1.a) Pill in a plastic
Product RF type: RF lucent


Li I -,,I
Ile 111,


iBD Ultra-Fine On,.n.i
Pen NE-dlee


100









Smart Health Aspirin


Distributed by: Wal-Mart Stores, Inc., Bentonville, AR, USA
Tag place: Side, parallel
Dimensions: h=90 mm, d=45 mm
Product physical type: 1.a) Pill in a plastic
Product RF type: RF lucent

Procanbid


Manufacturer: DSM Pharmaceuticals, Inc., Greenville, NC, USA
Tag place: Top, perpendicular
Dimensions: h=115 mm, 1=85 mm, w=70 mm
Product physical type: 1.b) Pill in a metal (blister)
Product RF type: RF opaque

Equate Flu Relief


Manufacturer: LNK international, inc., Hauppauge, NY, USA
Tag place: Side, perpendicular
Dimensions: h=90 mm, 1=85 mm, w=23 mm
Product physical type: 1.b) Pill in a metal (blister)
Product RF type: RF opaque

Ivax


Manufacturer: IVAX Pharmaceuticals, Inc., Miami, FL, USA
Tag place: Top, parallel
Dimensions: h=85 mm, 1=105 mm, w=105 mm
Product physical type: 1.b) Pill in a metal (blister)
Product RF type: RF opaque



Gentamicin Sulfate Injection


Manufacturer: Abbott Laboratories, North Chicago, IL, USA
Tag place: Side, parallel
Dimensions: h=40 mm, 1=90 mm, w=90 mm
Product physical type: 3.c) Liquid in a glass
Product RF type: RF absorbent









Xylocaine-MPF


Manufacturer: AstraZeneca LP, Wilmington, DE, USA
Tag place: Top, parallel
Dimensions: h=112 mm, 1=90 mm, w=40 mm
Product physical type: 3.c) Liquid in a glass
Product RF type: RF absorbent

Magnesium Citrate


Manufacturer:HUMCO, Texarkana, TX, USA
Tag place: Side, parallel
Dimensions: h=165 mm, d (diameter)=60 mm
Product physical type: 3.c) Liquid in a glass
Product RF type: RF absorbent

Diltiazem HCI Injection


Manufacturer: Ben Venue Laboratories, Inc., Bedford, OH, USA
Tag place: Top, parallel
Dimensions: h=50 mm, 1=120 mm, w=48 mm
Product physical type: 3.c) Liquid in a glass
Product RF type: RF absorbent

Pedialyte Freezer Pops


Manufacturer: Abbott Laboratories, Columbus, OH, USA
Tag place: Top, perpendicular f
Dimensions: h=240 mm, 1=90 mm, w=90 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent


OCuSoft


Manufacturer: OCuSoft, Inc., Rosenberg, TX, USA
Tag place: Side (cap), parallel
Dimensions: h=200 mm, d=50 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent
[a









MGP Bromaxefed Syrup


Manufacturer: Morton Grove Pharmaceuticals, Inc., Morton Grove, IL, USA
Tag place: Side, parallel
Dimensions: h=120 mm, d=44 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent

Tincture of Benzoin


Manufacturer: Smith & Nephew, Inc., Largo, FL
Tag place: Side, parallel
Dimensions: h=154 mm, d=42 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent

MGP Bromaxefed Syrup 1 Pint


Manufacturer: Morton Grove Pharmaceuticals, Inc., Morton Grove, IL, USA
Tag place: Front (behind the label), parallel
Dimensions: h=180 mm, 1=80 mm, w=60 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent

Equate Triacting Cold & Cough

I
Manufacturer: Perrigo, Allegan, MI, USA
Tag place: Top, perpendicular
Dimensions: h=140 mm, d=55 mm
Product physical type: 3.a) Liquid in a plastic
Product RF type: RF absorbent


Hurricaine Tropical Anesthetic Spray


Manufacturer: Beutlich Pharmaceuticals, Waukegan, IL, USA
Tag place: Side (cap), parallel
Dimensions: h=125 mm, d=34 mm
Product physical type: 3.b) Liquid in a metal
Product RF type: RF opaque









Tinactin Antifungal Spray Can


Manufacturer: Schering-Plough HealthCare Products, Inc, Memphis, TN, USA VALUE
Tag place: Top, parallel
Dimensions: h=170 mm, d= 53 mm
Product physical type: 3.b) Liquid in a metal
Product RF type: RF opaque

Albuterol Aerosol


Manufacturer: IVAX Pharmaceuticals, Ireland, Waterford, Ireland
Tag place: Top, parallel
Dimensions: h=97 mm, 1=52 mm, w=33 mm
Product physical type: 4.c) Gas in a metal
Product RF type: RF opaque

Combivent Inhalation Aerosol


Manufacturer: Boehringer Ingelheim Pharmaceuticals, --I%
Ridgefield, CT, USA .
Tag place: Top, parallel
Dimensions: h=82, 1=54 mm, w=32 mm
Product physical type: 4.c) Gas in a metal
Product RF type: RF opaque .

Equate Effervescent Pain Relief


Manufacturer: Tower Laboratories, Centerbrook, CT, USA
Tag place: Side, parallel
Dimensions: h=102 mm, 1=87 mm, w=54 mm
Product physical type: 1.b) Pill in a metal (blister)
Product RF type: RF opaque









APPENDIX B
PRODUCTS UHF TAG TYPE


Table B-1. Determination of product UHF tag type
Products Turn table testing Conveyor belt testing Tag Type
(average read rates) (average read rates)
Near/Far Near/Far Far Field
Field Far Field Field
Lovenox 100 100 3.26 2 Near/Far Field
NuvaRing 100 100 1.80 0.93 Near/Far Field
BD Ultra-Fine III 100 100 2.13 2 Near/Far Field
Equate Chest Rub 100 100 2.20 1.53 Near/Far Field
Ivax 100 100 3.06 2.86 Near/Far Field
Procanbid 100 100 2.26 3.00 Far Field
Equate Flu Relief 98.4 90.6 0.80 2.93 Far Field
Cardizem LA 91.96 100 1.00 3.33 Far Field
Smart Health 92.16 100 1.33 2.60
Aspirin Far Field
Gentamicin Sulfate 100 100 1.66 2.33 Far Field
Heparin Lock Flush 100 100 3.33 2.00 Near/Far Field
Xylocaine-MPF 100 100 1.40 1.33 Near/Far Field
Magnesium Citrate 3 0 0.86 0 Near/Far Field
Diltiazem HCI 100 100 2.53 1.26
Injection Near/Far Field
Pedialyte Freezer 68.4 80 1.33 2.33
Pops Far Field
OCuSoft 100 100 2.53 1.80 Near/Far Field
MGP Bromaxefed 79 0 1.00 0
Syrup Near/Far Field
Tincture of Benzoin 100 0 1.00 0 Near/Far Field
MGP Bromaxefed 11.52 0 1.00 0
Syrup 1 Pint Near/Far Field
Equate Triacting 100 100 2.53 1.80
Cold & Cough Near/Far Field
Hurricaine Tropical 100 100 1.00 2.06
Anesthetic Spray Far Field
Tinactin Antifungal 100 100 2.73 1.46
Spray Can Near/Far Field
Albuterol Aerosol 100 100 1.40 1.40 Near/Far Field
Combivent 100 100 2.13 1.33
Inhalation Aerosol Near/Far Field










APPENDIX C
PREPARATION OF PURCHASE ORDERS

Controlled PO
Table C-1. Worksheet of controlled PO
Product Name Orientations
Sample 1 Sample 2 Sample 3 Sample 4 Sample 5
Equate Flu Relief 0 0 0 0 0
Magnesium Citrate Oral Solution 45 0 45 0 180
Pedialyte Freezer Pops 0 0 0 0 0
OCuSoft 0 45 90 0 45
MGP Bromaxefed Syrup 90 90 45 90 45
MGP Bromaxefed Syrup 1 Pint 90 45 180 45 90
Lovenox 45 0 0 0 45
NuvaRing 180 180 180 180 180
Procanbid 0 0 45 0 45
Xylocaine-MPF 45 0 45 0 0
Equate Triacting Cold & Cough 0 45 0 0 45
Hurricane Tropical Anesthetic 0 0 45 45 0
Spray
Tinactin Antifungal Spray Can 45 0 45 0 0
Product Name Orientations
Sample 6 Sample 7 Sample 8 Sample 9 Sample
10
Equate Flu Relief 0 0 0 0 0
Magnesium Citrate Oral Solution 45 0 0 0 0
Pedialyte Freezer Pops 0 0 0 0 0
OCuSoft 0 0 45 90 45
MGP Bromaxefed Syrup 45 45 90 90 90
MGP Bromaxefed Syrup 1 Pint 135 90 90 0 180
Lovenox 0 45 45 0 45
NuvaRing 180 180 180 180 180
Procanbid 45 45 0 45 0
Xylocaine-MPF 45 0 0 45 45
Equate Triacting Cold & Cough 0 0 45 0 45
Hurricane Tropical Anesthetic 45 0 45 45 0
Spray
Tinactin Antifungal Spray Can 0 0 0 0 45









Pictures of Controlled PO Samples


Sample 1


Sample 2


Sample 3









Sample 4


Sample 5


Sample 6









Sample 7


Sample 8


Sample 9









Sample 10


Pictures of One Type of Product PO Samples
Sample 1















Sample 2









Sample 3


10 different samples were created for this type of PO. Above are only the pictures of 3 samples.
The other samples are just the different placement configuration of the same products as they are
in a case.


Picture of Medium Load PO Sample

30 different samples were created for this type of PO by taking out each products and placing
them randomly inside the tote. The RF absorbent products were put before the others. Below is a
picture of one sample.

Sample









Picture of Full Load PO Sample


50 different samples were created for this type of PO by taking out each products and placing
them randomly inside the tote. The placement order of the products is first the RF absorbent,
then the RF lucent and finally the RF opaque products.

Sample









LIST OF REFERENCES


1 FDA. 2005. Radiofrequency identification technology. FDA Consumer 39(2):37.

2 Charatan F. 2001. Fake prescription drugs are flooding the United States. Br Med J
322:1446.

3 Whiting R. 2004. Drugmakers 'jumpstart' RFID tagging of bottles. Information Week
999:28.

4 Dobkin DM, Weigand SM. 2005. Environmental effects on RFID tag antennas. IEEE
MTT-S International Microwave Symposium 1:135-138.

5 Lee YM, Cheng F, Leung YT. 2004. Exploring the impact of RFID on supply chain
dynamics. Proceedings of Simulation Conference 2:1145-1152.

6 Finkenzeller K. 2003. RFID handbook, fundamentals and applications in contactless
smart cards and identification. New York: Wiley. p 43-56.

7 Texas Instrument 2007. Texas Instruments LF reader family
http://www.ti.com/rfid/docs/manuals/appNotes/szzt012.pdf

8 Magellan Technology 2005. Comparison between 125 kHz low frequency (LF) & 13.56
MHz high frequency (HF). White Paper.

9 Magellan Technology. 2006. A comparison of RFID frequencies and protocols. White
Paper.

10 Chawla V, Ha DS. 2007. An overview of passive RFID. IEEE Applications and Practice
45(9):11-17.

11 ADT/Tyco Fire & Security, Alien Technologies, Impinj Inc., Intel Corporation, Symbol
Technologies Inc., Xterprise. 2006. RFID and UHF: A prescription for RFID success in
the pharmaceutical industry. White Paper.

12 Desmons D. 2006. UHF Gen 2 for item-level tagging. Proceedings from the 2006 RFID
World Meeting, Dallas, Texas.

13 Flores JLM, Srikant SS, Sareen B, Vagga A. 2005. Performance of RFID tags in near and
far field. IEEE International Conference on Personal Wireless Communications 1:353-
357.

14 Dobkin DM. 2007. RFID basics: antenna polarization.
http://www.rfdesignline.com/howto/202404293 ;jsessionid=RTGY5NAOM 1DNIQSNDL
OSKHSCJUNN2JVN?pgno=2

15 Kraus JD. 1988. Antennas. New York: McGraw-Hill. p 70-73









16 Laran RFID. 2004. A basic introduction to RFID technology and its use in the supply
chain. White Paper.

17 Chau TC, Welt BA, Eisentadt WR. 2006. Analysis and characterization of transponder
antennae for radio frequency identification (RFID) systems. Packaging Technology
Science 19:33-44.

18 Clarke RH, Twede D, Tazelaar JF, Boyer KK. 2006. Radio frequency identification
(RFID) performance: the effect of tag orientation and package contents. Packaging
Technology Science 19:45-54.

19 Sarma S, Engels DW. 2003. On the future of RFID tags and protocols. White Paper,
Auto-ID Center, Massachusetts Institute of Technology.

20 Clampitt HG. 2006. RFID certification textbook. Houston: PWD Group. p 103-120.

21 EPCglobal. 2007. Tag class definitions.
http://www.epcglobalinc.org/standards/TagClassDefinitions 1 0-whitepaper-
20071101.pdf

22 FDA. 2001. Prescription Drug Marketing Act Report to Congress.
http://www.fda.gov/oc/pdma/report2001/

23 WHO. 2006. Counterfeit medicines.
http://www.who.int/mediacentre/factsheets/fs275/en/

24 Messplay GC, Heisey C. 2006. Pharmaceutical pedigree requirements. Contract Pharma
July/August: 18-20.

25 Koh R, Schuster EW, Chackrabarti I, Bellman A. 2003. Securing the pharmaceutical
supply chain. White Paper.

26. Wilding R, Degado T 2004. RFID applications within the supply chain. Supply Chain
Practice 6(2):30-43.

27. Michael K, McCathie L. 2005. The pros and cons of RFID in supply chain management.
IEEE Proceedings of the International Conference on Mobile Business 1:623-629.

28. Koroneos G. 2006. RFID: A slow go in pharmaceutical adoption; proposed legislation to
help fight counterfeit drugs. Pharmaceutical Technology 30(4):24.

29. Philips Semiconductors TAGSYS Texas Instruments Inc. 2004. Item-level visibility in
the pharmaceutical supply chain: a comparison of HF and UHF RFID technologies.
White Paper.









30. ODIN Technologies, Unisys Corporation. 2006. Pharmaceutical item level RFID: Battle
of the frequencies. White Paper.

31. Nikitin PV, Rao KVS, Lazar S. 2007. An overview of near field UHF RFID. IEEE
International Conference on RFID 2007 1:167-174.

32. Wal-Mart Inc. 2006. Wal-Mart stores radio frequency identification.
http://www.walmartstores.com/GlobalWMStoresWeb/navigate.do?catg=339

33. United States Department of Defense. 2007. United States Department of Defense
Suppliers' Passive RFID Information GuideVersion 10.0
http://www.acq.osd.mil/log/rfid/supplierguide.htm

34. Dean DA, Evans ER, Hall IH. 2000. Pharmaceutical packaging technology. London, NY:
Taylor & Francis. p 65-80,326-329.

35. Jenkins WA, Osborn KR. 1999. Packaging drugs and pharmaceuticals. Lancaster,
Pennsylvania: Technomic Publishing Company. p 6-10.

36. Pilchik R. 2000. Pharmaceutical blister packaging, part 1 rationale and materials.
Pharmaceutical Technology November 2000:68-76.

37. Lahiri S. 2005. RFID sourcebook. Upper Saddle River: IBM Press. p 3.

38. Zhou X, Wang G. 2004. Study on the influence of curving of tag antennas on
performance of RFID system. 4th International Conference on Microwave and Millimeter
Wave Technology Proceedings 18-21 Aug 2004:122-125.

39. Siden J, jonsson P, Olsson T, Wang G. 2001. Performance degradation of RFID system
due to the degradation in RFID tag antenna. 11th International Conference on Microwave
and Telecommunication Technology:371-373.

40. Costenoble J. 2005. Rotary screen printing: the productive solution for HF/UHF RFID
labels. SGIA Journal 9 Fourth quarter:7-10.

41. Bix L, Sansgiry SS, Clarke R, Cardoso F, Shringarpure GS. 2004. Retailers' tagging
practices: a potential liability? Packaging Technology Science 17:3-11.

42. Cheng DK. 1993. Fundamentals of engineering electromagnetics. New York: Prentice
Hall. p 304-330.

43. Reitz JR, Milford FJ, Christy RW. 1993. Foundations of electromagnetic theory. Boston:
Addison Wesley. p 441-469.

44. Sweeney PJ. 2006. CompTIA RFID + study guide. Alameda, CA: Sybex. p 5-10.









45. Frieden DR. 1985. Principles of naval weapons systems. Annapolis, MD: Naval Institute
Press. p 23-27.

46. Banerjee SR, Jesme R, Sainati RA. 2007. Performance analysis of short range UHF
propagation as applicable to passive RFID. IEEE International Conference on RFID
2007:30-36.

47. Hodges S, Harrison M. 2004. Demystifying RFID: principles and practicalities. White
Paper.

48. Orfanidis SJ. 2004. Electromagnetic Waves and Antennas. p 11-14.
http://www.ece.rutgers.edu/-orfanidi/ewa/

49. Weis SA, Sarma SE, Rivest RL, Engels DW. 2004. Security and privacy aspects of low-
cost radio frequency identification systems. Security in Pervasive Computing
2802/2004:203-214.

50. Raumonen P, Sydanheimo L, Ukkonen L, Keskilammi M, Kivikoski M. 2003. Folded
dipole antenna near metal plate. 2003 IEEE Antennas and Propagation Society
International Symposium 1:848-851.

51. Foster BR, Burberry RA. 1999. Antenna problems in RFID systems. IEEE Colloquium
on RFID Technology 1999(123):3/1-3/5.

52. Sydanheimo L, Ukkonen L, Kivikoski M. 2006. Effects of size and shape of metallic
objects on performance of passive radio frequency identification. The International
Journal of Advanced Manufacturing Technology 30:897-905.

53. Ranasinghe DC, Hall DM, Cole PH, Engels DW. 2004. An embedded UHF RFID label
antenna for tagging metallic objects. Intelligent Sensors, Sensor Networks and
Information Processing Conference, 2004 1:343-347.

54. Ukkonen L, Sydanheimo L, Kivikoski M. 2004. Patch antenna with EBG ground plane
and two-layer substrate for passive RFID of metallic objects. IEEE Antennas and
Propagation Society International Symposium, 2004 1:93-96.

55. Ukkonen L, Engels DW, Sydanheimo L, Kivikoski M. 2004. Planar wire-type inverted-F
RFID tag antenna mountable on metallic objects. IEEE Antennas and Propagation
Society International Symposium, 2004 1:101-104.

56. EIAmin A. 2006. RFID tags turns metal packaging into antennas.
http://www.foodproductiondaily.com/news/ng.asp?id=72561-crown-qinetiq-rfid









BIOGRAPHICAL SKETCH

Dilek Dagdelen Uysal was born in Ankara, Turkey, in 1980. She received her Bachelor

of Science degree in industrial design from Ortadogu Teknik Universitesi (ODTU),

Ankara,Turkey, in 2003. In 2005 she joined the Department of Food Science at University of

Florida to pursue a M.S. degree with a minor in Packaging Science. She is the co-inventor of the

"method and apparatus for tracking transported items using RFID tags", U.S. Provisional Patent

Application Docket No UF-538P. Her research interests mainly include RFID, packaging and

product design.





PAGE 1

1 IMPROVING THE READABILITY OF RFID TAGGED PHARMACEUTICAL PRODUCTS AND FOOD SUPPLEMENTS INSIDE A MODIFIED SEALED TOTE By DILEK DAGDELEN UYSAL A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Dilek Dagdelen Uysal

PAGE 3

3 To my family and friends

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my s upervising committee chair and my advisor Dr. Emond, for his never ending guidance, support, and most importa ntly for being an irreplaceable mentor who gave me the unparalleled opportunity to work in a motivated environment with such a great team. I thank my co-chair, Dr. Balaban for helpin g me not only in my research but in any aspect of my education and also for believing in me I thank my supervisory committee member Dr. Engels for his valuable instruc tions, immense expertise in this field, and invaluable insight. I thank Tom Reese from Motorola; Jeff Wells Steve Dean, Sandy Dean and all the other members of Franwell for their generous oneon-one help, practical industry advice, and providing us up-to-date equipments. Finally, I would never be able to tha nk my husband, Ismail Uysal, enough for his patience, encouragement, and devotion along the way. Without him I could never imagine what I have achieved; he will always have my ad miration and love. I would like to extend my gratitude towards my family, especially my sister Dilber Dagdelen and my brother-in-law Yavuz Yagiz for their unconditional love and support. Last but not the least I would lik e to thank my friends and colleagues for th eir never ending support.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES.........................................................................................................................9 ABSTRACT...................................................................................................................................11 CHAP TER 1 INTRODUCTION..................................................................................................................13 2 LITERATURE REVIEW.......................................................................................................15 Fundamentals of Radio Frequency Identification................................................................... 15 Tag-Reader Communications.......................................................................................... 15 Frequencies......................................................................................................................16 Low frequency..........................................................................................................16 High frequency.........................................................................................................16 Ultra-high frequency................................................................................................17 Microwave................................................................................................................17 Antenna Polarization.......................................................................................................18 Tag Classifications..........................................................................................................19 Standards (Regulations)................................................................................................... 20 Pharmaceutical and Food Supplements Distribution Chain................................................... 21 3 EVALUATION OF RFID PERFORMANCE FOR A PHARMACEUTICAL DISTRIBUTION CHAIN ....................................................................................................... 24 Introduction................................................................................................................... ..........24 Materials and Methods...........................................................................................................26 Characterization of Pharmaceutical Products an d Food Supplements............................ 26 Physical characterization.......................................................................................... 26 Radio frequency characterization............................................................................. 28 Tagging the Products....................................................................................................... 29 Conveyor Belt Testing..................................................................................................... 29 Turn Table Testing.......................................................................................................... 30 Tote Testing.....................................................................................................................31 Results and Discussions........................................................................................................ ..32 Readability of Products on the Production Line.............................................................. 32 Readability of Products in Free Space............................................................................. 36 Readability of Multiple Products inside a Closed Environment..................................... 38 Summary.................................................................................................................................40

PAGE 6

6 4 USING RF REFLECTIVE MATERIAL TO IMPROVE RF WAVE DIS TRIBUTION INSIDE A CLOSED ENVIRONMENT................................................................................ 41 Introduction................................................................................................................... ..........41 Materials and Methods...........................................................................................................42 Preparation of Experiment Setup..................................................................................... 42 Taking the Measurements................................................................................................ 44 Analysis of Data..............................................................................................................44 Results.....................................................................................................................................46 RF Mapping of Non-Metal Sided Tote...........................................................................46 RF Mapping of 2 Metal Sided Tote................................................................................. 49 RF Mapping of 3 Metal Sided Tote................................................................................. 51 RF Mapping of 4 Metal Sided Tote................................................................................. 53 RF Mapping of 5 Metal Sided Tote................................................................................. 55 Discussions.............................................................................................................................57 Behavior of Reflected RF Waves.................................................................................... 57 Insulation.........................................................................................................................58 Security Issues.................................................................................................................58 Summary.................................................................................................................................59 5 EVALUATION OF READABI LITY CONTENTS OF PURCHASED ORDERS ............... 60 Introduction................................................................................................................... ..........60 Materials and Methods...........................................................................................................61 Tote Design.....................................................................................................................61 Preparation of Represen tative Purchase Orders .............................................................. 62 Controlled PO..................................................................................................................63 PO With One Type of Product........................................................................................ 64 Medium Load PO............................................................................................................ 65 Full Load PO...................................................................................................................65 Testing Procedure............................................................................................................65 Data Analysis...................................................................................................................66 Results.....................................................................................................................................68 One Read Attempt........................................................................................................... 68 Second Time Read Attempt.............................................................................................69 Third Time Read Attempt............................................................................................... 71 Discussions.............................................................................................................................72 Placement Order of the Products in the Medium and Full Loaded Totes....................... 72 Performance Differences of Modified Totes................................................................... 73 Other Application Methods of Metal Coating................................................................. 74 Efficiencies of 2nd and 3rd Read Attempts from the Warehouse Perspective.................. 74 Summary.................................................................................................................................75 6 CONCLUSION..................................................................................................................... ..76

PAGE 7

7 APPENDIX A PRODUCTS USED IN THE TESTS..................................................................................... 78 B PRODUCTS UHF TAG TYPE.............................................................................................. 83 C PREPARATION OF PURCHASE ORDERS........................................................................ 84 LIST OF REFERENCES...............................................................................................................91 BIOGRAPHICAL SKETCH.........................................................................................................95

PAGE 8

8 LIST OF TABLES Table page 1-1. Related ISO RFID standards.................................................................................................21 3-1. Physical characterization of pharmaceu tical products depending on their produ ct type and package material.......................................................................................................... 28 3-2. Radio frequency ca tegorization of products .......................................................................... 28 3-3. Means of read rates out of 100 read cy cles of individual produc ts on a turntable in different orientation ........................................................................................................37 3-4. Means of read rates of 100 totes and probability of gettin g 100% read rate with HF and UHF systems......................................................................................................................38 5-1. Statistical analysis of read rates of products on turn table* .................................................. 63 5-2. Example of a calculation of read perf orm ance of a tote if the totes offered 2nd time read attempt*....................................................................................................................... .......67 5-3. Example of a calculation of read perf orm ance of a tote if the totes offered 3rd time read attempt*....................................................................................................................... .......67 5-4. Statistical analysis of read performances of 100 POs with regular, 3, and 5 m etal sided tote offered 1 read attempt*............................................................................................... 68 5-5. Statistical analysis of read performances of 100 POs with regular, 3, and 5 m etal sided tote offered 2nd time read attempt*.................................................................................... 70 5-6. Statistical analysis of read performances of 100 POs with regular, 3, and 5 m etal sided tote offered 3rd time read attempt*..................................................................................... 71

PAGE 9

9 LIST OF FIGURES Figure page 1-1. Linear polarized antenna radiation ........................................................................................18 1-2. Circularly polari zed antenna radiation .................................................................................. 18 3-1. Picture of a turn table testing set-up ................................................................................... ...31 3-2. Probability of reading and not reading products in low speed with a) HF system and b) UHF system.......................................................................................................................33 3-3. Probability of reading and not reading pr oducts in m edium speed with a) HF system and b) UHF system............................................................................................................34 3-4. Probability of reading and not reading prod ucts in high speed with a) HF system and b) UHF syste m.......................................................................................................................35 3-5. Change in read rates of product types acco rding to RF frequency an d the orientation of the product.................................................................................................................... .....37 3-6. Distribution of the means of read rates of 100 totes with using HF and UHF system s........ 39 4-1. 2 metal sided tote....................................................................................................... ............42 4-3. 4 metal sided tote....................................................................................................... ............43 4-4. 5 metal sided tote....................................................................................................... ............43 4-5. RF mapping inside a tote with different levels of m etal coatings set-up.............................. 44 4-6. Layer arrangements of mapping tests.................................................................................... 45 4-7. Color scale for dBm color coding fo r all of the signal strength graphs................................. 46 4-8. Mapping results of 0 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d) Layer 4. ..............................................................................................................................48 4-9. Mapping results of 2 metal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d) Layer 4. ..............................................................................................................................50 4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4...............................................................................................................................52 4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4...............................................................................................................................54

PAGE 10

10 4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4...............................................................................................................................56 5-1. 5 metal sided tote with a grid pattern texture. ....................................................................... 62 5-2. Content of the product, e qua te effervescent pain relief, used in the preparation of one type of product POs...........................................................................................................64 5-3. Distribution of the means of read rates of 100 POs with regular, 3, and 5 m etal sided tote offered 1 read attempt................................................................................................. 69 5-4. Distribution of the means of read rates of 100 POs with regular, 3, and 5 m etal sided tote offered 2nd time read attempt...................................................................................... 70 5-5. Distribution of the means of read rates of 100 POs with regular, 3, and 5 m etal sided tote offered 3rd time read attempt....................................................................................... 72

PAGE 11

11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IMPROVING THE READABILITY OF RFID TAGGED PHARMACEUTICAL PRODUCTS AND FOOD SUPPLEMENTS INSIDE A MODIFIED SEALED TOTE By Dilek Dagdelen Uysal December 2007 Chair: Jean Pierre Emond Cochair: Murat O. Balaban Major: Food Science and Human Nutrition One of the biggest advantages of radio fre quency identification (RFID) technology is that it can be used to augment the security of the s upply chain. A straightforw ard application area for which security is of utmost importance is th e distribution and marketing of pharmaceutical products and food supplements. This research st ates that an RFID technology can be used to develop an RFID system that can reliably read a ll the tagged products insi de a tote while trying to provide security against unauthorized readers, during distribution of these products. The challenge in using RFID to automatical ly, without human intervention, identify the tote, its contents, and the RFID enabled security features are due to the problems related to tag readability near and through diffe rent materials. The solutions addressing these problems are developed for sealed totes by evaluating of proper system set-ups and tote designs. The products and the purchase orders (PO) in the pharmaceutical supply chain were analyzed to provide a repeatab le and comparable evaluation. Each product were tagged and tested on a conveyor belt, on a turn table and in side a tote with both high frequency (HF) and ultra-high frequency (UHF) RFID systems. The results show that UHF systems have better readability performances than HF for all tested speeds levels on the conveyor belt where HF

PAGE 12

12 systems never achieve 100% readability. UHF systems are orientation independent (P=0.875 > 0.05= ) where HF systems are not (P=2.93E-07 < 0.05= ). Moreover UHF systems have higher read rates at reading tags inside a tote (97.61%) than do HF systems (71.70%). After concluding that UHF is more suitable frequency for the pha rmaceutical application of RFID, we evaluated methods of improving tag readability within totes. Using reflective materials inside the tote to control the propagation of UHF radio waves were studied by mapping of the RF field inside totes with 0,2,3,4,5 metal sides. Accordin g to the results, 5 metal sided tote has better RF propagation than any other configurations. In the comparis on of read performances of regular (97.61%), 3 metal sided (98.76%), and 5 metal sided (97.77 %) to tes, the 3 metal sided tote has the highest read rate. This tote also has around 15% more probability of identifying all the tags within the tote than the other totes. The pe rformance of each tote was also ev aluated as if they were offered 2nd and 3rd time read attempts to simulate the real time process and similar results was observed. Finally it is concluded that the readabil ity of tagged pharmaceutical products and food supplements inside a sealed tote can be improved by utilizing reflective materials inside a tote as in 3 and 5 metal sided totes which also enables tote contents of the tote undisclosed against unauthorized readers.

PAGE 13

13 CHAPTER 1 INTRODUCTION Radio frequency identification (RFID) technologies are the la test branch of information technologies to begin changing our everyday live s. Several pharmaceutical manufacturers and distributors are implementing RFID tec hnologies following the US Food and Drug Administrations (FDAs) 2004 recommendation. The FDA encouraged the use of RFID technologies in the pharmaceutical supply chains to improve security and prevent counterfeit products from reaching consumers.1 Over the counter medicine, as well as prescrip tion drugs and food supplements need to be fully controlled due to the high risk of counter feiting. According to World Heath Organization (WHO) between 5% and 8% of pharmaceuti cal products are count erfeit worldwide,2 which costs an average of $40 billion lost every year.3 When the drugs reach to th e wholesaler they are often broken down from cases and pallets and shipped in reusable plastic totes where totes contain the items of a purchase order (PO) placed by a retailer s, hospitals or another wholesalers. One of the best ways to keep a tight control during their transportation from the distribution center is to place every order in a sealed tote where the produc ts inside are labeled with an RFID tag and the tote is scanned without opening it. With the popularity of blister packs and liqui d products in the pharmaceutical industry, the probability of not reading an RFID tag is significant when the tagged products are placed randomly within a tote due to the attenuation of the RF signal.4 Furthermore, a mixed purchase order with just a few blister packages may creat e enough interference to reduce the read rate (the percentage of RFID tags read) of a tote significan tly during the scanning of that tote. This research states that an RFID technology can be used to develop an RFID system that can reliably read all the tagged pr oducts inside a tote while tryi ng to provide security against

PAGE 14

14 unauthorized readers, during distribution of thes e products. Therefore, the objective of this research is to improve security and readability of the RFID tagged pharmaceutical products and food supplements inside a sealed tote by proper de signs of the tote and portal. The problems in the readability of different types of POs are addressed by development of a novel tote design as well as a system encompassing all the aspects from the product picking to scanning process. The findings of this thesis have significant contributions not only to the pharmaceutical industry but also have direct implications for researchers developing new RFID technologies and will also help with the implementation of RFID to real life systems. It enables a secure distribution of pharmaceutical produ cts by suggesting a reliable RFID system set-up for the item level pharmaceutical application which can also be applied to the other item level applications. Moreover, it serves as a guide for the beneficial use of reflective materials to improve readability performances of UHF systems. In the following chapters, fi rst background information about both the RFID technology and the pharmaceutical distribution chain are pr ovided. Then, the performance of RFID technology was evaluated for a pharmaceutical dist ribution chain. In th is chapter the most suitable frequency, UHF or HF, for the item level pharmaceutical application was analyzed. In chapter 4, the usage of reflective material to im prove RF wave distribu tion inside a tote was studied by mapping of RF power insi de the totes with different le vels of metal coatings. The chapter 5 evaluates the readability performances of the totes th at are found to have better RF power distribution compared to a regular tote.

PAGE 15

15 CHAPTER 2 LITERATURE REVIEW Fundamentals of Radio F requency Identification Radio frequency identifica tion (RFID) technologies ar e automatic identification technologies that use tags attached to objects to wirelessly communicate information about that object to a reader. The primary building blocks of RFID systems consist of a tag, a reader, and a host system that processes the data collected from tags by the reader. The reader sends radio frequency (RF) wave s through its antenna and collects the RF waves emitted or reflected from the tag. The information, such as a unique serial number for every tag, is carried with these RF waves. RF waves can be absorbed or reflected by certain materials, which impacts the readab ility of the tags by the reader. Passive RFID systems are beginning to be widely deployed in retail and pharmaceutical supply chains to improve inventory visibility, asset management, supply chain efficiencies, and supply chain security.5 Even though passive RFID systems are bei ng deployed and piloted for improved supply chain management and security, the technologies and systems being implemented are rapidly evolving and their use is not well understood. Significant technical, oper ational, and social problems remain to be solved to achieve the ma ximum possible benefits from the use of RFID systems. Many of these problems are intimately connected to the tags, the frequencies, and the standards being used. Tag-Reader Communications There are two methods of communications betw een the passive tag and the reader: near field inductive coupling and far field backscatter communications. In inductive coupling, the reader antenna coil generates an alternating magnetic field that when part of it is enclosed by the

PAGE 16

16 tag antenna a current is induced in the tag an tenna providing a power su pply to the tag. The information in the tag is communicated to the re ader by having the tag change how much power it draws from the readers magnetic field. The modulation in the tag load is detectable by the reader. Far field backscatter communications use reflected, or backscattered, reader emitted signals to communicate from the tag to the reader. While inductive coupling relies on the fact that the reader emitted signal is still electrically connected to the antenna, backscatter communication works when the emitted signal is not connected to the reader antenna. The tag is powered up by harvesting electri cal and magnetic energy from the electromagnetic field created by the reader.6 Frequencies There are 4 primary ranges of frequencies within which passive RFID systems operate: low frequency (LF), high fre quency (HF), ultra-high frequenc y (UHF), and microwave. Low frequency LF tags operate in the frequency range less than 135 kHz by using inductive coupling. Passive tags using LF fre quencies have demonstrated read-ranges up to 2.0 m.7 Although low frequencies enable RF waves to go through metal a nd water, they have longer read times, higher costs, and larger tag sizes. The maximum rate of tags that existing LF system can read is limited to 20-25 tags a second.8 Their application today is mostly limited to animal ID, access control, and vehicle immobilizers. High frequency HF tags operate in the ISM frequency band centered at 13.56 MHz and are similar to LF tags in the sense that they use inductive coupli ng. Passive tags using HF waves may have readranges up to 1.5 m depending upon antenna sizes.9 RF waves with this frequency can penetrate through water; however, metal objects still constitute a problem. HF tags have faster read rates

PAGE 17

17 than do LF tags where up to 800 tags/s can be read with the newer standard protocols.9 Their primary applications today are in smart cards, access control systems, vehicle immobilization systems, subway cards, and the Mobile Speedpass system. Ultra-high frequency UHF tags operate in the frequency range fr om 860 MHz to 960 MHz. Unlike LF and HF tags, they use electromagnetic coupling communications, i.e., b ackscatter communi cation in the far field of the RF radiation. UHF passive ta gs have exhibited a maximum read-range of 20m.10 The new Gen 2 standards provide a read-rat e up to 1000 tags/s for the passive tags.11 Unlike HF and LF, UHF is more affected by materials; it is highly absorbed by liquids and reflected by metals, especially for far field tags. However, re cent introduction of near field tags operating in the UHF band provides similar adva ntages that the lower frequenc ies have, such as being less effected by water.12 UHF tags are being deployed in supply chain applications at the case and pallet level and passive and semi-passive UHF tags are used for toll collection applications. Microwave Microwave tags operate in frequency bands centered at either 2.45 GHz or 5.8 GHz. They use far field communications as a result of which electromagnetic coupling is employed. They have the fastest read-rate s, although their read ranges are significantly less than those achievable with UHF passive tags. In summary, all frequencies have their respec tive advantages and di sadvantages. As the carrier frequency gets higher, tag sizes get smal ler in general, the read range and read rate increases but the impact of environmental materials increases.13 These properties should be considered carefully for different system designs and applications in order to utilize the most appropriate RFID technology.

PAGE 18

18 Antenna Polarization Polarization is defined as the or ientation of an electric field of the wave. Most antennas radiate via either linear or circular polarization. In the linear polar ization, the radiation pattern of an antenna stays in the same plane at all tim es. Depending on the propagation direction with respect to the earth su rface; it can be vertically or horizontally polarized (Figure 1-1).14 Figure 1-1. Linear pola rized antenna radiation.14 A circularly polarized antenna radiates an el ectric field which is the sum of two fields that have equal amplitude and magnitude with a 90 phase difference; thus it rotates in a circular motion completing a full 360 within one wavelength (Figure 1-2)14.15 Circularly polarized antennas can either be right-handed (clockwise rotation) or left-hande d (counter clockwise rotation). Figure 1-2. Circularly po larized antenna radiation.14

PAGE 19

19 Tag Classifications While we are concerned with passive tags, other types of RFID tags exist. The tags may be classified according to their source of operating and communicating power and communication method. There are 4 basic types of tag: active, se mi-active, semi-passive, and passive. Active tags have their own power source (typically a battery) and are able to actively generate and transmit an electromagnetic signal. They typically have higher performance, larger memories, and longer read-ranges th an do the other types of tags.16 However, their application is limited primarily due to their higher costs, larger form factors, and limited power sources which limits their lifetime. Semi-active tags, like active tags, have an internal power source and use active communication; however, they do not continuously communicate like active tags. Instead, they are often activated manually, such as the keyless entry system in a vehicle or a garage door opener. The use of active communication signi ficantly increases the communication robustness in the presence of noise, allowing for more reliable communications at sh ort ranges and overall longer possible communication ranges as compared to passive communications. Semi-passive tags, like active and semi-activ e tags, have an on-tag power source, but unlike these two types of tags, semi-passive tags use passive communication, either load modulation or backscatter, to communicate with a reader. Th e most common type of semipassive tag today is a battery as sisted passive tag operating in the UHF band. By using their ontag power to operate the logic and functionality of the tag, se mi-passive tags are able to communicate at roughly an order of magnitude further distance th an can passive tags, and they communicate at roughly an order of magnitude less distance than do active and semi-active tags. Passive tags contain no on-tag power source. Instead they ha rvest all of their operational energy from the readers communication signal and use passive communication techniques,

PAGE 20

20 either load modulation or backscat ter, to communicate with the reader The need to harvest all of their operational energy from the readers co mmunication field limits the communication range of passive tags to no more than a few meters at best. The biggest potential for the mass deployment of RFID systems lies within these tags due to their low cost and small sizes. However, the pe rformance of passive tags is more sensitive to orientation and environmental materials than ar e semi-passive, semi-active, and active tags. Their placement with respect to the polarization of the antenna and product package and content has a significant effect on readability.17, 18 The tags have been classified by the Au to-ID center according to their operation characteristics and functionality as Class 0 through Class 5.19 Class 0 Passive read only tags. Class 1 Passive write once, read many tags. Class 2 Passive rewritable tags Class 3 Semi-passive tags Class 4 Active tags power up itself Class 5 Active tags ability to power up other tags as well. Standards (Regulations) Regulations related to RFID systems are cr eated to ensure safe operation with other devices that use RF such as nearby radio and television stations, mobile radio services (police, security services, industry) and cellular phone to prevent a ny interference. There are many standards governing RFID syst ems, but The International St andards Organization (ISO) and EPCglobal are the major organizations that set st andards for RFID systems on data, air interface, conformance and its applications. Table 1-1 pres ents ISO standards related to this research.20 EPCglobal is another organization which regulat es the electronic produ ct code (EPC) that has a unique number for every item in the supply chain. The standards mainly focus on different classes of UHF tags as well as readers.21

PAGE 21

21 Table 1-1. Related ISO RFID standards Pharmaceutical and Food Supplements Distribution Chain Prescriptions drugs and over-the-counter (O TC) products such as food supplements and personal health care products constitute th e goods distributed via the pharmaceutical supply chain. Prescription drugs are the ones that affect the distribution system most with their highly restricted regulations. Before the products reach retail pharmacies or hospitals they pass through at least 4 different stages. The active ingredient of the pro duct is produced in a chemical plant and carried to the manufacturers for packaging after which th e wholesalers distribute the product to the final seller (retailers or hospitals). ISO Standards Title ISO 18000 -1 Generic parameters for the air interface for globally accepted frequencies ISO 18000 -2 Parameters for air inte rface communications below 135 kHz ISO 18000 -3 Parameters for air inte rface communications at 13.56 MHz ISO 18000 -4 Parameters for air interface communications at 2.45 GHz ISO 18000 -5 Parameters for air interface communications at 5.8 GHz ISO 18000 -6 Parameters for air inte rface communications at 860-930 MHz ISO 18000 -7 Parameters for air inte rface communications at 433 MHz ISO 15693 Identification cardscontactle ss integrated circuit(s) cardsvicinity cards for 13.56 MHz ISO 10374 Freight containers automatic identification ISO 17363 Supply chain application fo r RFID Freight containers ISO 18185 Freight containers radio frequency communication protocol for electronic seal

PAGE 22

22 In the United States 90% of the distribution of prescription drugs are made by only five major companies. These companies are primar y, authorized distribut ors who have an ongoing relationship with the manufacturer and purchase a large majority of their drugs directly from them. Secondary wholesalers purchase their pro ducts generally from authorized wholesalers with a rare purchase directly from the manufacturer or from other sources. Secondary wholesalers are used by pharmacies to obtain ra re products or to purchase drugs when the pharmacy is in an area that none of the pr imary distributors have sufficient service. According to FDAs report to the Congress, it is estimated that there are mo re than 6500 wholesalers in US. Of these, 83% are small (fewer than 20 empl oyees), 11% are medium-sized (with 20 to 99 employees), and 6% are large (with more than 100 employees).22 When it comes to the safety of pharmaceutical products throughout the chain, the wholesalers play a very important role because most of the counterfeited products are introduced to the market during the distribution process. A counterfeited drug is defined by WHO as follows: A medicine that is deliberately and fr audulently mislabeled with respect to identity and/or source. Counterfeit medicines may include products with the correct ingredients but fake packaging, with the wrong ingred ients, without active ingredient s or with insufficient active ingredients.23 To prevent counterfeiting, the FDA has only rece ntly indicated that they intend to enforce the pedigree provision of Prescription Drug Ma rketing Act (PDMA) in the regulations in 21 CFR part 203 which concentrates on the regul ation of secondary wholesalers.24 PDMA of 1988 requires a pedigree, documentation on a produc t history, for each product to track them throughout the supply chain to ensure the safety and the effectiveness of the products. Hence, for a complete pedigree a tracking and tracing of every product es pecially during their

PAGE 23

23 transportation by wholesalers are required. The adoption of RFID technology offers efficient preparation of e-pedigree which w ill enable the localizat ion of every product in the market. This will expedite the distribution progress while adding more security features and saving companies both time and money.

PAGE 24

24 CHAPTER 3 EVALUATION OF RFID PERFORMANCE FOR A PHARMACEUTICAL DISTRIBUTION CHAIN Introduction Implementation of RFID technology to pharm aceutical supply chain is highly favorable mainly due to the safety and security concerns in this field. By pr oviding better tracking and tracing with a complete electronic pedigree, RFID will not only deter counterfeiting products but also provide complete supply chain visibility which precludes tampered products as well as products with unacceptable status.25 Other significant adva ntages of RFID include improvements in supply chain efficiencies and asset management like in other supply chains.26, 27 After the recognition of more benefits and the recommendation of FDA in 2005, RFID started to attract a greater interest in the pharmaceutical industry. However, only ten pharmaceutical products are currently being planned to be tagged by the end of 2007 which indicates a slow progress towards its realization.28 One of the main reasons is that the companies are still waiting for the conf lict to be resolved among the two major RFID systems, high frequency (HF) and ultra-high frequency (UHF ), before investing in this technology. The three main reports in the literature which study both systems for a side-by-side comparison for the pharmaceutical industr y are prepared by Texas Instruments29, ODIN technologies & Unisys Corp30 and leading UHF vendors11. The first two papers suggest that HF has more advantages over UHF at the item leve l. Their common conclusion is that even though UHF is not speed dependent and has a greater read range; HF should still be used for the item level pharmaceutical packaging due to its s uperior performance with liquid products. However, according to the paper by UHF vendor s, the recent introduction of UHF Gen 2 near field tags changed the performance of UHF systems significantly These tags have antenna that are more efficient at harvesting the energy in the magnetic field than the more

PAGE 25

25 commonly used UHF tags. This enables better performance through liquids providing similar performance to HF tags. The same UHF tag can al so have both near and far field features which offer the advantages of both systems.12, 31 Moreover, Wal-Mart, the world's largest retailer, and the US Department of Defense (DoD) are lead ing the adoption of passive RFID systems by requiring passive UHF tags to be placed on all pallets and cases delive red to their supply chains.20, 32, 33 Hence, if a company chooses to use an HF system, it might be forced to tag its products with both HF and UHF tags, which migh t increase the cost of overall implementation. Both systems still have problems to be solved before practical implementation worldwide. The author believes that the most im portant issue that needs to be addressed is a thorough frequency performance analysis for these tw o systems. After all, the real advantage of using passive RFID is the ability to utilize complete automation which requires fail-proof performance. For better evaluation of the systems, each step in the pharmaceutic al distribution chain through which the products flow should be consider ed and simulated. The usage of RFID starts right after the packages are fill ed during the manufacturing proce ss. In this step, each tagged product is scanned individually on th e conveyor belt and then cased or palletized to be sent to the wholesalers. The wholesalers separate the produ cts according to their ID and placed them onto the shelves. The orders of th e retailers are put into the seal ed totes which are scanned and confirmed before the distribution. Then, the retail pharmacy rescans each tote without opening it. If the content is confirmed the seal is op ened and the products are placed onto shelves for sale. If it is not the tote is sent back to the wholesaler. The objective of this chapter is to analyze th e advantages and disadva ntages of using HF and UHF systems in pharmaceutical supply chains. State-of-the-art technology is used for the

PAGE 26

26 evaluation of both systems to prevent any bias. While simulating each step in the product flow through the distribution chain following criteria are used for comparison: Scanning speed by testing produc ts in the production line, Orientation and material sens itivity by turn table test, Performance of reading several products at the same time in a closed environment by tote testing. Materials and Methods Characterization of Pharmaceutica l Products and Food Supplements The pharmaceutical industry has a wide vari ety of products in terms of the physical properties of both the products and their packages. In orde r to analytica lly gauge the performances of both RFID systems and to pr ovide a repeatable and comparable evaluation, characterization of the products n eeded. Two primary types of char acterization were performed: Physical characterization: Ch aracterize the products accordin g to products types and the types of materials used in their packaging Radio frequency characterization: Characterize the products accordi ng to their effects on RF waves. Physical characterization The physical state of pharmaceutical products is usually categorized into 4 groups; solid, semi liquid, liquid and gas.34 Examples of products that fall in these categories are as follows: Solid products: Tablets capsules, powders, granules, patches etc. Semi-liquid products: Creams, gels, foams etc. Liquid products: Syrups, oral li quids, droplets, solutions etc. Gas products: Liquefied or pressurized gasses After a market research it wa s found that this list is not comprehensive enough to identify all the products distributed via pha rmaceutical supply chain. As it is indicated in chapter 2, OTC

PAGE 27

27 products are also considered as pharmaceuticals w ith a bigger market share than the prescription drugs.35 Some of the very common products in this group are health care products that are not consumed as drugs per se but used during a treatment. Such products can also be prescribed along with actual drugs. Hence, to comprise these products in the list an other category is created as follows: Sub-products: Needles, syringes, health care products, etc. For the packaging materials, the products are categorized based on only their primary containers such as in plastic, in metal or in gla ss. Paper was not included in this list since it is not a common material that functions as the main containment of the physi cal product itself. Recent developments in plastic enabled it to be the most widely used material in the pharmaceutical industry where it could be used in 99% of all products within 5-10 years.34 The common applications of plastic are bottles, clos ures, blister packs and film layers. Unlike plastic, metal usage as a whole container has declin ed in recent years, but since the blister packs are being used more frequently (approximately 85% of solid drugs in Europe are packed in blisters36) metal is still a very popul ar material for pharmaceutical products. Moreover the applications of metal are more widespread than any other materials: such as lids, aerosol containers, blister packs, closers, and sachet s. Glass is the least common material for pharmaceutical products. However, it is still very valuable especially for the liquid ones. Bottles for the liquids, ampoules and vials are differe nt applications of glass packaging in pharmaceutical industry. Based on the above information and discussions the Table 3-1 is prepared to summarize how the pharmaceutical products were categorized according to their physical properties for this research.

PAGE 28

28 Table 3-1. Physical characteri zation of pharmaceutical products depending on their product type and package material Radio frequency characterization The physically characterized products are grouped into 3 categories, RF lucent, RF absorbent and RF opaque, according to their contai nments/materials interaction with RF waves. RF lucent products are defined as the products th at do not affect RF waves. RF absorbent products cause attenuation on the RF waves by ab sorbing their energy and similarly RF opaque products do not let RF waves pass through by reflecting them.37 In Table 3-2 the products that contain substances with high dielec tric constant, such as water, were placed in the RF absorbent category since it absorb s electromagnetic RF waves and pr oducts with conductors, such as metals, are placed in the RF opaque group. Table 3-2. Radio frequency categorization of products Product Type Package Material a) In plastic b)In metal c) In glass 1. Solid X X N/A 2. Semi-liquid X X N/A 3. Liquid X X X 4. Gas N/A X X 5. Sub-products X X N/A Package Categories 1. RF lucent 1.a, 2.a, 4.c, 5.a 2. RF absorbent 3.a, 3.b, 3.c 3. RF opaque 1.b, 2.b, 3.b, 4.b, 5.b

PAGE 29

29 Tagging the Products For tagging the products the most suitable tags and the best locations to place them to get the highest read rates during the production line, individual read, and inside the tote were chosen according to following criteria: Package content Package surface availability for the size of th e tag to prevent performance loss due to antenna curving38, 39 Application of tags onto products during manufacturing and labeling40, 41 However, the products were not only evalua ted independently but also in a group of different other products to minimize tag collis ion problems which ma y occur in the tote application. 48 products (6 RF lucent, 25 RF absorbent, 17 RF opaque) were fi rst tagged with UHF Symbol (Rockville, MD) RFX6000 1x1 far field and UHF Avery Dennison (Clinton, SC) Gen 2 Satellite near/far field inlays re spectively. According to the prel iminary tests results of conveyor belt and turn table tests the tag which has the highest performance depe nd on the paired-t test with a significant level of 0.05 was chosen for each product for the UHF system. The determination of UHF tag type can be found in Appendix B. For the HF system, HF Rafsec (Fletcher, NC) ISO 15693 14 x 31 mm inlays were used during the conveyor belt and turn table tests. However, these tags were replaced by Omron (Schaumburg, IL) V720-D52P04 inlays due to having higher read distances for the tote testing. Conveyor Belt Testing The tagged product are scanned on the conveyor belt 5 times with 3 different speed levels; 0.17 m/s as the low speed, 0.34 m/s as the medium speed, and 0.51 m/s as the high speed. A line was created on the conveyor belt to keep products in the same position throughout the

PAGE 30

30 belt. At the end of the line th e antennas were placed to be on the top or the side of the product based on the tag position for each package to simu late the manufacturing line. The antenna used for UHF tags was a Symbol linear conveyor belt antenna with a 30 x 50 mm dimensions and for HF, a 190 mm loop antenna with Feig Electronics (Hessen, Germany ) tuning board which was tuned properly right after the set-up was used. A Symbol XR 400 fixed reader and a Feig Electronics OBID i-scan ID ISC.LR200 reader were used for UHF system and HF system, respectively. To compare the two RF system s, the probability of reading a product was calculated for each product type and speed level. Turn Table Testing As shown in Figure 3-1, the already tagged products for the conveyor belt testing were put onto a turntable which was pe rpendicular to the antenna with its center 100 mm away. The antenna position was decided to be parallel to the ground which is a common placement in the smart shelves. This placement will also provide a better setup for the tote testing discussed in the next section. The read rate of a product out of 100 read cycl es was recorded for 5 different angles which were 0, 45, 90, 135and 180 with 0 indicating its upright position. The same readers as in the conveyor belt testing were used for both HF and UHF systems. However, the Symbol AN200 circularly polarized ante nna for the UHF system and a 250x 250 mm loop antenna which was tuned with Feig Electronics tuning board for the HF system was employed. For the analysis, the products were considered within their groups as in the previous test. However, instead of a full probability comparison, th e means (first order stat istics) of read rates were evaluated. Paired-t test was conducted between both systems within the same orientations and ANOVA (analysis of variance) was performed between the orie ntations to decide whether the system was orientation dependent or not.

PAGE 31

31 Figure 3-1. Picture of a turn ta ble testing set-up. Circularly polarized antenna was placed under the products attached to a turn table. Tote Testing A total of 100 different totes, 50 full loaded (with 48 products), 30 medium loaded (with 24 products), 10 with one type of product (with around 32 products) and 10 with controlled products (with 13 products), were scanned on a co nveyor belt with a low speed using both UHF and HF systems. Each tote was scanned 3 times (see Chapter 5 for more information about the preparation of totes). In the UHF set-up the totes was scanned with using a Symbol XR440 reader and 1 Symbol AN200 circularly polari zed antenna which was placed under the conveyor belt. Based on the previous test resu lts all the products were tagged with UHF near/far field inlays for better performance. Moreover, besides Avery Deni son Gen 2 Satellite tags, Omron Gen 2 V750

PAGE 32

32 Ninja near/far field tags wh ich are found to have higher re ad range were placed on some products containing liquid. Unlike UHF, in HF setup a tunnel with 4 ante nnas tuned with Feig tuning boards was created on the conveyor belt such that the tags were able to be s canned in any orientation inside the tunnel. The same reader used for conveyor belt and turn table te sting was employed along with Feig ID ISC.ANT.MUX 8 Channel multiplexer. The data is analyzed in two different ways. First the read rates of totes, i.e., the read percentage of the content of the tote, are ev aluated. Then, the percentages of getting a 100% read rate from a tote out of 100 totes are compared. Results and Discussions Readability of Products on the Production Line Figures 3-2, 3-3 and 3-4 show the continge ncy analysis of response by products in different speed levels with HF and UHF system s. In the figures, (red colored areas) represents not reading a product and (green colored areas) re presents reading a product at least one time. According to Figure 3-2 in the low speed all the products that are in the RF absorbent and RF opaque groups were read with HF system. However, 12% of the RF lucent products was not detected. Note that even though some products were containing small portion of conductive materials such as aluminum lids or needles, they were still categorized as RF lucent products (see Appendix A for the products categories). Hence, not reading some produ cts in this group can still be attributed to sensitivit y of the system either to orie ntation or material. The overall probability of reading products in low speed wi th HF system was 97.50%. In contrast, UHF system had 100% probability of reading products by detecting all the pro ducts on the line for the same speed level.

PAGE 33

33 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 0 1 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 1 a) b) Figure 3-2. Probability of readi ng and not reading products in low speed with a) HF system and b) UHF system.

PAGE 34

34 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 0 1 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 0 1 Figure 3-3 illustrates the pr obability of reading and not reading products in medium speed. Where 23.64% of RF absorbent, 32.00% of RF lucent, 17.50% of RF opaque products were not read by the HF system, the UHF system read all except 2.50% of RF opaque products. The performance of the HF system for medium speed has decreased 21.36% compared to low speed by having a total of 76.67% probability of r eading all products. However UHF has a very negligible decrease which is less than 1% by reading 99.17% of all products. a) b) Figure 3-3. Probability of readi ng and not reading products in medi um speed with a) HF system and b) UHF system.

PAGE 35

35 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 0 1 responce 0.00 0.25 0.50 0.75 1.00 RF absorbent RF lucentRF opaque products 0 1 Finally, Figure 3-4 demonstrates the readability of products in high speed. With the HF system, almost half of the products in RF absorbent and RF opaque groups and 68.00% of the products in RF lucent group could not be read. On the contrary with UHF only 3.64% of RF absorbent and 5.00% of RF opaque groups could not be read. The overall HF performance showed even a higher decrease of 39.13% compared to medium speed by having 46.67% of readability of all products whereas UHF had only a slight performance loss (2.50%) compared to medium speed by reading 96.67% of all products. a) b) Figure 3-4. Probability of readi ng and not reading products in high speed with a) HF system and b) UHF system.

PAGE 36

36 We conclude that HF is a speed dependent system with a gradual decrease in the probability of reading products on the conveyor be lt with increasing speed. More importantly, it could never achieve a 100% read rate even at the low speed level. Even though the UHF system is also affected by conveyor belt speed, it is not as si gnificant as HF. Moreover, the performance of the system is sufficient to get a 100% read at the low speed level. Finally, it is worth mentioning that the speed level for the UHF system can be increased in small steps because the performance difference is marginal between lo w and medium speeds. Depending on the product type the manufacturer can choose to use higher speeds without any significant adjustments to the system setup. Readability of Products in Free Space The statistical means of read rates out of 100 read cycles of individual products on a turntable in different orientations are shown in Table 3-3. Accordi ng to pair-t test results HF has a statistically better performance than UHF only at reading RF abso rbent products at 90. In all other product-orientation combinations UHF has e ither significantly better results (reading all types of products at 0 and 45) or has a sim ilar performance with no statistical differences. Moreover, if the read rates of orientations averaged for every product and system are compared, UHF is found to be signif icantly better (P=0.00135 < 0.05= ) at reading different types of products. Moreover, Figure 3-5 demonstrates that even though th ere are no statistical differences between 5 different orientations (0, 45, 90, 135, 180) in UHF (P=0.875 > 0.05= ), HF is observed to be orientation de pendent with a significant di fference between orientations (P=2.93E-07 < 0.05= ) where 90, 135, 180 are similar and better than 0 and 45 at scanning a tag on different products.

PAGE 37

37 0 10 20 30 40 50 60 70 80 90 100 110 0 45 90 135180AngleRead Rat e rf absorbent hf rf absorbent uhf rf lucent hf rf lucent uhf rf opaque hf rf opaque uhf Table 3-3. Means of read rate s out of 100 read cycles of indi vidual products on a turntable in different orientation *Same letters represent no difference. Figure 3-5. Change in read rate s of product types according to RF frequency and the orientation of the product. 0 Degree 45 Degree 90 Degree 135 Degree 180 Degree HF UHF HF UHF HF UHF HF UHF HF UHF RF absorbent 45.45 B 85.45 A 36.36 B 73.45 A 90.90 A 63.63 B 81.72 A 72.72 A 63.27 A 72.72 A RF lucent 40.00 B 100 A 40.00 B 100 A 100 A 100 A 100 A 100 A 100 A 100 A RF opaque 25.00 B 94.12 A 61.25 B 100 A 87.50 A 100 A 98.50 A 100 A 87.50 A 100 A

PAGE 38

38 To conclude, HF reading RF absorbent produc ts better than UHF at one degree proves the theory that HF is less material sensitive than UHF.9 However, while HF might be superior when reading liquid products, it can be claimed that it is sti ll negatively affected by metal like UHF. Furthermore, a multiple antenna setup is required for HF to minimize its reliance on orientation whereas a single antenna is sufficien t for UHF which increases its cost efficiency. Readability of Multiple Products inside a Closed Environment Table 3-4 summarizes the experimental results for 100 different product mixes corresponding to distinct Purchase Orders (POs). The POs were scanned with both HF and UHF systems. The mean distributions for different types of POs were also added to the results for a detailed overview. The mean of total r ead rates with the UHF system is 97.61.52, [97.09, 98.12] which is substantially higher than the m ean read rates for the HF system that is 71.70.90, [69.80, 73.56] with a 95% confidence interval. Figure 3-6 provides a quick visualization of the performance differences and da ta distribution of the two RFID systems. As can be observed from this figure, the UHF sy stem is a more reliable system by having less variance in performance while reading different totes when compared to the HF system. Table 3-4. Means of read rates of 100 totes and probability of ge tting 100% read rate with HF and UHF systems PO Types Number of Samples Read Rates Probability of getting 100% read rate UHF HF UHF HF Control 10 98.46 74.62 8.00 0.33 One type 10 97.18 65.42 6.00 0 Medium tote 30 96.67 80.28 12.00 0 Full tote 50 98.08 67.22 21.00 0 Total 100 97.61 [97.09, 98.12] 71.70 [69.80, 73.56] 46.00 0.33

PAGE 39

39 read rate 50 60 70 80 90 100 HF UHF system Figure 3-6. Distribution of the m eans of read rates of 100 totes w ith using HF and UHF systems. Another significant statistic is the occurrence of 100% read rates. Table 3-4 shows that the UHF system has outperformed the HF system in this category, as well. Out of 100 totes, the contents of 46 totes were fully scanned with the UHF system where as not even 1 tote could be read entirely with the HF system. The goal of using RFID tagged products is to obtain 100% read rates for all totes. To sum up, UHF has a superior performance th an HF at reading products in a closed environment such as in a tote. Getting such a lower reads from HF can be attributed to not only HF being speed and orientation depended as proved in previous tests but also having a lower performance at scanning multiple products together. The individual results for different totes showed that, in general, the read rate decrease d as the amount of products included in the totes increased.

PAGE 40

40 Summary Recent innovations in RFID technology prom ise many advantages for pharmaceutical supply chains. However, there are many details that need to be considered regarding the system setup, some of which were highlighted in this chapter. Three different tests were performed to analyze various properties of HF and UHF systems. Even though both systems have their pluses and minuses, UHF pulls forward when pharmaceutical applications are considered in par ticular. The main reason is that UHF systems have less overall dependency on system setup fo r optimal performance. Other reasons include the fact that they are less affected by speed and orientation and finally becoming less sensitive to materials as well, with the recent intr oduction of near/far field UHF tags. To conclude, this chapter of the research recommends the use of UHF systems in the item level pharmaceutical applications. Even though a better HF system set-up (reader/antenna configuration) could be used to improve the in dividual product read ra te, UHF systems have a greater opportunity to be used in a physically secure tote due to the angular dependence of HF systems.

PAGE 41

41 CHAPTER 4 USING RF REFLECTIVE MATERIAL TO IMPROVE RF WAVE DISTRI BUTION INSIDE A CLOSED ENVIRONMENT Introduction This chapter states that RF propagation can be guided to a better distribution of RF power by the reflective materials inside a closed environment. The motiva tion is that the metal reflects UHF radiation; therefore, metal coatings on the side s and within the totes can be used to increase the received power of the tag inside a tote which might result in the increase of overall readability of tote content. The metal can also be utilized as a physical se curity barrier to prevent unauthorized readers scanning the tags inside a metal cage which provi des another security aspect usually concerned in the RFID systems. As the principle idea, electromagnetic waves reflect off metals at the same angle they encounter the surface with a phase reversal (180 phase shift).42, 43 During this reflection insignificant power loss is observed since metals have lower index of refraction.44 The wave energy is also affected by interfer ence of waves at the same frequency. This interference can either be constructive or destru ctive depending on the phase, amplitude and the polarization of waves. For instance 0 to 120 and 240 to 360 phase differences result in constructive interference whereas 120 to 240 phase differences cause destructiv e interference in the waves with same polari zations and same frequencies.45 In summary, the objective of this part of the research is to determine the RF power distribution within totes that have different levels of metal coati ngs. In order to accomplish this goal, mappings of RF wave power inside totes w ith 0 (control), 2, 3, 4, and 5 sides coated with reflective materials were experimentally determined.

PAGE 42

42 Materials and Methods Preparation of Experiment Setup 5 different totes with different levels of al uminum coating as a reflective material were prepared. The totes that are tested are listed as follows: Non-metal sided tote as a control unit 2 metal sided tote has metal coati ng on the 2 adjacent sides (Fig. 4-1) 3 metal sided has metal coating on the 2 adjacent sides and the top (Fig. 4-2) 4 metal sided tote has metal coati ngs on 4 adjacent sides (Fig. 4-3) 5 metal sided tote has metal coating on all sides except the bottom (Fig.4-4) The totes were placed 50 mm away from the top of the UHF circularly polarized antenna with a 6 dBi gain, Symbol AN200, which is connected to a Sy mbol XR400 reader. Figure 4-1. 2 metal sided tote. Figure 4-2. 3 metal sided tote.

PAGE 43

43 Figure 4-3. 4 metal sided tote. Figure 4-4. 5 metal sided tote.

PAGE 44

44 Taking the Measurements The RFID system was operated with th e highest output power of 30 dBm. The distribution of RF wave power inside the to tes was measured with the Agilent/HP 8595E RF spectrum analyzer. The frequency of the spectr um analyzer was centere d at 915 MHz with a 200 MHz span. As shown in Figure 4-5, a 38 mm omni directional antenna wi th a 1 dBi gain was placed at the center of each cell on a grid platfo rm with 4x4 cells and the signal strength was recorded at every cell. This process is repe ated throughout the tote where the platform was moved to the centers of 7 intervals that divide the tote into equal spaces. Figure 4-5. RF mapping inside a tote with different levels of metal coatings set-up. Analysis of Data During the analysis the data is compresse d into 4 layers inst ead of 7 for better visualization of the power distribu tion. Layer 1 is the closest layer to the front and layer 4 is the furthest as illustrated in Figure 4-6. The data on each layer is shown with a surface chart where

PAGE 45

45 each corner except the ones on the edges represents a value taken from the measurements. Since it was impractical to take data on the sides of the totes, a su itable value was assigned to the corners at the edges in order to map the entire tote. This value is determined in such a way that for the data points at the edges if there is no me tal coating, the values of closest row or column are averaged and if there is a metal coating 10 dBm was assigned as a minimum value to power up the tag46 to show that the tags can not be read on these surfaces. Figure 4-6. Layer arrangements of mapping tests. Layer 1 is the closest layer to the front and layer 4 is the furthest. In the results and the discussion sessions the to tes with metal coatings are evaluated as if there were 10 mm insulators after the metal surf aces. The insulator borders are presented in the surface charts as dashed lines.

PAGE 46

46 The scale for the RF wave power in dBm was color coded from red to blue as signal strength increase, the color changes from blue to red as shown in Figure 4-7. Light blue colors show the null spots where a tag cannot be read. Figure 4-7. Color scale for dBm color codi ng for all of the signal strength graphs. Results RF Mapping of Non-Metal Sided Tote Figure 4-8 demonstrates the RF power distri bution inside a regular tote with no metal coating. Even though a null spot is not detected anywhere, the signal strength at the bottom of the tote drops to as low as -10 to -5 dBm. These and their surrounding zones lower than 0 dBm can be considered as critical areas which have a greater possibility of creating a zone where reading a tag may not occur when the tote is fi lled with products. The low power at the bottom of the tote, the places close to the antenna, can be ascribed to the antenna radiation pattern. In a typical UHF antenna, the waves are radiated in a cone like shape whic h has narrow radiation beam close to the antenna causing more null spots or low power levels in these areas as it is also verified by the results of non-metal sided tote. The maximum power that is observed in all layers is between 10 to 15 dBm and the dominant power level is between 5 to 10 dBm.

PAGE 47

47 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 7654321 bottom D C B A Top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 a) Layer 1 b) Layer 2 Figure 4-8. Mapping results of 0 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.

PAGE 48

48 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 c) Layer 3 d) Layer 4 Figure 4-8. Mapping results of 0 me tal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote.

PAGE 49

49 RF Mapping of 2 Metal Sided Tote According to Figure 4-9 in the tote with 2 me tal coated sides, the same data points have the lowest powers but higher values within a ra nge of -5 to 0. This clearly shows that the distribution of RF waves is cha nged by the waves reflected from the metal sides. However there is no significant difference in the average overa ll power by having a layer, layer 4, which has a maximum power between 5-10 dBm. a) Layer 1 b) Layer 2 Figure 4-9. Mapping results of 2 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A Top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 50

50 c) Layer 3 d) Layer 4 Figure 4-9. Mapping results of 2 me tal sided tote in dBm a) Layer 1, b) Layer, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 51

51 RF Mapping of 3 Metal Sided Tote The results of the tote with 3 metal coated sides are shown in Figur e 4-10. Notice that the average power has significantly increased. Inside the free space of the tote, the lowest points are located only in a small area, the righ t top corners of layer 3 and layer 4, with values between -10 dBm and -5 dBm. Moreover the maximum power is higher than the previ ous totes with a value between 15-20 dBm instead of 10-15 dBm. a) Layer 1 b) Layer 2 Figure 4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator borde r

PAGE 52

52 c) Layer 3 d) Layer 4 Figure 4-10. Mapping results of 3 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 53

53 RF Mapping of 4 Metal Sided Tote As illustrated in Figure 4-11, the tote with 4 metal coated sides has similar power levels on average with the 3 metal sided tote. However fo r the first time a null spot is observed inside a tote at the left bottom corner of layer 4. As in the non-coated and 2 metal sided tote, the lowest power is distributed at the bottom of the tote inside the free space of the tote. This indicates that the sides across each other created a multi path where the waves canceled each other. Due to having a slight angle on the si des this effect is not observed for the top areas. a) Layer 1 b) Layer 2 Figure 4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator borde r 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator borde r

PAGE 54

54 c) Layer 3 d) Layer 4 Figure 4-11. Mapping results of 4 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 55

55 RF Mapping of 5 Metal Sided Tote Figure 4-12 shows that the tote with 5 metal co ated sides has the best RF power levels of all the totes. It has more uniform wave dist ribution along the edges as well as a higher average power level. The lowest signal strength which is only observed at one point is between 0-5 dBm inside the free space of the tote Moreover the highest power is between 35-40 dBm which is almost 3 times more than the highest valu e measured in the other configurations. a) Layer 1 b) Layer 2 Figure 4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 56

56 c) Layer 3 d) Layer 4 Figure 4-12. Mapping results of 5 metal sided tote in dBm a) Layer 1, b) Layer 2, c) Layer 3 and d) Layer 4. The antenna was placed 50 mm away from the bottom of the tote. 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border 7654321 bottom D C B A top 35-40 30-35 25-30 20-25 15-20 10-15 5-10 0-5 -5-0 -10--5 -15--10 Insulator border

PAGE 57

57 Discussions Behavior of Reflected RF Waves This part of the project face with two very important problems that surround the propagation of RF waves inside closed environments. First, the occurrence of destructive interference of the waves by meta l surfaces, which can be observed from the results of tote # 4, explains how RF waves cancel each other while having multiple paths. Second, the path loss occurre nce causes an attenuation of the RF waves. As it is known, the energy of the waves decreases proportionally to the inverse square of the distance6, 47, 48 that they travel. Hence the power loss will be greater inside metal coated totes since the waves will travel more due to the reflection in the path of RF waves. However in this chapter, it was shown that both these problems can be addressed by properly guiding the RF waves with reflective mate rials. For example, in the example of path loss, which is simple unavoidable, the average RF wave power level in a region can still be kept the same by the joined force of multiple waves that are reflected from different angles instead of a single wave. This is observed in all the tote s with metal surface, that the average RF wave power level was even increased compared to the control tote. More over the lowest power regions remain similar in every tote, but depending on the metal surface c onfiguration they have different signal strengths, which is an other example of the joined force. In the tote designs, it is found that the top metal surface plays a grea ter role in increasing RF power when the results of tote #3 and #5, wh ich has the highest measure power respectively, compared to the others. Furthermore the selection of antenna type based on polariza tion also plays an important role in the distribution of RF waves. Notice th at the RF power distribution is not symmetric in any totes due to the use of circ ularly polarized antenna. This type of polariza tion enables to

PAGE 58

58 decrease the probability of destructive interferen ce occurrence of the waves reflected from the metal surfaces across each ot her in the totes #4 and #5. Insulation The other interesting ob servation from the results of these tests is the power measured near the metal surface in the case of layer # 4 tote # 5. Here, the high power zone affect and strengthen the field around the metal surface adjacent to these regions. More power level may turn on the tag and backscatter the signal which will enable the tag to be read even when it is in close proximity to a metal surface. This may lead us to use thinner insulator which can be studied as a future work. Security Issues One of the main drives for this research is to keep the order safe by sealing the tote to prevent loss in any step of the di stribution. The other s ecurity issue can be considered as keeping the order, i.e., the contents of the tote, undisclosed. One of the dr awbacks of an RFID system is that it raises security and privacy concerns by making unprotected tags vulnerable to eavesdropping, traffic analysis, sp oofing or denial of service as defined by Weis S.A. and others.49 Among them spoofing is a real threat to the supply chain systems where an unauthorized reader can scan, change or copy the in formation of the tag. For example in the case of a sealed pharmaceutical tote, someone can si ngle out and counterfeit only the totes that contain valuable or special products which the counterfe iter is interested in. In order to preclude the disclosu re of the totes contents, the readability of the tags inside the tote by unauthorized readers s hould be prevented. The results of this chapter showed that the metal can also be used as a shield against un authorized readers by preventing RF waves passing through it. For instance, if 5 meta l sided totes are stacked in a pall et on a metal plate, a complete protection against unauthorized readers is achie ved. Same effect can be ensured for 4 metal

PAGE 59

59 sided totes when top and bottom metal plates are used on a pallet. For 3 metal sided totes, a special configuration of stacking is needed where the metal sides of half of the totes are reversed and the totes are put in an order to cover all possible intrusion sides. Summary In this chapter the metal effect on the dist ribution of electromagnetic RF waves inside a closed environment was evaluated by RF mapping of totes with different levels of reflective material coatings. It is always claimed that the metal causes attenuation in the UHF signal by creating multi paths and in terference of the waves; however this phenomenon can be controlled and even be used as an opportunity to increase RF power level by proper use of metal. Tote number 5 is a good example, wher e all the sides except the bottom are coated with metal. The highest measured RF power in this tote was almost 3 times higher than the control tote (nonmetal sided tote). Metal can also be utilized as a barrier agai nst RF waves. When tote #4 and #5 palletized a full protection against unauthorized readers can easily be achieved. This prevents disclosure of the totes contents pr oviding more security. Finally, better distribution of RF waves via use of reflective material may increase the probability of reading tags, which will be explored and evaluated in the next chapter

PAGE 60

60 CHAPTER 5 EVALUATION OF READABILITY CONTENTS OF PURCHASED ORDERS Introduction The contents of the totes used for distributi on of pharmaceutical produ cts vary in size and material. The random orientations and mixing of products within the tote s make it probable that any liquids and metals that are part of one pr oduct or its packaging will block the path of RF waves thus negatively impacting the readability of the totes conten ts. The results of Chapter 4 showed that the metal can be utilized to increas e the RF power inside a tote. This chapter will investigate how the use of metal in a tote design affects the readability of the tote. The negative effects of metal on the tag an tenna, such as detuning, are well known. For instance, when the tag is placed too close to th e metal, the tag might not be read due to the changes in the antennas radiation pattern, input impedance, radiation efficiency or resonant frequency.50 To minimize these negative impacts the tag should be separated from the metal. A good example is the degradation of antenna radiation in metal products being much higher than the ones with liquid or any other materials, whic h requires the tag to have better insulation from the product itself.51 Moreover the size and the shape of the metallic object also affect the radiation pattern of the antenna.52 However, metal can also be used in ways to actually improve the performance of an RFID system. Applications, such as using a meta llic surface as a reflector or the ground plane of an antenna to harvest more energy for the tag, are becoming more common.53, 54, 55 Researchers have recently begun investigating the possible use of the metal p ackage material itself as an antenna.56 This chapter explores a novel constructive use of metal to generate multiple paths via reflected waves, potentially creating a viable co mmunication path to a ta g that would otherwise

PAGE 61

61 be shielded by the contents of the tote. This is accomplished by developing a modified tote to improve the readability of its contents with a UHF system by using reflective materials inside. Different modified totes were evaluated with a set of representative POs, where each PO is offered 1, 2 and 3 times of read attempts. Materials and Methods Tote Design The tote is designed based on the work of the previous chapter that suggests the metal can be utilized to improve readabil ity if better distribution of refl ected RF waves is achieved. The first design was to use a full metal cover on the 5 sides leaving the bottom open to enable the waves entering through there since during the RF mapping analysis th is tote showed the best RF wave propagation inside. However, inside 5 metal sided tote the tags placed on the 10 mm insu lator, used in the previous chapter, from the metal surface could not be read due to metal feature of detuning the antenna. The thickness of the insulator wa s increase by 5 mm starting from 10 mm and it was found out that a minimum insulation of 25 mm be tween the metal side and the tag was required to read the tag when the tote is closed. According to researches the bigger the size of the metal the more separation is required between the tag antenna and the metal surface.45 To decrease the insulator thickness for more available free space fo r the products, the concentration of the metal was decreased by using metal grid texture on th e surfaces that has 10 mm gap between each 10 mm metal stripes instead of a full metal cover sh own in Figure 5-1. Different thickness levels for insulators starting from 25 mm were tried and the minimum required thickness was found to be 10 mm. This texture also provided a similar effect with the full cover by still reflecting the RF waves without letting them go through the surfaces.

PAGE 62

62 Along with the 5 metal sided tote, the 3 metal sided tote with metal grid texture was also tested since it had the second best RF propagation to evaluate a nd better understa nd the usage of reflective material at im proving the read rate. Figure 5-1. 5 metal sided tote with a grid pattern texture. The bottom side is not covered with metal surface. Preparation of Representative Purchase Orders The POs which only represents the worse cas e scenarios were prep ared to lower the sample number that needs to be tested and to observe if there is a significant difference between the modified totes. A total of 100 POs was categ orized into 4 groups which were controlled PO, PO with only one type of product, medium load PO and full load PO. The distribution of 100 POs based on their categories was; 10 controlled, 10 with one type of product, 30 with medium load and 50 with full load POs. The detailed information about the products and pictures are ava ilable in the appendix section.

PAGE 63

63 Controlled PO To prepare the controlled POs first the products that had the worst read rates in turn table tests in Chapter 3 were chosen. Then, for ever y product both HF and UH F results were assessed together within orientations and t test was conducted to find the differences between the orientations. If there were differenc es, t test with a significant level ( =0.05) is performed to rank the read rates. The results of statistical analysis for every product that are used in this type of PO are shown in Table 5-1. Table 5-1. Statistical analysis of re ad rates of products on turn table* 0 45 90 135 180 Equate Flu Relief 76.50 B 100 A 100 A 100 A 100 A Magnesium Citrate Oral Solution 0 B 7.00 B 50.00 A 50.00 A 0 B Pedialyte Freezer Pops 0 C 50 B 100 A 100 A 100 A OCuSoft 46.18 B 47.50 B 66.17 A B 100 A 100 A MGP Bromaxefed DM RF Syrup 100 A 48.00 B 50.00 B 100 A 100 A MGP Bromaxefed DM RF Syrup 1 Pint 16.33 A 0 B 0 B 0 B 0 B Lovenox 50.00 B 50.00 B 100 A 100 A 100 A NuvaRing 100 A 95.00 A 100 A 100 A 50.00 B Procanbid 50.00 B 50.00 B 100 A 100 A 100 A Xylocaine-MPF 50.00 B 50.00 B 100 A 100 A 100 A Equate Triacting Cold & Cough 50.00 B 50.00 B 100 A 100 A 100 A Hurricaine Tropical Anesthetic Spray 50.00 B 50.00 B 94.00 A 100 A 100 A Tinactin Antifungal Spray Can 50.00 B 50.00 B 100 A 100 A 100 A *Same letter represents no differe nce between the orientations.

PAGE 64

64 Finally the worse orientations for each product were selected. If there was more than one, each one was assigned a number which was th en selected randomly. 10 different PO was created out of 13 different products that were in their worse orientations. A detailed worksheet for this study can be found in Appendix C. PO With One Type of Product The tote was filled completely with equate effervescent pain relief packages which contain the tablets inside aluminum patches. This product is chosen instead of any blister packages to create more challenge during scanni ng. As shown in Figure 5-2 there is no air gap between each aluminum patches which do not leav e any space for the waves to pass through it. The only way to reach a tag in the middle is by utilizing the air gap between each product which can be blocked easily by the patches with even a small movement in the tote. The packages were arranged inside the tote such that they were all positioned in the same orientation as if they were in a case. 10 different POs were created out of 28 or 32 products depending on the case size that fits inside the tote. Figure 5-2. Content of the product, equate effervescent pain relief, used in the preparation of one type of product POs.

PAGE 65

65 Medium Load PO Medium loaded tote contained 24 products that have 50% RF opaque and 50% RF absorbent products. The samples were prepared by taking all of the produc ts outside and placing them again randomly inside the tote and each sample scanned 3 times without changing any placement of the products. However the products w ith liquid were first put inside before other products. This type of placement is also a comm on practice in real world applications as it is confirmed by McKesson, one of the largest wholesalers. 30 different POs were te sted in this category. Full Load PO The totes with full load POs had 48 products, 75% of which were RF opaque (18) and RF absorbent (18). More samples (50) were prepared for this type simply due to having a higher population number by the permutation of random or ientations for more products. Similar to medium load POs, for every sample, all the prod ucts were taken outside and placed back inside the tote randomly. The placement order of the pr oducts is first the RF ab sorbent, then the RF lucent and finally the RF opaque products. The pr oducts with blister packs were put in such a way that the metal surface w ould be facing to top. Testing Procedure The same types of tags that are used in ch apter 3 for the UHF setup were applied onto products before the POs were prepared. Each PO was tested 3 times for th e validity of statistical analysis by using the same speed level (0.17 m/s) on a conveyor belt within a regular tote (non metal sided tote), 3 metal sided tote, and 5 metal sided tote. The content of the totes were scanned with one Symbol UHF ci rcularly polarized antenna placed under the conveyor belt and Symbol XR440 reader.

PAGE 66

66 Data Analysis During the analysis two-way analysis of vari ance (AOV) and t test statistical methods with a 5% significance level were performed for the evaluation of percentages of read rates. In addition, the probability of getting 10 0% read rate with regular tote and the modified totes were compared. To simulate the real time process the data was also analyzed as if the totes that were not 100% read were given a 2nd and 3rd time of read attempt. Hence the performances of different tote designs at reading 100 totes were evaluated into 3 groups: a) The totes only offered 1 read attempt b) The totes that were not read entirely offered a 2nd time read attempt c) The totes that were not read entirely offered a 3rd time read attempt For the 2nd and 3rd trials, instead of scanning each PO 9 times (3 times for each 1st, 2nd and 3rd trials) 3 times of reading were conducte d without writing over each other. Then, according to trial version, the 6 combinations out of 3 scans were evaluated as if the data were written over each other. The read rates were av eraged and the probability of getting 100% read rate calculated along with the time spent on reading that tote. For ex ample; a full tote analysis in a group b was performed as in Table 5-2. Similar calculations were con ducted for group c totes; however the 6 combinations were abc, acb, bac, bca, cab, cba (Table 5-3.) Moreover, the evaluation of the other trials provided us w ith a means to compare the average times of readin g 100 POs within different types of totes.

PAGE 67

67 Table 5-2. Example of a calcula tion of read performance of a tote if the totes offered 2nd time read attempt* *The read rates of same tote samples are repres ented by a, b and c. The numbers next to them are the numbers of produc ts that are read inside the tote a nd the numbers in the parentheses are the product IDs. Table 5-3. Example of a calcula tion of read performance of a tote if the totes offered 3rd time read attempt* *The read rates of same tote samples are repres ented by a, b and c. The numbers next to them are the numbers of produc ts that are read inside the tote a nd the numbers in the parentheses are the product IDs. Read rates of samples out of 48 products for a PO: a= 47 (13) b= 47 (20) c=48 Sample order Read rate out of 48 Time spent on scanning probability of getting 100% read rate ab 47 2 0 ac 48 2 100 ba 47 2 0 bc 48 2 100 ca 48 1 100 cb 48 1 100 Average 47.66 = 99.31% 10/6 = 1.66 66.66% Read rates of samples out of 48 products for a PO: a= 46 (18,27) b= 47 (18) c= 47 (6) Sample order Read rate out of 48 Time spent on scanning probability of getting 100% read rate abc 48 3 100 acb 48 2 100 bac 48 3 100 bca 48 2 100 cab 48 2 100 cba 48 2 100 Average 48 = 100% 14/6 = 2.33 100%

PAGE 68

68 Results One Read Attempt Table 5-4 indicates the data of reading 100 POs without offering another read attempt. According to the results the aver age read rates of a total of 100 different POs within totes have statistical differences (P= 3.25E-05 <0.05). Howeve r the regular tote and the 5 metal sided tote have similar values of 97.61.52 and 97.77.45 re spectively with no si gnificant difference. Even though the 5 metal sided tote has better performances than the regular tote in most of the PO types, its performance degr ades significantly for full totes. On the contrary, the 3 metal sided tote has an almost 1% better read rate than the 5 metal sided tote and more than 1% better read rate than the regular tote w ith a value of 98.76.34. It has the highest performance not only in the average read rate but also in every PO type with a smaller variance (Figure 5-3). Furthermore with the 3 metal sided tote 18% more totes than the regular tote and 15% more totes than the 5 metal sided tote were read en tirely which serves as th e real indicator of the dramatic improvement in performance. Table 5-4. Statistical analysis of read perf ormances of 100 POs with regular, 3, and 5 metal sided tote offered 1 read attempt* *Same letter represents no di fference between the totes. PO Type # of Samples Read Rates Probability of getting 100% read rate regular 3 metal sided 5 metal sided regular 3 metal sided 5 metal sided control 10 98.46 99.49 99.74 8.00 9.33 9.66 one type 10 97.19 97.81 97.81 6.00 6.33 6.66 medium loaded 30 96.66 99.17 97.18 12.00 24.00 13.33 full loaded 50 98.08 98.56 97.72 21.00 24.33 19.33 Total Results 100 97.61 B 98.76 A 97.77 B 46.00 64.00 49.00 Margin of Error 0.52 0.34 0.45

PAGE 69

69 read rates 87 89 91 93 95 97 99 101 3 side 1 5 sided 1 regular 1 Column 1 Each Pair Student's t 0.05 Figure 5-3. Distribution of the m eans of read rates of 100 POs with regular, 3, and 5 metal sided tote offered 1 read attempt. Circles separated each other represents significant differences between the totes. Second Time Read Attempt Table 5-5 shows that there is significant differences between totes at reading 100 POs that are offered a 2nd time read attempt if 100% read is not achieved (P= 6.2E-04<0.05). Even tough there is an increase in the performance of 5 me tal sided tote by having a better read rate at reading full tote, no significant differences was observed with a regular tote. However the 3 metal sided tote still has the highest read rate on average, 99.26.26, with the least variance as illustrated in Figure 5-4. For the probability of getting 100% read rate, the 3 metal sided tote shows a superior performance with reading 75 POs completely out of 100 whereas the 5 metal sided and regular totes reads 66 POs and 58 POs, respectively (Table 5-5). Moreover the 3 metal sided tote requires less time than the other totes, as much as 11.08% less than the regular tote and 9.71% less than the 5 metal sided tote.

PAGE 70

70 read rates 89 90 91 92 93 94 95 96 97 98 99 100 101 3 side 1 5 sided 1 regular 1 Column 1 Each Pair Student's t 0.05 Table 5-5. Statistical analysis of read perf ormances of 100 POs with regular, 3, and 5 metal sided tote offered 2nd time read attempt* *Same letter represents no difference between the totes. Figure 5-4. Distribution of the m eans of read rates of 100 POs with regular, 3, and 5 metal sided tote offered 2nd time read attempt. Circles separa ted each other represents significant differences between the totes. PO Type # of n Read Rates Probability of getting 100% read rate Time of Read regular 3 metal sided 5 metal sided regular 3 metal sided 5 metal sided regular/ 6 3 metal sided/6 5 metal sided/6 control 10 99.23 99.74 100 90.00 96.66 100 12.00 11.00 10.33 one type 10 98.33 98.85 99.38 70.00 73.33 83.33 14.00 13.66 13.00 mediu m loaded 30 97.80 99.49 98.24 55.56 87.78 63.33 48.33 36.00 47.00 full loaded 50 98.57 99.11 98.58 50.66 63.33 57.33 79.00 75.66 80.66 Total Results 100 98.38 B 99.26 A 98.70 A 58.00 75.00 66.00 153.33 136.33 151.00 Margin of Error 0.45 0.26 0.37

PAGE 71

71 Third Time Read Attempt The results of read rates if the totes are offered a 3rd time read attempt is demonstrated in Table 5-6. The average read rate with the 5 metal sided tote has continued to increase to reach a level that eliminates the significant difference between the 3 metal sided tote, whereas the regular tote continues to have the lowest r ead rate (P=1.5E-03<0.05). Figure 5-5 shows the distribution of read rates accordi ng to tote type. When it comes to reading all the products, the 3 metal sided tote has 80% POs with 100% read ra te whereas the 5 metal sided tote has 75% POs and the regular tote has only 62% POs. More importantly the 3 metal sided tote required 13.77% and 18.99% less time than the 5 metal sided and th e regular tote, respectiv ely, to achieve this performance. Table 5-6. Statistical analysis of read perf ormances of 100 POs with regular, 3, and 5 metal sided tote offered 3rd time read attempt* *Same letter represents no di fference between the totes. PO Type # of n Read Rates Probability of getting 100% read rate Time of Read regular 3 metal sided 5 metal sided regular 3 metal sided 5 metal sided regular/ 6 3 metal sided/6 5 metal sided/6 control 10 99.23 100 100 90.00 100 100 13.00 11.00 10.33 one type 10 98.75 99.06 99.69 80.00 80.00 90.00 16.33 98.00 15.00 medium loaded 30 98.19 99.58 98.89 60.00 90.00 73.33 62.66 39.33 58.66 full loaded 50 98.75 99.29 98.86 54.00 70.00 67.35 105.66 94.00 102.33 Total Results 100 98.63 B 99.43 A 99.37 A 62.00 80.00 74.67 198.33 160.66 186.33 Margin of Error 0.43 0.25 0.34

PAGE 72

72 read rates 89 90 91 92 93 94 95 96 97 98 99 100 101 3 side 1 5 sided 1 regular 1 Column 1 Each Pair Student's t 0.05 Figure 5-5. Distribution of the m eans of read rates of 100 POs with regular, 3, and 5 metal sided tote offered 3rd time read attempt. Circles separated each other represents significant differences between the totes. Discussions Placement Order of the Products in the Medium and Full Loaded Totes As described in the methods and materials sect ions the products are pl aced in a specific order inside the medium and full loaded totes. Main reasons for such a placement include both physical constraints and the aim to improve the read rates. As an example of a physical constraint, the liquid products ar e usually heavier; thus can dama ge other products if they are placed on top. Moreover, glass is a common pa ckage material for the liquid products and placing them at the bottom makes th em more stable and offers more protection from breakage. Regarding issues with read rates, liquid produc ts tagged with near field tags should be placed closer to the antenna as they have a short read range. Placing them at the bottom ensures that they are closer to the antenna. In the case of the blister packs, they are placed last with their metal surfaces facing the top. This placement prevents adjacent product tags from getting detuned by the metal of the

PAGE 73

73 blister packs since the tags below will have enough air gaps and in almost all the blister packs the tags are located on the side. Fina lly, please note that this particular placement described above is easily possible with automation during product pick-up in the warehouse. Performance Differences of Modified Totes The results of the mapping tests, Chapter 4, ar e validated in this section. By using an aluminum coating on the sides of the tote, the RF wave propagation was guided in such a way that it reaches everywhere inside the tote by creating more waves resulting in more RF power coming from different directions. Hence, the 5 metal sided tote performed better at reading controlled and one type product POs than all the other totes. However, when the tote is filled completely with the products, it s performance has dropped substantia lly. One of the reasons for this observation is due to the fact that in the regu lar tote the waves emitted outside of the tote is also utilized to read the products placed closer to the edges on the top areas. Conversely, in metal coated totes, these waves are blocked by the metal surfaces and the sides convex angel limits the waves coming from inside to those ar eas. Via the same reasoning, the 3 metal sided tote has a better performance by utilizing both the outside waves and the reflected waves inside. On top of everything it should al ways be considered that all th e results are affected by the tote movement on the conveyor belt, as the wave pa ttern inside the tote changes along with this motion. Even a slight change within each read a ttempt, such as the location of the tote on the conveyer belt, cause completely different wave patterns to occur. 5 metal sided tote is affected most by this phenomenon simply because it has more metal surfaces and thus more reflections within the tote. Therefore it di splays a high variance in overall performance due to these random wave patterns, which re sult in a non-linear (x2) increase in its performance, whereas other totes display a linear increase.

PAGE 74

74 To improve the readability performance of the 5 metal sided tote, there are multiple solutions such as using a wider bottom surface w ith more perpendicular sides. Moreover, coating only half the side surfaces with metal in stead of the whole surfa ce might result in better performance as well. This is mainly because, in the case of a full tote, th e top half of the side metal surfaces are not being used effectively (the reasons described above) and in the case of a small loaded tote waves reflecting from th ere do not contribute in reading as much. Other Application Methods of Metal Coating The metal coating can be applied to the totes in different ways as they can either be part of the insulator or the tote. For example, metal can be embedded to the plastic of the tote during injection molding or metal spray paint can be applied onto the plastic after the production of the tote. Hence, metal layer can be put inside or ou tside of the tote material ; however the layer will be more protected if it is placed inside. Efficiencies of 2nd and 3rd Read Attempts from the Warehouse Perspective The results of this chapter can be used by th e distributor on how to decide the number of read attempts that should be conducted for the totes. Accordi ng to the results for the 3 metal sided tote, 30.60% and 20.00% of the remain ing totes can be read entirely in the 2nd and 3rd attempt, respectively. For the 5 metal sided tote these numbers change to 33.30% in the 2nd attempt and 25.50% in the 3rd attempt. However in the regular tote fewer totes can be read 22.20% in the 2nd attempt and only 9.50%in the 3rd attempt. When time consumption (given in the results se ction of this chapter) and the percentages above are both taken into consid eration, it is observed that usi ng the conveyor belt to scan the tote for the 3rd attempt is not very efficient for the ware houses. Depending on the order size, it is better to do the scanning manually for the 2nd or 3rd attempts. For example, according to the

PAGE 75

75 results, in the case of the 3 metal sided tote, only 1 or 2 products are missed when the read rate is less than 100%, which can be easily read in a manual scan. Summary In this chapter, a regular tote, a 3 metal side d tote and a 5 metal sided tote were compared at scanning 100 different POs with UHF system on a conveyor belt. Most challenging POs with different sizes were created. The metal coating of the tote was textured with a grid like pattern, which enabled to use thinner insulator. The data is analyzed as if the totes were offered up to 3 read attempts. As a conclusion there were no significant diffe rences in the averag e read rates between the regular tote and the 5 metal sided tote at all attempts. However, 3 metal sided tote had always the best performance in all totes by ha ving highest read rate s and reading more POs entirely within a better time range. Finally the distributor is suggest ed to attempt up to 2nd read attempt to the totes not read 100% on the conveyor belt for more time efficiency.

PAGE 76

76 CHAPTER 6 CONCLUSION Pharmaceutical distribution chain is very vulnerable to counterfeiting which is observed very often in this industry. The main objective of this research is to improve the security of and readability of the RFID tagged pharmaceutical pr oducts and food supplements inside a sealed tote by proper designs of the tote and portal. Several tests were performed to analyze different aspects of an RFID system as well as to gauge the improvement introduced by novel contribu tions to the tote and portal design. The first test compared the performances of UHF and HF systems and show ed that the UHF RFID system is more suitable to be used in item level of pha rmaceutical applications th an the HF system with its higher performance of read ing products on a high speed operating conveyor belt, on a turn table and inside a tote. The second phase of the research found out that the use of reflective materials as part of the sealed tote provides a control over the RF propagation which leads to a better RF power distribution by RF mapping of 0, 2, 3, 4, and 5 metal sided totes. The 5 metal sided tote had the highest RF power level which was almost 3 times higher than the regular (0 metal sided) tote. This chapter also indicates that metal can be used to enhance the security of the tote by preventing disclosure of the RFID tagged products inside a tote against unauthorized readers. The 3, 4 and 5 metal sided totes can be palletized in a special configuration which prevents RF waves reaching to the totes contents. The final test includes the scanning of differe nt types of POs with modified totes along with a regular tote (contr ol unit). The results show that the 3 metal sided tote with a grid texture on the metal surfaces has improved readability of products inside the sealed tote by having higher average read rates and more 100% r ead POs with the least time consumption.

PAGE 77

77 Even though the 3 metal sided tote has the high est read rates, its shielding effect against unauthorized readers is achievable bu t not as efficient as the 5 metal sided tote. However, if the 5 metal sided tote is to be used due to offering more protection, the read rates can be improved further by using wider bottoms or side s with different angles. To sum up, this thesis proves that a security and a readability of tagged pharmaceutical products and food supplements inside a sealed tote can be improve d by utilizing reflective materials in the tote design with a prope r system choice and setup.

PAGE 78

78 APPENDIX A PRODUCTS USED IN THE TESTS Lovenox Manufacturer: Aventis Pharma Specialties, Maisons-Alfort, France Tag place: Top, perpendicular Dimensions: height (h)=168 mm, length (l)=140 mm, width (w)=65 mm Product physical type: 5. a) Sub-pr oducts (syringes) in a plastic Product RF type: RF lucent NuvaRing Manufacturer: N.V. Organon, Oss, The Netherlands Tag place: Side, perpendicular Dimensions: h=110 mm, l=120 mm, w=19 mm Product physical type: 5. b) Sub-products in a metal Product RF type: RF opaque BD Ultra-Fine III Manufacturer: BD, Franklin Lakes, NJ, USA Tag place: Top, perpendicular Dimensions: h=85 mm, l=130 mm, w=72 mm Product physical type: 5. a) Sub-products in a plastic Product RF type: RF lucent Equate Chest Rub Manufacturer: The Mentholatum Co., Inc., Orchard Park, NY, USA Tag place: Top, parallel Dimensions: h=75 mm, l=55 mm, w=55 mm Product physical type: 2.a) Cream in a plastic Product RF type: RF lucent Cardizem LA Manufacturer: Biovail Corporat ion, Mississauga, ON, Canada Tag place: Side, perpendicular Dimensions: h=95 mm, l=35 mm, w=35 mm Product physical type: 1.a) Pill in a plastic Product RF type: RF lucent

PAGE 79

79 Smart Health Aspirin Distributed by: Wal-Mart Stor es, Inc., Bentonville, AR, USA Tag place: Side, parallel Dimensions: h=90 mm, d=45 mm Product physical type: 1.a) Pill in a plastic Product RF type: RF lucent Procanbid Manufacturer: DSM Pharmaceuticals, Inc., Greenville, NC, USA Tag place: Top, perpendicular Dimensions: h=115 mm, l=85 mm, w=70 mm Product physical type: 1.b) Pill in a metal (blister) Product RF type: RF opaque Equate Flu Relief Manufacturer: LNK international, inc., Hauppauge, NY, USA Tag place: Side, perpendicular Dimensions: h=90 mm, l=85 mm, w=23 mm Product physical type: 1.b) Pill in a metal (blister) Product RF type: RF opaque Ivax Manufacturer: IVAX Pharmaceuticals, Inc., Miami, FL, USA Tag place: Top, parallel Dimensions: h=85 mm, l=105 mm, w=105 mm Product physical type: 1.b) Pill in a metal (blister) Product RF type: RF opaque Gentamicin Sulfate Injection Manufacturer: Abbott Laboratorie s, North Chicago, IL, USA Tag place: Side, parallel Dimensions: h=40 mm, l=90 mm, w=90 mm Product physical type: 3. c) Liquid in a glass Product RF type: RF absorbent

PAGE 80

80 Xylocaine-MPF Manufacturer: AstraZeneca LP, Wilmington, DE, USA Tag place: Top, parallel Dimensions: h=112 mm, l=90 mm, w=40 mm Product physical type: 3. c) Liquid in a glass Product RF type: RF absorbent Magnesium Citrate Manufacturer:HUMCO, Te xarkana, TX, USA Tag place: Side, parallel Dimensions: h=165 mm, d (diameter)=60 mm Product physical type: 3. c) Liquid in a glass Product RF type: RF absorbent Diltiazem HCI Injection Manufacturer: Ben Venue Laborator ies, Inc., Bedford, OH, USA Tag place: Top, parallel Dimensions: h=50 mm, l=120 mm, w=48 mm Product physical type: 3. c) Liquid in a glass Product RF type: RF absorbent Pedialyte Freezer Pops Manufacturer: Abbott Laborator ies, Columbus, OH, USA Tag place: Top, perpendicular Dimensions: h=240 mm, l=90 mm, w=90 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent OCuSoft Manufacturer: OCuSoft, Inc., Rosenberg, TX, USA Tag place: Side (cap), parallel Dimensions: h=200 mm, d=50 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent

PAGE 81

81 MGP Bromaxefed Syrup Manufacturer: Morton Grove Pharmaceuti cals, Inc., Morton Grove, IL, USA Tag place: Side, parallel Dimensions: h=120 mm, d=44 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent Tincture of Benzoin Manufacturer: Smith & Nephew, Inc., Largo, FL Tag place: Side, parallel Dimensions: h=154 mm, d=42 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent MGP Bromaxefed Syrup 1 Pint Manufacturer: Morton Grove Pharmaceuti cals, Inc., Morton Grove, IL, USA Tag place: Front (behind the label), parallel Dimensions: h=180 mm, l=80 mm, w=60 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent Equate Triacting Cold & Cough Manufacturer: Perrigo, Allegan, MI, USA Tag place: Top, perpendicular Dimensions: h=140 mm, d=55 mm Product physical type: 3.a) Liquid in a plastic Product RF type: RF absorbent Hurricaine Tropical Anesthetic Spray Manufacturer: Beutlich Pharmaceuticals, Waukegan, IL, USA Tag place: Side (cap), parallel Dimensions: h=125 mm, d=34 mm Product physical type: 3. b) Liquid in a metal Product RF type: RF opaque

PAGE 82

82 Tinactin Antifungal Spray Can Manufacturer: Schering-Plough HealthCare Products, Inc, Memphis, TN, USA Tag place: Top, parallel Dimensions: h=170 mm, d= 53 mm Product physical type: 3. b) Liquid in a metal Product RF type: RF opaque Albuterol Aerosol Manufacturer: IVAX Pharmaceutical s, Ireland, Waterford, Ireland Tag place: Top, parallel Dimensions: h=97 mm, l=52 mm, w=33 mm Product physical type: 4. c) Gas in a metal Product RF type: RF opaque Combivent Inhalation Aerosol Manufacturer: Boehringer Ingelheim Pharmaceuticals, Ridgefield, CT, USA Tag place: Top, parallel Dimensions: h=82, l=54 mm, w=32 mm Product physical type: 4. c) Gas in a metal Product RF type: RF opaque Equate Effervescent Pain Relief Manufacturer: Tower Laborator ies, Centerbrook, CT, USA Tag place: Side, parallel Dimensions: h=102 mm, l=87 mm, w=54 mm Product physical type: 1.b) Pill in a metal (blister) Product RF type: RF opaque

PAGE 83

83 APPENDIX B PRODUCTS UHF TAG TYPE Table B-1. Determination of product UHF tag type Products Turn table testing (average read rates) Conveyor belt testing (average read rates) Tag Type Near/Far Field Far Field Near/Far Field Far Field Lovenox 100 100 3.26 2 Near/Far Field NuvaRing 100 100 1.80 0.93 Near/Far Field BD Ultra-Fine III 100 100 2.13 2 Near/Far Field Equate Chest Rub 100 100 2.20 1.53 Near/Far Field Ivax 100 100 3.06 2.86 Near/Far Field Procanbid 100 100 2.26 3.00 Far Field Equate Flu Relief 98.4 90.6 0.80 2.93 Far Field Cardizem LA 91.96 100 1.00 3.33 Far Field Smart Health Aspirin 92.16 100 1.33 2.60 Far Field Gentamicin Sulfate 100 100 1.66 2.33 Far Field Heparin Lock Flush 100 100 3.33 2.00 Near/Far Field Xylocaine-MPF 100 100 1.40 1.33 Near/Far Field Magnesium Citrate 3 0 0.86 0 Near/Far Field Diltiazem HCI Injection 100 100 2.53 1.26 Near/Far Field Pedialyte Freezer Pops 68.4 80 1.33 2.33 Far Field OCuSoft 100 100 2.53 1.80 Near/Far Field MGP Bromaxefed Syrup 79 0 1.00 0 Near/Far Field Tincture of Benzoin 100 0 1.00 0 Near/Far Field MGP Bromaxefed Syrup 1 Pint 11.52 0 1.00 0 Near/Far Field Equate Triacting Cold & Cough 100 100 2.53 1.80 Near/Far Field Hurricaine Tropical Anesthetic Spray 100 100 1.00 2.06 Far Field Tinactin Antifungal Spray Can 100 100 2.73 1.46 Near/Far Field Albuterol Aerosol 100 100 1.40 1.40 Near/Far Field Combivent Inhalation Aerosol 100 100 2.13 1.33 Near/Far Field

PAGE 84

84 APPENDIX C PREPARATION OF PURCHASE ORDERS Controlled PO Table C-1. Worksheet of controlled PO Product Name Orientations Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Equate Flu Relief 0 0 0 0 0 Magnesium Citrate Oral Solution 45 0 45 0 180 Pedialyte Freezer Pops 0 0 0 0 0 OCuSoft 0 45 90 0 45 MGP Bromaxefed Syrup 90 90 45 90 45 MGP Bromaxefed Syrup 1 Pint 90 45 180 45 90 Lovenox 45 0 0 0 45 NuvaRing 180 180 180 180 180 Procanbid 0 0 45 0 45 Xylocaine-MPF 45 0 45 0 0 Equate Triacting Cold & Cough 0 45 0 0 45 Hurricaine Tropical Anesthetic Spray 0 0 45 45 0 Tinactin Antifungal Spray Can 45 0 45 0 0 Product Name Orientations Sample 6 Sample 7 Sample 8 Sample 9 Sample 10 Equate Flu Relief 0 0 0 0 0 Magnesium Citrate Oral Solution 45 0 0 0 0 Pedialyte Freezer Pops 0 0 0 0 0 OCuSoft 0 0 45 90 45 MGP Bromaxefed Syrup 45 45 90 90 90 MGP Bromaxefed Syrup 1 Pint 135 90 90 0 180 Lovenox 0 45 45 0 45 NuvaRing 180 180 180 180 180 Procanbid 45 45 0 45 0 Xylocaine-MPF 45 0 0 45 45 Equate Triacting Cold & Cough 0 0 45 0 45 Hurricaine Tropical Anesthetic Spray 45 0 45 45 0 Tinactin Antifungal Spray Can 0 0 0 0 45

PAGE 85

85 Pictures of Controlled PO Samples Sample 1 Sample 2 Sample 3

PAGE 86

86 Sample 4 Sample 5 Sample 6

PAGE 87

87 Sample 7 Sample 8 Sample 9

PAGE 88

88 Sample 10 Pictures of One Type of Product PO Samples Sample 1 Sample 2

PAGE 89

89 Sample 3 10 different samples were created for this type of PO. Above are only the pictures of 3 samples. The other samples are just the di fferent placement configuration of the same products as they are in a case. Picture of Medium Load PO Sample 30 different samples were created for this ty pe of PO by taking out each products and placing them randomly inside the tote. The RF absorbent products were put before the others. Below is a picture of one sample. Sample

PAGE 90

90 Picture of Full Load PO Sample 50 different samples were created for this ty pe of PO by taking out each products and placing them randomly inside the tote. The placement orde r of the products is first the RF absorbent, then the RF lucent and fina lly the RF opaque products. Sample

PAGE 91

91 LIST OF REFERENCES 1 FDA. 2005. Radiofrequency identificati on technology. FDA Consumer 39(2):37. 2 Charatan F. 2001. Fake prescription drugs ar e flooding the United States. Br Med J 322:1446. 3 Whiting R. 2004. Drugmakers jumpstart RFID tagging of bottles. Information Week 999:28. 4 Dobkin DM, Weigand SM. 2005. Environmenta l effects on RFID tag antennas. IEEE MTT-S International Microwave Symposium 1:135-138. 5 Lee YM, Cheng F, Leung YT. 2004. Exploring the impact of RFID on supply chain dynamics. Proceedings of Simulation Conference 2:1145-1152. 6 Finkenzeller K. 2003. RFID handbook, fundamental s and applications in contactless smart cards and identification. New York: Wiley. p 43-56. 7 Texas Instrument 2007. Texas Inst ruments LF reader family http://www.ti.com/rfid/docs/m anuals/appNotes/szzt012.pdf 8 Magellan Technology 2005. Comparison betw een 125 kHz low frequency (LF) & 13.56 MHz high frequency (HF). White Paper. 9 Magellan Technology. 2006. A comparison of RF ID frequencies and protocols. White Paper. 10 Chawla V, Ha DS. 2007. An overview of pa ssive RFID. IEEE Applications and Practice 45(9):11-17. 11 ADT/Tyco Fire & Security, Al ien Technologies, Impinj Inc., Intel Corporation, Symbol Technologies Inc., Xterprise. 2006. RFID and UHF: A prescription for RFID success in the pharmaceutical industry. White Paper. 12 Desmons D. 2006. UHF Gen 2 for item-level tagging. Proceedings from the 2006 RFID World Meeting, Dallas, Texas. 13 Flores JLM, Srikant SS, Sareen B, Vagga A. 2005. Performance of RFID tags in near and far field. IEEE International Conference on Personal Wireless Communications 1:353357. 14 Dobkin DM. 2007. RFID basics: antenna polarization. http://www.rfdesignline.com/howto/ 202404293;jsessionid=RTGY5NA0 M1DNIQSNDL OSKHSCJUNN2JVN?pgno=2 15 Kraus JD. 1988. Antennas. New York: McGraw-Hill. p 70-73

PAGE 92

92 16 Laran RFID. 2004. A basic introduction to RF ID technology and its use in the supply chain. White Paper. 17 Chau TC, Welt BA, Eisentadt WR. 2006. Anal ysis and characterization of transponder antennae for radio frequency identification (RFID) systems. Packaging Technology Science 19:33-44. 18 Clarke RH, Twede D, Tazelaar JF, Boye r KK. 2006. Radio frequency identification (RFID) performance: the effect of tag or ientation and package contents. Packaging Technology Science 19:45-54. 19 Sarma S, Engels DW. 2003. On the future of RFID tags and protocols. White Paper, Auto-ID Center, Massachuse tts Institute of Technology. 20 Clampitt HG. 2006. RFID certification textbook. Houston: PWD Group. p 103-120. 21 EPCglobal. 2007. Tag class definitions. http://www.epcglobalinc.org/standards/TagClassDefinitions_1_0-whitepaper20071101.pdf 22 FDA. 2001. Prescription Drug Marketing Act Report to Congress. http://www.fda.gov/oc/pdma/report2001/ 23 WHO. 2006. Counterfe it medicines. http://www.who.int/mediacentre/factsheets/fs275/en/ 24 Messplay G C, Heisey C. 2006. Pharmaceutical pedigree requirements. Contract Pharma July/August:18-20. 25 Koh R, Schuster EW, Chackrabarti I, Bellman A. 2003. Securing the pharmaceutical supply chain. White Paper. 26. Wilding R, Degado T 2004. RFID applications within the supply chain. Supply Chain Practice 6(2):30-43. 27. Michael K, McCathie L. 2005. The pros and c ons of RFID in supply chain management. IEEE Proceedings of the Internationa l Conference on Mobile Business 1:623-629. 28. Koroneos G. 2006. RFID: A slow go in pharmaceu tical adoption; proposed legislation to help fight counterfeit drugs. Pharmaceutical Technology 30(4):24. 29. Philips Semiconductors TAGSYS Texas Instruments Inc. 2004. Item-level visibility in the pharmaceutical supply chain: a compar ison of HF and UHF RFID technologies. White Paper.

PAGE 93

93 30. ODIN Technologies, Unisys Corporation. 2006. Pharmaceutical item level RFID: Battle of the frequencies. White Paper. 31. Nikitin PV, Rao KVS, Lazar S. 2007. An overview of near field UHF RFID. IEEE International Conferen ce on RFID 2007 1:167-174. 32. Wal-Mart Inc. 2006. Wal-Mart stores radio frequency identification. http://www.walmartstores.com/Globa lW MStoresWeb/nav igate.do?catg=339 33. United States Department of Defense. 2007. United States Department of Defense Suppliers' Passive RFID Information GuideVersion 10.0 http://www.acq.osd.m il/log/rf id/supplierguide.htm 34. Dean DA, Evans ER, Hall IH. 2000. Pharmaceutical packaging technology. London, NY: Taylor & Francis. p 65-80,326-329. 35. Jenkins WA, Osborn KR. 1999. Packaging drugs and pharmaceuticals. Lancaster, Pennsylvania: Technomic P ublishing Company. p 6-10. 36. Pilchik R. 2000. Pharmaceutical blister pack aging, part 1 rationale and materials. Pharmaceutical Technology November 2000:68-76. 37. Lahiri S. 2005. RFID sourcebook. Uppe r Saddle River: IBM Press. p 3. 38. Zhou X, Wang G. 2004. Study on the influence of curving of tag antennas on performance of RFID system. 4th Internati onal Conference on Microwave and Millimeter Wave Technology Proceedings 18-21 Aug 2004:122-125. 39. Siden J, jonsson P, Olsson T, Wang G. 2001. Performance degradation of RFID system due to the degradation in RFID tag antenna. 11th International Conference on Microwave and Telecommunication Technology:371-373. 40. Costenoble J. 2005. Rotary screen printing: the productive soluti on for HF/UHF RFID labels. SGIA Journal 9 Fourth quarter:7-10. 41. Bix L, Sansgiry SS, Clarke R, Cardoso F, Shringarpure GS. 2004. Retailers' tagging practices: a potential liability? Packaging T echnology Science 17:3-11. 42. Cheng DK. 1993. Fundamentals of engineeri ng electromagnetics. New York: Prentice Hall. p 304-330. 43. Reitz JR, Milford FJ, Christy RW. 1993. Founda tions of electromagnetic theory. Boston: Addison Wesley. p 441-469. 44. Sweeney PJ. 2006. CompTIA RFID + study guide. Alameda, CA: Sybex. p 5-10.

PAGE 94

94 45. Frieden DR. 1985. Principles of naval weapons systems. Annapolis, MD: Naval Institute Press. p 23-27. 46. Banerjee SR, Jesme R, Sainati RA. 2007. Performance analysis of short range UHF propagation as applicable to passive RFID IEEE International Conference on RFID 2007:30-36. 47. Hodges S, Harrison M. 2004. Demystifying RFID: principles and pr acticalities. White Paper. 48. Orfanidis SJ. 2004. Electromagnetic Waves and Antennas. p 11-14. http://www.ece.rutgers. edu/~orfanidi/ewa/ 49. Weis SA, Sa rma SE, Rivest RL, Engels DW. 2004. Security and privacy aspects of lowcost radio frequency identification system s. Security in Pervasive Computing 2802/2004:203-214. 50. Raumonen P, Sydanheimo L, Ukkonen L, Keskilammi M, Kivikoski M. 2003. Folded dipole antenna near metal plate. 2003 IEEE Antennas and Propagation Society International Symposium 1:848-851. 51. Foster BR, Burberry RA. 1999. Antenna prob lems in RFID systems. IEEE Colloquium on RFID Technology 1999(123):3/1-3/5. 52. Sydanheimo L, Ukkonen L, Kivikoski M. 2006. Effects of size and shape of metallic objects on performance of passive radio fr equency identification. The International Journal of Advanced Manuf acturing Technology 30:897-905. 53. Ranasinghe DC, Hall DM, Cole PH, Engels DW. 2004. An embedded UHF RFID label antenna for tagging metallic objects. Intelligent Sensors, Sensor Networks and Information Processing Conference, 2004 1:343-347. 54. Ukkonen L, Sydanheimo L, Kivikoski M. 2004. Patch antenna with EBG ground plane and two-layer substrate for passive RFID of metallic objects. IEEE Antennas and Propagation Society Internati onal Symposium, 2004 1:93-96. 55. Ukkonen L, Engels DW, Sydanheimo L, Kivikos ki M. 2004. Planar wire-type inverted-F RFID tag antenna mountable on metallic objects. IEEE Antennas and Propagation Society International Symposium, 2004 1:101-104. 56. EIAmin A. 2006. RFID tags turns metal packaging into antennas. http://www.foodproductiondaily.com/news/ng.asp?id=72561-crown-qinetiq-rfid

PAGE 95

95 BIOGRAPHICAL SKETCH Dilek Dagdelen Uysal was born in Ankara, Tu rkey, in 1980. She received her Bachelor of Science degree in industrial design from Ortadogu Teknik Universitesi (ODTU), Ankara,Turkey, in 2003. In 2005 sh e joined the Department of Food Science at University of Florida to pursue a M.S. degree with a minor in Pa ckaging Science. She is the co-inventor of the method and apparatus for tracking transported item s using RFID tags, U.S. Provisional Patent Application Docket No UF-538P. Her research interests mainly include RFID, packaging and product design.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101204_AAAABR INGEST_TIME 2010-12-04T15:27:27Z PACKAGE UFE0021827_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 5605 DFID F20101204_AABFFY ORIGIN DEPOSITOR PATH dagdelenuysal_d_Page_47thm.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
0ded564b209b2849f65d9610a09baa8f
SHA-1
f6edf1686a6b2081eaad2af89d8969799f9e05e3
7113 F20101204_AABFHB dagdelenuysal_d_Page_68thm.jpg
0a52172f756ad52cf70a8bfa91bae650
d7cc3fa25659f04d6d94c62e46b66a4e7117a879
25864 F20101204_AABFGM dagdelenuysal_d_Page_57.QC.jpg
8c953d653493156097040ce028bb0ef5
169647c9a92ed1ed468018dfd14bbcb7ede3277a
5726 F20101204_AABFHC dagdelenuysal_d_Page_69thm.jpg
fea5dd271f8cca93aa6d4a41d022ed5f
4aae6e909d49a5852e9b03541b4980b9bcdf38dc
7192 F20101204_AABFGN dagdelenuysal_d_Page_57thm.jpg
6b0206165a6dc2fa29863c7c6b389d72
65c63a3c304be12afc62585df09f2eb6ee989026
5789 F20101204_AABFFZ dagdelenuysal_d_Page_48thm.jpg
a4fdacf5b0b1a326e51749ae5da48043
5792cb07c488903f48290ed37f296e981899a9ee
6838 F20101204_AABFHD dagdelenuysal_d_Page_71thm.jpg
d2df046bde4e7c9d8314b7318ab91cb8
5ec820c0aafa9d140d52f19ad0d435421335725d
7346 F20101204_AABFGO dagdelenuysal_d_Page_58thm.jpg
bfabf0ccbef72e3717da8722fde709d4
84ed4d2cd9ea086b311e3fde8b63aff7179aca5a
20224 F20101204_AABFHE dagdelenuysal_d_Page_72.QC.jpg
95bfe386a998cb45019a52ffca04a3ec
58df00d543fe9e727a6b0de90c5feeeefd56250c
25699 F20101204_AABFGP dagdelenuysal_d_Page_60.QC.jpg
9c46f527af7286cdff1870b315b6f495
ce91f081702590a9c822a3cbfb1388c0a4b6776c
27253 F20101204_AABFHF dagdelenuysal_d_Page_73.QC.jpg
0fac4c93751e9484882f049bc7533675
ea8e0d6b1a884fe9abeccda890f8bb56cffc5de0
6874 F20101204_AABFGQ dagdelenuysal_d_Page_60thm.jpg
f6652acc23da15b915484d97ce5483be
1dc485d3c6a1946a1e36759a8ab223b1c26e7321
26635 F20101204_AABFHG dagdelenuysal_d_Page_74.QC.jpg
ffab9bfbe5d69a42ed23cd01bb29e5f5
2db98e81a489073fb2ef7f9856e3c04bbddbcbcd
25324 F20101204_AABFGR dagdelenuysal_d_Page_61.QC.jpg
735fcf17816e9a671d58deda3aecae55
1f15151b37ae80af3d11d8657700c5c3fab979bd
14890 F20101204_AABFHH dagdelenuysal_d_Page_75.QC.jpg
b5455e93ed8e3b58428cdb26f923330a
78cdbeb8eb615d8f31d1c243caa7c4d65f9b11bd
5628 F20101204_AABFGS dagdelenuysal_d_Page_62thm.jpg
5afce30a6dbf2f954c6cb76684cc79eb
39fe5965628ce7a78986f142eb2a1b31fb8ff95f
4470 F20101204_AABFHI dagdelenuysal_d_Page_75thm.jpg
f85313e484a638c94eef40202f641108
d9cd327245c196f9538d5cf2d9e912358e09a183
23649 F20101204_AABFGT dagdelenuysal_d_Page_63.QC.jpg
f97536d9b482cfad2c6bbe197de47080
54d9d5bd90d5f9d81323d368ff1a114cb5483489
7108 F20101204_AABFHJ dagdelenuysal_d_Page_76thm.jpg
982e4067a029aee563840ecfc2d2b005
6e27e4f3123bc70214c0361a331c132aba1fbe3c
20272 F20101204_AABFGU dagdelenuysal_d_Page_64.QC.jpg
aa36285ce0e5a94f3899d39b8b1296e7
795ba735cfd386a404ab0d14a4eb2fba971b9cce
9976 F20101204_AABFHK dagdelenuysal_d_Page_77.QC.jpg
35a358b22203000fcf9269f136df0ebf
5a93c605f47490c69b836185b1de4a9ed071a483
5944 F20101204_AABFGV dagdelenuysal_d_Page_64thm.jpg
c8c41a974242358d4d2ca1bbca17f36c
b59e21256090cd5f0dd4cc3c148ef66d5e7b698d
3059 F20101204_AABFHL dagdelenuysal_d_Page_77thm.jpg
239c761001b05ef953dca40aa71fb42a
32cfde35bee00e4a95ec9f3bd2e0b7cab4aa34ff
23579 F20101204_AABFGW dagdelenuysal_d_Page_65.QC.jpg
fc72c21b90ee142da121630b31c73626
0f0849db9ee8da048dddc6eac07d2ab3e7670470
6587 F20101204_AABFGX dagdelenuysal_d_Page_65thm.jpg
e28dd5731f3b3c27c047ee689f4a7c58
d6b290e509a376bce2a73e50091023300418f0f5
4549 F20101204_AABFIA dagdelenuysal_d_Page_87thm.jpg
ab65f61477b8e14cd87010445d2f8313
bb17b2a162664e239242c7a6830c9654d79c0419
19636 F20101204_AABFHM dagdelenuysal_d_Page_78.QC.jpg
01dcb72c36041a4e393d555e8bff7299
c043a5ff568622ecc6e2dd9aec616e1ad4759da2
19276 F20101204_AABFGY dagdelenuysal_d_Page_66.QC.jpg
d9e67c7e62ca8f109c1b5be2c10a9635
5fd1fc87f09a36b94b3f8ea5bb10793098295e0b
13456 F20101204_AABFIB dagdelenuysal_d_Page_89.QC.jpg
253abe48dad5902056dbf231158f1b76
3dfa2f6fb95fdb844bd7145084d2f83a17b67b97
5957 F20101204_AABFHN dagdelenuysal_d_Page_78thm.jpg
b483224e855aa69d71cc48780cba4415
ea48b06c3a58cd99aace534dcfed6cdf485085cd
5340 F20101204_AABFGZ dagdelenuysal_d_Page_66thm.jpg
876fc3a17ab4f16548880f49abce5528
7b7fcc427b725e508b76af36795bee0d133ff77f
4353 F20101204_AABFIC dagdelenuysal_d_Page_89thm.jpg
bb9a231166c0713a9e8af84bed8395ff
718989cc5d61667a2f439d10b07676a0b3f807b7
19022 F20101204_AABFHO dagdelenuysal_d_Page_79.QC.jpg
22bbedfa387b412154a9acc1f2ca9a48
253048653efc00e0608a883d56f403014426b713
9702 F20101204_AABFID dagdelenuysal_d_Page_90.QC.jpg
cb0e861e2a0b5cb893f5875ff45e8da6
74e9700a54f3f95afc2686b14a0777deb1191dec
5960 F20101204_AABFHP dagdelenuysal_d_Page_79thm.jpg
5e69df80cc442c1dffd5d624e9bf8fe5
b1d13c820912641b25e1ab0f3b8148318b36b4c9
3246 F20101204_AABFIE dagdelenuysal_d_Page_90thm.jpg
9a27f799421cdc804215e53db30e6b67
88f20e12e3acbffe13609f51806a12290434b655
18441 F20101204_AABFHQ dagdelenuysal_d_Page_80.QC.jpg
876846ef9583161516444f20756317f7
1e3f4f36467ad2c481ab480dda12377c80a90878
23884 F20101204_AABFIF dagdelenuysal_d_Page_91.QC.jpg
b8b8dfa1852d5869672ddef21462994a
02f3ec5a7dfa9a4b868114bc3a604190ff0adbd3
5569 F20101204_AABFHR dagdelenuysal_d_Page_80thm.jpg
a03e092f0aa43213ef69384e02419c85
4549d3ae5d476a8aff8a0331bc538d5577370374
6946 F20101204_AABFIG dagdelenuysal_d_Page_91thm.jpg
80b1e044791d565fe08337aef5e50281
3920cf04fe0402055b35b01fb068bd8e67d6210a
5739 F20101204_AABFHS dagdelenuysal_d_Page_81thm.jpg
86f47e0ef97cf28d66d81587cde5c872
4d38c16273f01714682f850e05b320ebd73c5d27
23109 F20101204_AABFIH dagdelenuysal_d_Page_92.QC.jpg
a76ff9c7df6795195b782f42046a7772
b19caf0e05b9c377e726a2587c38a8cc1e34d9a3
16836 F20101204_AABFHT dagdelenuysal_d_Page_82.QC.jpg
012f4be0536b9a590c628d585aa8de6c
36e3380dddba2c31fcf83d774f9ca58270cfd3c4
6785 F20101204_AABFII dagdelenuysal_d_Page_92thm.jpg
a637c7aba3736634dbb384823188eb9d
ec6c183534b6144d2b318e77deff95130499a182
5253 F20101204_AABFHU dagdelenuysal_d_Page_82thm.jpg
4459e82faaed822e6823d3dafb180ece
eaeb5026dc41589d08126b4485ae26e1795b75b2
25611 F20101204_AABFIJ dagdelenuysal_d_Page_93.QC.jpg
f74c462b27dd31d48004e532da4b6492
1cc4b93bc2cd05d59b7a2ce74c1608f322b5c83c
6850 F20101204_AABFHV dagdelenuysal_d_Page_83thm.jpg
f1c1660ac5bdf8fbc7675a2dba8a899f
e3f99dc8da44724b068d0494db2e8da385010c0b
7048 F20101204_AABFIK dagdelenuysal_d_Page_93thm.jpg
8f459b6f8f180f2001114901a8fa496a
74d88408376130b3c58081853f05d2c76723a584
21473 F20101204_AABFHW dagdelenuysal_d_Page_84.QC.jpg
d3d30bde6dc3a9055ec3e2f170840770
17b090bdbcf82bd5f12cdb5c4c496552c70e6c63
24277 F20101204_AABFIL dagdelenuysal_d_Page_94.QC.jpg
30ef7cbac8e7f7d3efdbf6b379a39aef
ed740dce84c1dc2ba4454e51f85ad173194b6803
13919 F20101204_AABFHX dagdelenuysal_d_Page_85.QC.jpg
4d587de38c96e3622ccd444f508d3410
402207f8dfc383d3b76afc9efa3bb5479dbef9a7
9841 F20101204_AABFIM dagdelenuysal_d_Page_95.QC.jpg
0935623c6326de70859ee23abf032e5c
649c1309351524703326bf5f99e9429952f7ee0e
4644 F20101204_AABFHY dagdelenuysal_d_Page_85thm.jpg
dccd2ad5ee4d06abe36cceca6fb6baa5
339ca27995da1db0a57da04144cfa38b1be22cc4
13085 F20101204_AABFHZ dagdelenuysal_d_Page_86.QC.jpg
bb592056f3e7ef1a50f20f07699b0ef3
ea90a7268687b5b4e48ffaeddee92e6daa4e5299
113796 F20101204_AABFIN UFE0021827_00001.mets FULL
2fd9bb5fbd437d157c739834e18d0820
ba0ffdc73a343218d3382344406f21c2e8e37ca5
801579 F20101204_AABEJA dagdelenuysal_d_Page_42.jp2
e93b231bfd7fe12896d98b8d7178ed9c
0e113a0eac3401457ce6dc6b2137e4fe71e243b1
62839 F20101204_AABEJB dagdelenuysal_d_Page_18.jpg
52625017e05a621e1bfd9213de378c7b
ac8e3c5ec447cae8506976e8c55aa20619fefeca
25271604 F20101204_AABEJC dagdelenuysal_d_Page_54.tif
e602911809af4d7592c021a67b8e4321
874d825944c91c1fba2d00757ad6b9ee7f1a80d4
7582 F20101204_AABEJD dagdelenuysal_d_Page_73thm.jpg
1fa1c5025f09df5a209ea8a3ef07f693
40f5dec738c3f1ba7fae82ec1ef117c2e223e92b
195760 F20101204_AABEJE dagdelenuysal_d_Page_23.jp2
70f2957d54c33b41178ca8071193e75b
1cc5db6e3967215dbbcaa701090f28df53051fd1
940455 F20101204_AABEJF dagdelenuysal_d_Page_35.jp2
c81b1e1b917cc75dc301f565fbf17711
362893d44a14011d0b677adb6b9519aa440c3f66
1051952 F20101204_AABEJG dagdelenuysal_d_Page_16.jp2
389bfd9fe3f391a26d3368c8bace383a
3c36f90907142f63f7a9cb40aa47808d1f37dedc
1053954 F20101204_AABEJH dagdelenuysal_d_Page_66.tif
983216410f8412a715355d98789d35c6
4bee6ed6d2a463f7d04d043b03671a2b16b7acc1
26365 F20101204_AABEJI dagdelenuysal_d_Page_30.QC.jpg
73c55f4d1347b593f1ef2cf8438cee6c
8ddd2108cb1275aaa6f8674b4fae21b4c734439f
11053 F20101204_AABEJJ dagdelenuysal_d_Page_52.pro
8cf14226bf50f15364aaa1c86e193409
8997beae2287af143668a0d7d0466e6542f6ce24
1847 F20101204_AABEIU dagdelenuysal_d_Page_14.txt
fd84d6a7bfafd18835002993f7daf81e
dab97c777054811bf4fa29f56a60bd56eaf29946
75896 F20101204_AABEJK dagdelenuysal_d_Page_29.jpg
fadfef6cbbd444dcfa493a4e0b931dfc
b9bb28424f19b51718e6366a5e8f326f28775a2e
F20101204_AABEIV dagdelenuysal_d_Page_29.tif
d3e31a34277c335891ab91eb215da465
1d9d3735d8f8c55eef0d61b1547c69dd588fdef6
3472 F20101204_AABEJL dagdelenuysal_d_Page_33thm.jpg
3aa2ce6ae9ccca8dc8059d44567955d1
6a8158226adc59d065e6326cf4a9fbdf9db33158
26743 F20101204_AABEIW dagdelenuysal_d_Page_19.QC.jpg
6bca054bf7f6e145b3aa5e683f5ffddc
3f5b170f1f5b5bf47b54fa9186f219eff84fcfd4
1015346 F20101204_AABEJM dagdelenuysal_d_Page_65.jp2
e4b38495a193629614f16bf010f1ccbc
2f5e986b169b91af9cb67c1f4d16fa09cf572cbd
16487 F20101204_AABEIX dagdelenuysal_d_Page_40.QC.jpg
cfc259fc4e6f82cf6253e4eff2b26be7
6bc4edaabe6cd78839273a601ee21dbe7b776a28
F20101204_AABEKA dagdelenuysal_d_Page_10.tif
fd2f1c3a11582962420003dc3ca0808c
09c3fe157eecc2eb6817b2f3034c103cacc82ac2
50974 F20101204_AABEJN dagdelenuysal_d_Page_48.jpg
af8a19bce41fd9b9d1e6fbb17a2595c4
34c43c2d59bb65d1d863c6b4d0fcf0d7501563b8
52512 F20101204_AABEIY dagdelenuysal_d_Page_35.jpg
87252dea66a16151b39412b95edf8604
5f4341dfced7f52e175a010ed5a8eee907a7b45e
1051941 F20101204_AABEKB dagdelenuysal_d_Page_91.jp2
f73110b524ddd80cb94ec5f7451a6cdc
6191c07cc67725ae85e1692c9115d2ab6da5002f
925677 F20101204_AABEJO dagdelenuysal_d_Page_81.jp2
424bacf01cd6bf1acf966a47bcfa3353
4f56d9a40a3fb2dc81aa913469aef5667d0bdc06
33205 F20101204_AABEIZ dagdelenuysal_d_Page_78.pro
416b4f0c75dcad1b3be710c8eb3e6e5c
450f0c039b5ae871a63aca626d470d20e78d1547
24303 F20101204_AABEKC dagdelenuysal_d_Page_45.pro
666778dc79b3451aa3e19c3f91595d5b
782bb57cb86cd11972d094617ff38fb5de9590cc
F20101204_AABEKD dagdelenuysal_d_Page_86.tif
2927d339e0c678cde8ab7def5ddb37aa
fc6a5cf0d8e6ca287dd3b6e8f115795b91dce949
F20101204_AABEJP dagdelenuysal_d_Page_56.tif
f17fded8082d920b468be373f0ea1f63
a9587f9e19a4c83693b0c538d8a80a8a78033a3d
25819 F20101204_AABEKE dagdelenuysal_d_Page_58.QC.jpg
b53c82bcb52927e9162cfcac5b373d14
84cecc3fd28cf1d8f76886ea5e2d554560f182aa
1051976 F20101204_AABEJQ dagdelenuysal_d_Page_36.jp2
3c1cd7b24a4a1ea1ba1b2d999046d063
293f5976a0d054c74e8e7d9e4bc09f55ae46159f
81099 F20101204_AABEKF dagdelenuysal_d_Page_58.jpg
e13e7a749d151819654721629b6b747b
9f2daeeda7f71e558018b7c8cf3af8b3d7f77915
F20101204_AABEJR dagdelenuysal_d_Page_11.tif
c4ad66841fbc9df49afdf7036e45fe9e
e4d3235f87fccaf298df9dbd4ccaa758ad158a81
5678 F20101204_AABEKG dagdelenuysal_d_Page_46thm.jpg
40c98f04e4496b8af7851f69d259e129
61f72587f05ef5dcaeeaa1c78992714c13990809
62845 F20101204_AABEJS dagdelenuysal_d_Page_79.jpg
e455794ec86150d854cda3203129ccc7
4c58c0049fa162f9602fd795969c71966c32d859
F20101204_AABEKH dagdelenuysal_d_Page_87.tif
235fcdeee5f813a08f3600546569362d
1a2cfb665bcfb000acfdf0a33b15f35d30a7256d
24497 F20101204_AABEJT dagdelenuysal_d_Page_20.QC.jpg
f5031648d6ee20746547c1945833613f
1abde4808bf9267baa8e559e1d890e47ebca8557
767127 F20101204_AABEKI dagdelenuysal_d_Page_04.jp2
adaa5d08f8e493e506fb3ce309dd8430
1ac36c80c530c03df8a25670bb6f45867d9b432b
788908 F20101204_AABEJU dagdelenuysal_d_Page_28.jp2
5adb44dbb5d857d4ea5bbc26757a3624
8757af564a180cc449914a60e57ca55756295006
4442 F20101204_AABEKJ dagdelenuysal_d_Page_88thm.jpg
c37016e4d138e8fb45668dd614855bee
c0873c599b7d5c5a4f84a935d05e5bc2372dc0db
1262 F20101204_AABEJV dagdelenuysal_d_Page_37.txt
b46f3fd3052d59abbe7b6dbcb0ed0c0e
968c0f748b7010222ebbea233d4a917e96776a83
19458 F20101204_AABEKK dagdelenuysal_d_Page_46.QC.jpg
f0280e6964a1c170185182910302f134
767c0e2e5cd6c339514bc62a7f9cc599ae701ed6
F20101204_AABEJW dagdelenuysal_d_Page_44.tif
16979f110a3b8d5194c8b713ff2a4f07
6b5cfe12de4d5e0093c565dedc9f20940aa91646
70585 F20101204_AABEKL dagdelenuysal_d_Page_49.jpg
ffb2a150cc6f72f1d964d89dfeddf033
97489fc88a7b612fd18abaf6488c7c9b817b70f1
1979 F20101204_AABEJX dagdelenuysal_d_Page_27.txt
ce4b8106fb494c2276771111260525a0
9d77b30a03327b3816d9f33cbbf834a739329b4e
785 F20101204_AABELA dagdelenuysal_d_Page_56.txt
42d44f8310ad504aba9d7aba2c3912db
5fa20c48d3e69399c73435604bf3b9b5bcf8b32a
1011 F20101204_AABEKM dagdelenuysal_d_Page_45.txt
b038125ae22ec10d5e2714754881299f
81d0d8e14c75654169e22f7c3d5a065e17296b85
30245 F20101204_AABEJY dagdelenuysal_d_Page_10.pro
92b3a8ad997ec4633f61fba6023a9035
7b4a005363aa0cd60e9b26c33cbd0e86cd411fb2
23061 F20101204_AABELB dagdelenuysal_d_Page_26.QC.jpg
3b2c97e2142040ae26a448f53c2fd38b
4d301da04eb91f54bf66d6900f60b9022343567b
16026 F20101204_AABEKN dagdelenuysal_d_Page_50.QC.jpg
2c495fc7b8636f5c260695e9e929bfd9
b18f23150d1d49eedc9f649b516a2fe1f48132cb
730981 F20101204_AABEJZ dagdelenuysal_d_Page_50.jp2
534701d81e03d34bec34d99a68c6daf4
ac963bfa802dba160b710c8e26a0da5cf499dee8
1222 F20101204_AABELC dagdelenuysal_d_Page_10.txt
144c77bf67f9accdb943986b6f796b58
0afec116e3b6b740c3d66b03f8b4e3d8654bfe5d
29578 F20101204_AABEKO dagdelenuysal_d_Page_79.pro
4deed8bfd982f1e90014b4debbc80b92
c5af466bc6a7bbf34e7848b852356b48180a9053
4744 F20101204_AABELD dagdelenuysal_d_Page_39thm.jpg
893cd266387f31717d10aabe000c4121
fcf0df1bb3e0a4ebf94032e1bf760e95e03e7257
23551 F20101204_AABEKP dagdelenuysal_d_Page_34.pro
0d4d75ac81df19448ac9235d1093f679
2ace620ee6bf48e8889cc4b4771652169f43f2e5
72918 F20101204_AABELE dagdelenuysal_d_Page_26.jpg
895c4fff07120135ea763bb582bbfe86
7d72596df4a3b26d6913ef28fcc73621469a0846
23859 F20101204_AABELF dagdelenuysal_d_Page_41.QC.jpg
b551c5f63c5d3da495a4aeb76f1c252b
dfafcc78230247abd4855db0b0ff4b3cb1cf1792
F20101204_AABEKQ dagdelenuysal_d_Page_75.tif
fe148720e27c93bcb3567b8218175311
2f32d8bc1709cc5993ff7e7c325b8d51a2d3650c
17863 F20101204_AABELG dagdelenuysal_d_Page_18.QC.jpg
f9f88f5681f39708329a5a6469f1542e
092650bd6af1334f37c58e750b52f65586dbceb7
F20101204_AABEKR dagdelenuysal_d_Page_78.tif
80ff3ba396f95b28436c77f358f80da3
578beff89664907387a13e4bb15d84fd5770023f
1641 F20101204_AABELH dagdelenuysal_d_Page_66.txt
9ad8b6ded1123743408e9e1245aa238e
3c941f77ac14baafbcd3dfb0a22fefc8553f4e1d
9375 F20101204_AABEKS dagdelenuysal_d_Page_33.QC.jpg
5b3d3657c533fbd192763289050248c5
2031ca9bf26337f2feaeeb9e53ba9ec8c123b730
80912 F20101204_AABELI dagdelenuysal_d_Page_57.jpg
28679d3a1f521705d7d92090dea283cf
6ed38a175ddc1c47d6ad4963608b31397a187fbc
65411 F20101204_AABEKT dagdelenuysal_d_Page_72.jpg
7d55cc77d418e75f3a0751a2e790c374
08d91eda491e9efcc090fef42b465e22632d27ad
3289 F20101204_AABELJ dagdelenuysal_d_Page_03.QC.jpg
985e7ebda385ffd0d285884caf8fcdb9
a2104e6ed0f3d5ed501aa0865f3e0ec4cf7516b2
6634 F20101204_AABEKU dagdelenuysal_d_Page_07.QC.jpg
1b1d52d5ac1bfd4b72d9b7ed291ec001
5d2579895d38b4987551e0a5bb7393fb9e7d3609
45667 F20101204_AABELK dagdelenuysal_d_Page_26.pro
22f0c1479cf2ee0946aff2ca4e711d5d
a9f06f92a0faefa33abd102021a20b643154b811
1051868 F20101204_AABEKV dagdelenuysal_d_Page_78.jp2
716148f0addb58f1a12e5d2a71d03575
b2b9d49740108e2024dc83327cf466828a39531d
F20101204_AABELL dagdelenuysal_d_Page_77.tif
263bb3af2d0bd2e869b73dcab9e3cb2f
bfb0125308e9314f934a55b2594694b7cc8094ac
2092 F20101204_AABEKW dagdelenuysal_d_Page_60.txt
2bcb1f0c52f8e554378cce7a4775380a
7d23ffc6f5faa5ea6567ccf14d40fc7318f27c78
20258 F20101204_AABEMA dagdelenuysal_d_Page_23.jpg
90a676d14adb271ae26ed82e103f3227
a31eaeb4a464bb80e3d012297f5b4bf1251af164
83901 F20101204_AABELM dagdelenuysal_d_Page_22.jpg
bc500bee61b763d1ae7285287a60af85
758180886d9d4dc7bbf66cab3c09e7d9afe369bf
7034 F20101204_AABEKX dagdelenuysal_d_Page_74thm.jpg
f9b078ca5162d32bd37520254193503f
da21ae3b76e02016a51f701585395174e7ab175b
F20101204_AABEMB dagdelenuysal_d_Page_24.tif
f4306f97a57fee33f17d5b329850984e
db707b9ba11ea769a7d0fdc77ca11f0590903d47
61786 F20101204_AABELN dagdelenuysal_d_Page_64.jpg
bc66ec7ecea75b062a1ef32f396e41d6
1dc0b2351dd6e0a7418929659a5bef8b2a69e1f0
76395 F20101204_AABEKY dagdelenuysal_d_Page_32.jpg
87ca2d309c3a09331ecd724a71a45409
a0c02f1356ac78cb2816c8918b04eb5409a29409
40670 F20101204_AABEMC dagdelenuysal_d_Page_86.jpg
71c7e30bee16b152a093ff69366fe7b9
b953e510ea03610e5222bb8f6cd1478c0bedd1a9
F20101204_AABELO dagdelenuysal_d_Page_58.tif
bc3c042934ef98ee8f85cd2075cfc570
446670569e3c94eb7e0516555e20eb4116baa195
14166 F20101204_AABEKZ dagdelenuysal_d_Page_54.pro
a500063e309b65b1eeb8e116de085afa
ef2e4eb11d41c39222f28cbaca1e09fd0cf45483
1051974 F20101204_AABEMD dagdelenuysal_d_Page_44.jp2
2dc3974fd5f96ce898e1a9944eeb7f74
87a42f68252eb83107b25f0b4bad9d0c8b8c7ac8
3437 F20101204_AABELP dagdelenuysal_d_Page_43thm.jpg
eb51556fd5402ad9ff4721509dec936b
3f33c0900108e955859e8c0161638d698863dd7a
F20101204_AABEME dagdelenuysal_d_Page_67.tif
4d4b4835d0d39144f3e30e316d82a961
e02c276793588f97df00f2e4635270753133a03f
998912 F20101204_AABELQ dagdelenuysal_d_Page_89.jp2
a6a336b0e4b49e24d806c0d14415fb7b
22131614596bee7572d66cd4925548bf79ceb8db
18572 F20101204_AABEMF dagdelenuysal_d_Page_62.QC.jpg
b64bdccda3a836becbfafdbe2107938d
496b78a0411ebeec2e55a6a6e52dece39e6905ab
18837 F20101204_AABEMG dagdelenuysal_d_Page_81.QC.jpg
5e29797e56c6e6d64b317520df87d3a2
dd8706c43cb45f11efe87ac5cae1cd3fb60f3609
515 F20101204_AABELR dagdelenuysal_d_Page_01.txt
ffbd5ab053a3c38ef29895540840efa6
e062e700e886a03bd1276b19c7f155c93162517c
60594 F20101204_AABEMH dagdelenuysal_d_Page_69.jpg
c700df2fa7989f5d6351edea75c06cde
0ab49b35c5e538ef65ed28153703ac09f4c0674a
5315 F20101204_AABELS dagdelenuysal_d_Page_04thm.jpg
373b1084aed4b3bff634afe9bfbac95f
9c76edcdb4ffb9635a75613d90752ef223ca759a
23419 F20101204_AABEMI dagdelenuysal_d_Page_35.pro
019ceed5f93dcdd8bd0a5bd3c386be01
ea2ac1c8a0a5d73f9ae327852963effafb9f2f81
3003 F20101204_AABELT dagdelenuysal_d_Page_54.txt
7e4153639af4d46df50e4c6d884d7fd3
97dae0d39e6f4fd78602c09d9c8f9008e7035fa3
74957 F20101204_AABELU dagdelenuysal_d_Page_21.jpg
040fc70128379f40c434a54de7cf081e
1da9ee90caa615aaf232a1cf936d853b3e8211dc
647 F20101204_AABEMJ dagdelenuysal_d_Page_77.txt
321993863b2f8cb3836d5b32488a8171
82db12ac3036ec15b45077cbd9897be295f6fdda
F20101204_AABELV dagdelenuysal_d_Page_45.tif
46bc77d49338ce7c8c0b1b04c3e9fa5b
678ea61062618602bd00ea5d8e77df7897c66259
1340 F20101204_AABEMK dagdelenuysal_d_Page_03thm.jpg
ba490b7a457161a58a3cb917a8b84dbd
e92585075989788c807c85c551942c308b3877cd
27163 F20101204_AABELW dagdelenuysal_d_Page_25.QC.jpg
ef5205cdbb00e705621ba972bd0c2711
f8a1fa7d346006debb854d1b887465a50b6ec25d
7213 F20101204_AABEML dagdelenuysal_d_Page_13thm.jpg
04c5d9611c2b58476be1fb424b73d690
154d76d73bafc27115faa9be2ae058cd6e8b4425
1517 F20101204_AABELX dagdelenuysal_d_Page_12.txt
d5e4d2f7fd594491907018e1f1f2648c
d9a5d1b7f35d72da53d2e2e0dc9cd2e92e313b0d
1033695 F20101204_AABENA dagdelenuysal_d_Page_80.jp2
537a2c513716b27babc72648d241daec
bac20027ab7705f7eadae007e94d9bccc30f799f
50898 F20101204_AABEMM dagdelenuysal_d_Page_39.jpg
82258e3a87ed24084d0cb3553e4eef37
ed1a5c287f59be5b1507c3de049133eaa626aebf
6273 F20101204_AABELY dagdelenuysal_d_Page_63thm.jpg
3002f7cb5bce59346058739a70c26dc3
dd7ae606fe7186476bb60427b58dec1166af7983
864145 F20101204_AABENB dagdelenuysal_d_Page_12.jp2
1e013b6ad28d700321e77121a3d02c84
857c2aa2a829b934652f6d9cbf63b09adcf71a72
1051949 F20101204_AABEMN dagdelenuysal_d_Page_09.jp2
6a7ca172225fb3c230ba0a86bd82de17
e6f29d773be4f799e17301169b6db2b30d13091d
19354 F20101204_AABELZ dagdelenuysal_d_Page_52.QC.jpg
081b08e590422195788853b08efb58d7
b75236be2ac2bfb287ad4992c975a14d886c4495
25107 F20101204_AABENC dagdelenuysal_d_Page_43.jpg
9fb3090073fdd82fe5b9e06a1287249e
b8ec67fd817ed8d23ff5552f39b76b5201805e6a
1957 F20101204_AABEMO dagdelenuysal_d_Page_46.txt
f0d14df7ce0ed225f087daff2217592e
3532bac729c37fd98d635be93a87821ea5d8bb18
867283 F20101204_AABEND dagdelenuysal_d_Page_45.jp2
b67009a0c2264cc871e2f9df08afce5d
1e3650ee780808cfeed9042c22a46d4f61f340ab
1129 F20101204_AABEMP dagdelenuysal_d_Page_44.txt
905d3e7c10d07f79fe43858c74a7898b
b2a8d60ac7dc8ccd67c19c5da2c101db9954f35d
270932 F20101204_AABENE dagdelenuysal_d_Page_01.jp2
8ed429f9063caf5333138e6a08c00966
c1afc6b66f2968afc9ec284606f08283609bdcfe
60974 F20101204_AABEMQ dagdelenuysal_d_Page_80.jpg
600f99fb0c5da07ea3b574e3fb9c89c0
d617d98751820edeed96da23681ff0f6cc910a15
52645 F20101204_AABENF dagdelenuysal_d_Page_54.jpg
ce0c5ede1643b34fa308dde53debbe0a
121331390fdd5ae10ccf1568768f444b93508e93
F20101204_AABEMR dagdelenuysal_d_Page_43.tif
35cfee431674ed49657ece6d35ea08da
f6d38c134cea7e6dae08bfb4b0bd64f9c664d454
49677 F20101204_AABENG dagdelenuysal_d_Page_27.pro
7548d232fc920f9a19cf66f9607c092a
a9910738949b46f9930520dd344df30b96795207
2128 F20101204_AABENH dagdelenuysal_d_Page_13.txt
47a5327b5ebd2cc96adb2f2609aea8ba
376009069082a5bea372de6bf011af04c5a3c66f
5888 F20101204_AABEMS dagdelenuysal_d_Page_67thm.jpg
13c6c28b217d175327c7f5a6921998c6
3a31343f30b9bf0eafb98bff7cffb2c4a6d80d61
F20101204_AABENI dagdelenuysal_d_Page_12.tif
675475c8a1fbe9e9fa0c30f645c65d74
68902e7cf398cab1fc83f7fe556f2bb5b69f9daa
4968 F20101204_AABEMT dagdelenuysal_d_Page_31thm.jpg
cd273ed7af3ebb0d6b130898bc0be530
1d07be8b7016edba06832d034870c8403c7ad7da
F20101204_AABENJ dagdelenuysal_d_Page_69.tif
1eb0fab301d8320cb3d66a77791aa295
893b79777c17cdc546168590b663561be8caeb47
26914 F20101204_AABEMU dagdelenuysal_d_Page_22.QC.jpg
1a5f03398f13d9dce611047e473a2860
19bbe62038717ed21af1cd96c6473a68d8aa9d19
5800 F20101204_AABENK dagdelenuysal_d_Page_70thm.jpg
c58fb600169d66e389429873830729bd
9f9d26fb26613b875dd71a0a4eeb8ee4ff76abcd
25377 F20101204_AABEMV dagdelenuysal_d_Page_76.QC.jpg
f8d944ae5598ee26fcc106e0b9118be5
084b024bf4f775ad3c02836e9dfb397e47109b71
793235 F20101204_AABENL dagdelenuysal_d_Page_59.jp2
4f0d6877846a354b3e8cf01184edd100
9617b8361255bfa96b1465811009ef660cc16b08
5416 F20101204_AABEMW dagdelenuysal_d_Page_54thm.jpg
b28a47dcd7a9ec1c6f2d9493a8ce4325
2e0ee09464c0317e538818a65eb360f2125a017b
6915 F20101204_AABEOA dagdelenuysal_d_Page_27thm.jpg
bdf7249d8e274656af378c74369e5e36
12e4a3ec4a6d175ae33c61002a70df23501dda82
18992 F20101204_AABENM dagdelenuysal_d_Page_59.QC.jpg
72dcad61d75968829c7b46afd2050eb4
de33210133f2039a1b4a486a024ea64dd4fde892
26966 F20101204_AABEMX dagdelenuysal_d_Page_75.pro
fa5a60fd88cb7337b32ada78b9a17d33
856a07fd4ff7da65010f1be02c400435a92584d8
77812 F20101204_AABEOB dagdelenuysal_d_Page_20.jpg
9cec1003a83eb2b5f0f6b704dcd1c6d3
4f0c83cee6343bac9090d523ec79eddca237590e
81864 F20101204_AABENN dagdelenuysal_d_Page_68.jpg
6d8473dc502bb2805fd73e0c972f5017
6d9dd3e95e81fc616f19e6a8e7a7e84d007c52f8
1051980 F20101204_AABEMY dagdelenuysal_d_Page_85.jp2
fa888327333713555b35846b355149f7
18b697cc7ad36babd88e747f69967f68e481dc40
52194 F20101204_AABEOC dagdelenuysal_d_Page_30.pro
054d97038841782245b8457a820a8d42
532bf8bc6d68dc9e4190dfe841e3710195e5dc20
2225 F20101204_AABENO dagdelenuysal_d_Page_07thm.jpg
915313bc63281b32e1f2e43d75587d3a
588c27cc898aeb7a3c775a02ccd987b9331258a6
81745 F20101204_AABEMZ dagdelenuysal_d_Page_38.jpg
ee296ae8fbed951ef8bd26a305f86a4c
f2543af14df280876bee7b887546a4e83e287b90
F20101204_AABEOD dagdelenuysal_d_Page_19.tif
fffed4a6b8d3c11c1b4894e05b96858c
bf0f90c95ff1a5f25c998fe2af1a4da8188dbd53
45726 F20101204_AABENP dagdelenuysal_d_Page_75.jpg
150dc0959f6a5ed8a7f6ca6ba3950e61
d521a8e5aa36190eda877516c61445fbdf23183a
27256 F20101204_AABEOE dagdelenuysal_d_Page_39.pro
1c07f4e060d975b1079394c3da4fabf7
3fea742dd39da9ed558b81a90d00ab049ccc33c7
985273 F20101204_AABENQ dagdelenuysal_d_Page_82.jp2
e83d2abd3f0b6d481d23197eaa7dde52
b5a5cd08eded6f68598caa0eda8b92ed2f253316
68745 F20101204_AABEOF dagdelenuysal_d_Page_84.jpg
0965822a46fa43ce36c5b71f9fe8a983
5b7b75ac845aa83eb2a3237e336dee67d0fc3443
1308 F20101204_AABENR dagdelenuysal_d_Page_79.txt
4b3453cac242620072d6c57c7f7f3278
ca0cdee528b40501aad1a387ca910eb19fa52f0c
7247 F20101204_AABEOG dagdelenuysal_d_Page_25thm.jpg
eb894ec6b6cb43ed19d7fe11fe46e8b9
67c0721edd950aa11b71e8ab89baf6308cc00524
53704 F20101204_AABENS dagdelenuysal_d_Page_82.jpg
e60b5f71e538b9463ddd1c4c32c6de01
d213f2cabb11d444cc66d8cf7f6f145fd7d941ab
6305 F20101204_AABEOH dagdelenuysal_d_Page_21thm.jpg
139aa92e76407378590de6014eb10609
0b33031723533923d5481fb8860fe45552326cd3
3104 F20101204_AABEOI dagdelenuysal_d_Page_95thm.jpg
e4dcf0d851086f918ea382dd30525c32
e3257cfb2fa896a08d6cc902e13eddb3bdcff5fd
F20101204_AABENT dagdelenuysal_d_Page_38.tif
e393418958fcf890ae7c8ca0ad3320af
28f9352e85d764e4d34c41c957543d82a08505df
F20101204_AABEOJ dagdelenuysal_d_Page_34.tif
5c7c26ba540ee9cf9d038a18adf8b376
02b5df378910ab1dd83fb5c407fe91468ec215cc
F20101204_AABENU dagdelenuysal_d_Page_84.tif
fab79bc986333e5c68a9beb50b7cf410
1ab3c830bf6fcc8fcd8b0d5c434cb77b84b04040
78872 F20101204_AABEOK dagdelenuysal_d_Page_36.jpg
fc431772cbb39a75d593785d48f55439
ffc33c4d742002e15753a63149d3c8d3b14ad5d0
1990 F20101204_AABENV dagdelenuysal_d_Page_61.txt
a456762c2cc0234e1fdcd57d9889905c
1096d3cc3bbd4da9c9953fbcde70d81a8cb1153a
821351 F20101204_AABEOL dagdelenuysal_d_Page_37.jp2
1b03cdee112e19ec208842b74770808b
9bd03772763664c6df4c526088c60928a0d688ec
85164 F20101204_AABENW dagdelenuysal_d_Page_73.jpg
754a007a5f42c19fd7b975b458bf32fd
589d6f822612d356b6b81cff49be385143b82f18
53 F20101204_AABEOM dagdelenuysal_d_Page_86.txt
a49c2bbe6d4ae5501f516d60019a77ad
cfa0aeb3a94a2357f1dc07d16b37775e59254553
49212 F20101204_AABENX dagdelenuysal_d_Page_16.pro
635a5a933bacb953c9f7c8addb1af9b9
362bab5f9e868dfab4023ac3b84959a09f01c3e6
6795 F20101204_AABEPA dagdelenuysal_d_Page_20thm.jpg
b05ff41051c3d7d1216a853a1f8774ac
d5fce0fcbeaf3be4b5ce31580573433f403b55f7
2109 F20101204_AABEON dagdelenuysal_d_Page_74.txt
8b627ffa43d2149c6687aa5fc69e58ef
c4e9ac7d0f70e6cec84a85ebb5fe0d322102d13b
1005800 F20101204_AABENY dagdelenuysal_d_Page_79.jp2
3e06080c5a5c22d97d54d2577c63b75a
1bdbad99cba7293cea295c3890e38df7d85d226a
6891 F20101204_AABEPB dagdelenuysal_d_Page_61thm.jpg
3d5bbf4907778dc068c5790c630d8c23
ac4066ee6bdc7e98759a57f6e3b7d5958598dae5
6836 F20101204_AABEOO dagdelenuysal_d_Page_32thm.jpg
f39487a3915292ad8632e8fc6a6fbffb
9e943799ae913eb41eace212803ecd6178c37e9f
1906 F20101204_AABENZ dagdelenuysal_d_Page_26.txt
c0fed9cf6466e6e1abbe522f35975893
b53910dbf30ae22bb2bde1cff6cf80298d0c7280
13301 F20101204_AABEPC dagdelenuysal_d_Page_87.QC.jpg
aa33affa9adc40ac74c0dcfaa8d62449
5fb2a6e7a0fb403158305af8d27fd76ef63f9341
45654 F20101204_AABEOP dagdelenuysal_d_Page_29.pro
cce9b4482e2817529b5ea64d07ef35ef
916716e5a7b161dd4b5215565c1082009cd83e61
6129 F20101204_AABEPD dagdelenuysal_d_Page_09thm.jpg
2ac1c3f942bb49610deb950f5f53faf8
256283609742744a30120716986ad5f0bfcdddfb
64013 F20101204_AABEOQ dagdelenuysal_d_Page_12.jpg
4d902c9abbc8398288e9d4b889f31f9a
11a7854c78926f071e46757dce3c35eabd7056e4
1051984 F20101204_AABEPE dagdelenuysal_d_Page_57.jp2
39456e6f8bb87c3f28a56e8951a717e2
25d4dae3ddf59f16d8fd39766e7c2e86e2a1a4f8
25036 F20101204_AABEOR dagdelenuysal_d_Page_27.QC.jpg
8ffb79208ea22a7bf73958e843a5b3d1
f1e823f3538bbe45569a68d9e64b232968c31256
18037 F20101204_AABEPF dagdelenuysal_d_Page_08.QC.jpg
1b46588539ec770fd95bce045335cb38
59be5b6661a517693f7a023f6cbdae4917f700c4
23953 F20101204_AABEOS dagdelenuysal_d_Page_71.QC.jpg
3261a0a26ee4c4a1ab83f2a92973ae08
a408c492429f7c12ceddcae4c9e3f90b77cccc6a
F20101204_AABEPG dagdelenuysal_d_Page_33.tif
3103908efc4bf1c3f11b99f730b5cf9b
f01b03ce07ea39c9d0490eae9791c53617d796b7
F20101204_AABEOT dagdelenuysal_d_Page_53.tif
4c59bdf4b9bce2e050db806ab4bd4268
af423ca796870e42cc197d1ad1a07c6ffd51be34
7416 F20101204_AABEPH dagdelenuysal_d_Page_24thm.jpg
e53587a85bd6fd6c881f660e15e3e94b
943b11746f02d2a97bce1e162f05e52f5814226c
10067 F20101204_AABEPI dagdelenuysal_d_Page_03.jpg
6d23e7264bac2bca85ae13ddfb84b5c4
cd906b521658395fd3a160fd731060d070a10a64
2257 F20101204_AABEOU dagdelenuysal_d_Page_08.txt
c1b24eb74e8b38b8e374b88806f5dcf3
11c48d68c196e06e7687de19d6dde3f3624fc9f1
1848 F20101204_AABEPJ dagdelenuysal_d_Page_29.txt
7f3a4ede5ecd327012fa6c306113d46d
7dd41e4d76f2c30bda815239c9bbf811482047bf
5156 F20101204_AABEOV dagdelenuysal_d_Page_50thm.jpg
33bf1646190c2d817e0d8ccb90358d60
16274b256bf1274d41b1c7c92c945321270523f4
F20101204_AABEPK dagdelenuysal_d_Page_83.tif
49dcbaf1d17a346d37c9bb0c4c72dc73
0d3bb94291eab84cf3e3daed4f9aa5964b460f28
F20101204_AABEOW dagdelenuysal_d_Page_87.txt
9ef71d5a1aacd42a0b45c1a0c42edcb0
8c3f4e27433360eea0e35d395fa5c01e2694dbcf
728362 F20101204_AABEPL dagdelenuysal_d_Page_48.jp2
5d005f0186b5b235054ac46cbb0ea69d
191377f6bda69d1db43a939b21d933957b200560
25316 F20101204_AABEOX dagdelenuysal_d_Page_68.QC.jpg
0d4dab8c0ea028e09b78368af2509aef
da26705e030314bdf598f3ed9e902bcc37fc925f
47905 F20101204_AABEQA dagdelenuysal_d_Page_32.pro
7ea24ef3b3c5cd3527657ac252750940
0476fa69bedce0b72f839b771ae9e5e789ffcd24
F20101204_AABEPM dagdelenuysal_d_Page_38.jp2
5b8b6096f779d65955964c35ec09a673
ab7b4fc562880e64b43e6043cc92a864713e64ad
50703 F20101204_AABEOY dagdelenuysal_d_Page_47.jpg
020d5348b3ef1af6d0019ae21f72e432
b4075a7cf631105dc087bffab0174ada7f575a7d
25743 F20101204_AABEQB dagdelenuysal_d_Page_82.pro
45d013e164b4cea0c081bc224083b3af
224b12413d2004e1fb85ced146fcbd4cc7759803
95263 F20101204_AABEPN dagdelenuysal_d_Page_05.pro
2bdeb46ecbc1df1f8cc27f383d405a21
f7a7fbc1a9c3a4a1e3602335bad1b96075689f00
4595 F20101204_AABEOZ dagdelenuysal_d_Page_86thm.jpg
9de970cd0ce23b6bf07645691ffc5abd
8db24b863ea18b2bafaf5f38e97f60377b1f0acf
3366 F20101204_AABEQC dagdelenuysal_d_Page_02.QC.jpg
9f04762fdcc9a8dbec93508495f3cc43
9f71fae8c78bc1d2d54d320752093c807408785f
19770 F20101204_AABEPO dagdelenuysal_d_Page_70.QC.jpg
79b9876a3b6f4c50d7fce7eb7e4056a3
6fd1320e09d5eca4a28a8e011007c1925ef6d1e1
51824 F20101204_AABEQD dagdelenuysal_d_Page_13.pro
4e14feaa881c56d5494d25741e4b10ec
8831b5490b08faed1ef84a529e7beacc03b2da53
91126 F20101204_AABEPP dagdelenuysal_d_Page_93.jpg
9d3e79bda952b37dcb3c16be0af15db1
145c3eb837ed8ff9053ecabfb9e1d5a0aa3db233
12688 F20101204_AABEQE dagdelenuysal_d_Page_88.QC.jpg
8f0d368eae01f38eccba2dc2be6e6084
cf0f4ef85f3e31b41204661b1ab1477b1017a004
497 F20101204_AABEPQ dagdelenuysal_d_Page_33.txt
f4e9e6d7ccbd376cd79d88af7edf0f77
ada0ac82a58b72acf065903862638e8822c6d336
2737 F20101204_AABEQF dagdelenuysal_d_Page_84.txt
0db71956d5619e9243bec1d067b44f24
6c32cc48e37db3eb543c778c20e0645434d3e332
24890 F20101204_AABEPR dagdelenuysal_d_Page_83.QC.jpg
9e5fa7bdece1f32752d37d80a1a03db6
e274c5637ff2ff93e1224267bd71e4ad23278a5d
F20101204_AABEQG dagdelenuysal_d_Page_36.tif
a71169ac6483984b41c5ab8c180118ec
38f1151a802d4c2255bd9f114fae4a40d89ddb55
5744 F20101204_AABEPS dagdelenuysal_d_Page_05thm.jpg
817ba43220745b905f91259528b2d02d
df343b0aa596936c4d23feaf42a1b4c1ccfa81f3
2148 F20101204_AABEQH dagdelenuysal_d_Page_73.txt
3ee9e30f22950abea155a19767f60bca
b7bb799e244509922a94f72fd758d3f5c4465c41
9029 F20101204_AABEPT dagdelenuysal_d_Page_01.pro
e454a07db35c5ba8c8a6557917a52d96
6c8cc9f59e252f8edfaecd2784e2e2f553db8a3a
49552 F20101204_AABEQI dagdelenuysal_d_Page_61.pro
f2a1a55f60ebcebe841adbd9026249c8
8f554b37efa48ac1ff9a1cc0c93a5fbcc2f13833
1610 F20101204_AABEPU dagdelenuysal_d_Page_69.txt
bf21719abdd7f299d2b94b4dc35ff6d4
901418b56a29218a83656b74b4b904c1cb597eb0
16110 F20101204_AABEQJ dagdelenuysal_d_Page_07.pro
45bd25dddce6ab38c568b21f13c2fb8a
1ee3e2758cad1ec509b98800434c7e470f1fe13a
82616 F20101204_AABEQK dagdelenuysal_d_Page_91.jpg
37d04dec0b0eb4d2691414975e0aba4d
001442744bd26265c60d569195c7cf7e96e740b5
F20101204_AABEPV dagdelenuysal_d_Page_63.tif
a2eee2e216881327a8cfd673a66c5d3e
c5da0e962359aed0ab73a5db88e3ebdaf2129f5a
F20101204_AABEQL dagdelenuysal_d_Page_81.tif
4cdb59b800e9ca01335b076249067102
203344886420a04b59c1a8e4d7642f55e68c8ba8
F20101204_AABEPW dagdelenuysal_d_Page_42.tif
abd4e6844ab9f191c963e753926adaa8
443f93f6efe5ff77504e16bc3a5e7955ba13f323
24455 F20101204_AABERA dagdelenuysal_d_Page_29.QC.jpg
a0c5f4c7a799859e046dd24f762b710d
75bb4e24277628721d419a038898f3c415ddadbb
2205 F20101204_AABEQM dagdelenuysal_d_Page_24.txt
892b2f71dead4baff6faa013ea12c21d
8f028a853b42cc7c03652ce3cc1a2aad33c80cda
16883 F20101204_AABEPX dagdelenuysal_d_Page_48.QC.jpg
5250584316dd4a4a4639300c4dc7b5bd
b1929fe027954d12600a4a7ac112bc1442939c52
1318 F20101204_AABERB dagdelenuysal_d_Page_55.txt
9081c9c3edf027046e28ae399a73823a
efb83873692eaee59a115a553e2c9f5f94e2a15d
F20101204_AABEQN dagdelenuysal_d_Page_01.tif
5e4f6ec2bb39c9c5ea125e4f4dd21680
4f4b139cf3d63c69ae2c6569f3c200f244740844
15811 F20101204_AABEPY dagdelenuysal_d_Page_42.QC.jpg
9029c096ea4aef8c34025c3cb7283ba3
6767c59ffda65ecca42667b627da44d12ab933ca
F20101204_AABERC dagdelenuysal_d_Page_60.tif
ea42460e9cbfe9ac862f007376d0f375
4ed19f0767c8aa6eee5b286a420c66c9ddd119b4
2104 F20101204_AABEQO dagdelenuysal_d_Page_92.txt
bcc5b5139697062a924068880e17e18c
3253fe67f211e2482b8a7ff8112cd65ce4a6c2eb
1949 F20101204_AABEPZ dagdelenuysal_d_Page_21.txt
5cc4b314434e960ce3dcd477f97db968
e73b4b6836eda80d5b7194bd64e79fa4b5d88616
F20101204_AABERD dagdelenuysal_d_Page_94thm.jpg
5b24d72a3bf1ea88842129bc89ff3c14
f2b2f3e0bd81805bd20c4eb384f08e87c1533271
1977 F20101204_AABEQP dagdelenuysal_d_Page_41.txt
420451da50cf97675564c096c546019a
5735561906405b62c1be783d5b298c0acd9c12d4
19622 F20101204_AABERE dagdelenuysal_d_Page_69.QC.jpg
6e7fbcfbacbc049d3f422f8b5907f023
c86bfc5a32be14a2d98117f08f200a0fac1ee3b4
45409 F20101204_AABEQQ dagdelenuysal_d_Page_89.jpg
424b18a96a353bd42dd2d264125073ff
a719f878b762837c1f2fe58ca7edeb86d0c8808d
F20101204_AABERF dagdelenuysal_d_Page_28.tif
4a17aa862fc4539e5c65052533545d3e
7d5d636fed0614b1a952e805fa70aac2656309ce
88188 F20101204_AABEQR dagdelenuysal_d_Page_94.jpg
bb288dde043d5243fd41579174fa7907
c75712d926d4d1e028fbc1d86aa3ba978eca97f0
F20101204_AABEQS dagdelenuysal_d_Page_60.jp2
292d309995e62e526d692dd8e271d9ea
cfecea8b250e6f3df5ed8a2e4ed138444a59f78b
81855 F20101204_AABERG dagdelenuysal_d_Page_05.jpg
4f085f6853e3e63f5b878e64f52c7942
1397de744a6aeec230786d341d3e69c2b08e2d07
20818 F20101204_AABEQT dagdelenuysal_d_Page_12.QC.jpg
a762976eb9f769a30b367432bcd0db29
a92e3b90536b0b38afaddd218775eec961aaefc1
594295 F20101204_AABERH dagdelenuysal_d_Page_90.jp2
8ad7a6e10e6da05e35003dd25b4fc0ff
85115ac09b7d00615bc47632e85af2b37779cd80
6052 F20101204_AABEQU dagdelenuysal_d_Page_84thm.jpg
fbeb2d4a9753c02f2ad56fe23a8d54e6
36642f2b06a5082e7d1839e348c553aa3b087155
4914 F20101204_AABERI dagdelenuysal_d_Page_08thm.jpg
e425eed48a184e523e67111cb6765c5b
396a584db5563e888768f7443619dfcd2045aa61
1051985 F20101204_AABEQV dagdelenuysal_d_Page_22.jp2
c35b7feab6d013f1c2a8d5fc92b91f9e
e87ebeee829329aa2347151579659340194017e8
23832 F20101204_AABERJ dagdelenuysal_d_Page_14.QC.jpg
a5119a6d1b82a80292b2c311303f4531
06d11ddda0a81078eba6d06bdb24c386f627673a
1051861 F20101204_AABERK dagdelenuysal_d_Page_88.jp2
0aad9b46a1dd00e3997b106624a7c780
a5b24edcb6966725e30564355074d8d506a02cfc
5375 F20101204_AABEQW dagdelenuysal_d_Page_59thm.jpg
c1736f25e108a68ff50088736fafe5d6
096127391ee0910e269e4a6966722aa9b91572a2
1051967 F20101204_AABERL dagdelenuysal_d_Page_32.jp2
4bf664e24b0d9581f47f92eaf6c3fa36
4fbc48ba2a0882936bb9d21ed35dd0a2f071ed58
36750 F20101204_AABEQX dagdelenuysal_d_Page_46.pro
65f9c684ceca08ccd98c112a7699eb19
638fd7d8e399cebaa6831de12e14ace60786239d
10261 F20101204_AABESA dagdelenuysal_d_Page_02.jpg
3aa506883d87bd52679cc3f798fff48d
8afc2b10d21d038538abbe13a05df4fec9a3e73a
2548 F20101204_AABERM dagdelenuysal_d_Page_47.txt
da60571d9412a01b70e4727791552e41
eb9d4158a3132ab6f4ebbe4a71efffdab38e0086
1357 F20101204_AABEQY dagdelenuysal_d_Page_02thm.jpg
b8d93bcd98d806b0242684bf9e6ee90c
8d340bffa9b2670ce1987ca808e8360e7c97082e
57528 F20101204_AABESB dagdelenuysal_d_Page_04.jpg
e9f1f6275bf681df5b82ce6c0f57c309
63685db8b666e474e27b365a99774b36b03f89c0
1051983 F20101204_AABERN dagdelenuysal_d_Page_20.jp2
a8bdcdc6dec90cc917ab99506940cf8e
d2df2e61fcb3516951473efd6f9036918994e88f
6417 F20101204_AABEQZ dagdelenuysal_d_Page_26thm.jpg
542f4b10fae957725d50f8f13f451445
bc55c9282a162e0178894fbdc006fb19009e1813
97064 F20101204_AABESC dagdelenuysal_d_Page_06.jpg
766786c4de1aba536bb5eb0d10d2d812
e00ee4445c985823afe225d3007b26701f8741f6
75576 F20101204_AABERO dagdelenuysal_d_Page_71.jpg
83905d4dede6d199f00f476ae3616f0f
57dafb7afc0ab86a61a69b47219f67148b5ca50e
21268 F20101204_AABESD dagdelenuysal_d_Page_07.jpg
a48d30df4a92c18875bb7b2419678cb0
566a75c0ed76f37b50a3a8fcb0e9cbf6368b549f
1965 F20101204_AABERP dagdelenuysal_d_Page_36.txt
16ab7314ceec493ac00b72e4ac64fd24
a6e15adefeb7a84a2606fea0d6e13f7376462fbb
61945 F20101204_AABESE dagdelenuysal_d_Page_08.jpg
fddcdf3a1410a82a6b7898b9d5cfbe6c
4cdc01e7d50ecf986ed5bad7a9b34365280c9aec
F20101204_AABERQ dagdelenuysal_d_Page_32.tif
9c6b3687fa28b9972801ec4f4e6661d5
f14957dfd40165131908887c01f9727fa2b55f6c
82660 F20101204_AABESF dagdelenuysal_d_Page_09.jpg
11e2f0adca4a82bd70ff06365844cf33
002af6f3df5362740cc7e61aa7c72ec8d43450b5
1051972 F20101204_AABERR dagdelenuysal_d_Page_11.jp2
ee34d0370038117036320db4059c48c9
b564a387820994cf4e6de3d4ddecdc6a158c5e9b
41583 F20101204_AABESG dagdelenuysal_d_Page_10.jpg
61daaa38fcba91d8513f2665c2882ea4
3a614840deeb00b4e2be8f583e7cc6e532549369
5956 F20101204_AABERS dagdelenuysal_d_Page_72thm.jpg
202bae2b03b422d0d4c535c377270591
d71891a4b9665b9de3379db01827443760c07493
81113 F20101204_AABESH dagdelenuysal_d_Page_11.jpg
56ab0c7172fe949f82983931e3b090a2
bb4945861114a242c1fce665c8eaa6cff7394ee0
75264 F20101204_AABERT dagdelenuysal_d_Page_15.jpg
14f7a37b8057a728ec6f68560e36ad10
59d25b90067210169a36e64797e39a71d9bd2a64
84172 F20101204_AABESI dagdelenuysal_d_Page_13.jpg
5224c463c21f70b2e3d41dca65d39af4
aca02bd36c1954666a76d26cc4243988d4374ae9
F20101204_AABERU dagdelenuysal_d_Page_95.txt
13de366dd9df9fc7108881925f6f6255
804f709df5f8ea2a1cddd8e37d54ce00d62e3893
74348 F20101204_AABESJ dagdelenuysal_d_Page_14.jpg
10573642eed71197d051f4cda4b832aa
d8d7c3ffaeb62e41fd63e63696642f638d037fbf
53076 F20101204_AABERV dagdelenuysal_d_Page_22.pro
407997d22e1ee56f4348bea3d02913b1
a6a47fb88c0c163719f6122caf8f516b19015f84
78367 F20101204_AABESK dagdelenuysal_d_Page_16.jpg
770cef838c6cf91bf169ccd60ec6c5a8
0c0ef0bc2017cb64738fc0e5db0f6289e398fed0
147642 F20101204_AABERW UFE0021827_00001.xml
0ddeef56ad3783eb20fd0e8a81553b50
fc0cb818284406cd078470c0ec7f9fb2d84d3f7f
77795 F20101204_AABESL dagdelenuysal_d_Page_17.jpg
be55f1b70d05bf00bea1a3f26e1c9eef
b352e3c3e9f94d8b0808d6920fa7f2e6e24eff95
53403 F20101204_AABETA dagdelenuysal_d_Page_45.jpg
4aa3ca7a70e7b64a3c769e59ae1e34b2
e5a41ec4c2a04f9c5292a1d5c2ca9068e6686286
85211 F20101204_AABESM dagdelenuysal_d_Page_19.jpg
f08792cfe98fab9c92bef3e9cd7849a2
94f370559d4da2f0a06655228a5a878d833d2056
61179 F20101204_AABETB dagdelenuysal_d_Page_46.jpg
60bd3caf76096b83cfe40723cdc02852
6849376d600b7fdbe14f5593e3820f87317a54be
86430 F20101204_AABESN dagdelenuysal_d_Page_24.jpg
69fff6a960efaa3f14051ced4419c6ac
49cd413bf93c1d69b2cd520251c8ae7eecdc9605
48348 F20101204_AABETC dagdelenuysal_d_Page_50.jpg
0ff8d40606b7f3939002fafdcba043e3
95fce64f295fa7c2db470da8e2ee58318a2fb4d7
84830 F20101204_AABESO dagdelenuysal_d_Page_25.jpg
a21d3a20ca0c81324168d988dea89385
fdb6fe797e009a13b4fee09cd53df9a68b083b37
26889 F20101204_AABERZ dagdelenuysal_d_Page_01.jpg
f3af030e26d05a6cfc99f54d9a629dca
e13f7d86ce8f7bf8134ca237610537ccc892d1af
67448 F20101204_AABETD dagdelenuysal_d_Page_51.jpg
5471be21f70c3f47084504f6dcdeb2c7
5a3c305808166742797447b5316c5c4bcfb5bd66
78684 F20101204_AABESP dagdelenuysal_d_Page_27.jpg
ae4f5d34f0aea8fbb2ad1703c4a48a9d
9d9642c5e90370169330fa094fb676caa95b0b6c
57958 F20101204_AABETE dagdelenuysal_d_Page_52.jpg
973a2dea3ca2ccdcb2e59171940b4ede
1e11a8c403fcb8b7dc031800c8b9aec8e942277f
63329 F20101204_AABESQ dagdelenuysal_d_Page_28.jpg
4245019aa57a6ef3b0089be67ce8311e
836347cab2be45043b9739c8debfa355492cd619
69058 F20101204_AABETF dagdelenuysal_d_Page_53.jpg
7c4bdde45f0bb89606334df689f734ae
741c1dc56a21848a8d2fe41774581bba3711b9e4
82731 F20101204_AABESR dagdelenuysal_d_Page_30.jpg
c2df4facfd56a413a7bfe008f4a5ac24
5c2be904342c5a376c11b27f7d9753db3aec5035
76794 F20101204_AABETG dagdelenuysal_d_Page_55.jpg
0589d5a2c1ba338a07052cf876cbf620
f77a38c034b785260bb73382ccde5539b1b24362
52367 F20101204_AABESS dagdelenuysal_d_Page_31.jpg
ca8bd53139173a9ed9a7594d02826460
0c84e27075e3b4dc07d7dab00472ddfa5a042ce4
58804 F20101204_AABETH dagdelenuysal_d_Page_56.jpg
5ca92408a6cccd4aabc55c090ed4144c
8581e52148081e315aee04ef312fbe0d981a3417
28837 F20101204_AABEST dagdelenuysal_d_Page_33.jpg
544ca35be01e16c766e910ce86b47cbd
753a3b0dfcc84157c19303f7905c2304c2f0de32
58295 F20101204_AABETI dagdelenuysal_d_Page_59.jpg
6ec965f4e04fcda0d043ff15403d9822
8d0f6ccf8f9d6b6413b16d7048429ff10bc61c7c
48963 F20101204_AABESU dagdelenuysal_d_Page_34.jpg
c3484d48efe3152610572bbeb3f8a2cc
4d7dd738c56a1b0d5c0304dd4f88323cfdabc8d5
81242 F20101204_AABETJ dagdelenuysal_d_Page_60.jpg
ac9bdbe3053cdb6442f0c41806557ff1
0c08fd3e5a24d4ec198883844d561ceeae09246a
57122 F20101204_AABESV dagdelenuysal_d_Page_37.jpg
d76e9f18b87ea0f55975bc31484f4bf8
1992c98c844bc0fa6e555c8b7f0175afc87da4d3
79475 F20101204_AABETK dagdelenuysal_d_Page_61.jpg
7fd0ee019b02d40c7534014ac6c00f11
33566e63a381d8944fea44bbb596249fcea52bab
52072 F20101204_AABESW dagdelenuysal_d_Page_40.jpg
8acd6221a2ec4f9b0335f766f7387138
2111cfe296c8a4dfabb031001f8bbab811840f30
56686 F20101204_AABETL dagdelenuysal_d_Page_62.jpg
4c1925a0cf028387b5cb228deb12b5be
c952376efd6ba127f51b222506f4d8a8a20b466e
76292 F20101204_AABESX dagdelenuysal_d_Page_41.jpg
435b22bad6f0dd5e67bf03336d2a35d2
51c19e9720c089cf12d9582e64cbfe36777a2124
75269 F20101204_AABETM dagdelenuysal_d_Page_63.jpg
9ff9e4997cd8ce20a92ae97d8a79a3fa
a19e51226a709ff911fb36cf572c2004121269dc
30942 F20101204_AABEUA dagdelenuysal_d_Page_90.jpg
0ce8bd4bf2ab78d31ad77aac9be338ef
158bf5ccf158fee896d560d24f70e04cbb0e5723
73193 F20101204_AABETN dagdelenuysal_d_Page_65.jpg
b3b84e7dfc0added87c82bd095ee69a4
6b73cf67030cac1c1235bbd1bb526a2e1e98f441
52285 F20101204_AABESY dagdelenuysal_d_Page_42.jpg
42c4e0977848c4e1453d8cc52de8cd00
2fa312fdf77a30019be57e74dbd5742fead396de
80856 F20101204_AABEUB dagdelenuysal_d_Page_92.jpg
560948da40f801044cb3f7d84d29304a
c8aa4d5b9f83542636d3b15ce9971f7373c962c8
60210 F20101204_AABETO dagdelenuysal_d_Page_66.jpg
d3a010c5dbb08a81652629395c0e1b6c
2512854fdb1a9ecd710a5d482fc016e086e4ab45
62090 F20101204_AABESZ dagdelenuysal_d_Page_44.jpg
5220f85ce76c0720e60976caeb95870f
daf34b576126b9bb7b5e130ea9654c2a9fab43bb
31523 F20101204_AABEUC dagdelenuysal_d_Page_95.jpg
e142455f5cbbc39c9aaef7531cd222e3
43e4c7606fd3f6e8b324a606ce4c3a4ade41285e
67983 F20101204_AABETP dagdelenuysal_d_Page_67.jpg
47d48e28e5b55d568de6b6e9a0645982
3003c882e1fa04a1780ed57fb326c3b9cf923d73
29439 F20101204_AABEUD dagdelenuysal_d_Page_02.jp2
38b39d590cd5a82e92beb9c167edefaa
585cafd08da3e1c9fab5607c2782e2bf73da13d3
63783 F20101204_AABETQ dagdelenuysal_d_Page_70.jpg
7a418d1c7c923d6b5476a0a1e9ec08b9
6db2223a16adbf6305ed2a14b45aeb886fcb1a5b
25057 F20101204_AABEUE dagdelenuysal_d_Page_03.jp2
49b0dc07bc8c8cde44a9fb13ddf8e6e5
c3f84be68fb96be28c3a8d1ce5d0e7ad4a3ac16a
82827 F20101204_AABETR dagdelenuysal_d_Page_74.jpg
57ee542f42ec52b61e9c76a818b152ee
864ac4f9e160a0b0b1cacc5ba48d4374fe003deb
25723 F20101204_AABFAA dagdelenuysal_d_Page_49.pro
ea6b256ac188683b29e893cdbd2bd529
4d1c378b3d084a28d195e36a343ed6fc47cf495e
1051951 F20101204_AABEUF dagdelenuysal_d_Page_05.jp2
29ecfde1a42d512d5687d914d7bc43a5
f641fc08b7d199efaf00d56a6139b7f0bfac6bb5
80803 F20101204_AABETS dagdelenuysal_d_Page_76.jpg
c64f61cc610559d88abba8552b7f1999
e767d2a62672d1875810439b15c1c9cdeda87a16
16271 F20101204_AABFAB dagdelenuysal_d_Page_50.pro
20f133f125d0fe955e0612af6dd5979f
31927cdb940da300a7733a7750432e94bb73cd2d
1051977 F20101204_AABEUG dagdelenuysal_d_Page_06.jp2
3bfe92913dc3e4efb201786cf2a62f10
3e686ee5ee81d69fcbc0d199dcead47f2baca399
30339 F20101204_AABETT dagdelenuysal_d_Page_77.jpg
614beb35632c8f8a59bc852e543b94db
e0aeefca3af020a7db2ca87a2df0458e25c09c49
26305 F20101204_AABFAC dagdelenuysal_d_Page_51.pro
6d1b8b9a7f961fcef2e10cd2b4e493b7
a0fceb381496b44aaaa7fd8eb23e7b6cd49f56b1
417471 F20101204_AABEUH dagdelenuysal_d_Page_07.jp2
e5cb7b17d0e73c75fbc3f8b89e54333a
e897b974fa33d0b35fbd89d724628df41f58afdd
67524 F20101204_AABETU dagdelenuysal_d_Page_78.jpg
526507e8e3d392aff7c04306dac21f93
fb6b100c301f74d8043705432391c0dfd4f11ab1
29897 F20101204_AABFAD dagdelenuysal_d_Page_53.pro
30ef68f046dfcf5fa5125f6fa95984bf
0c08fc6c1b06165270f86330712c2d5237ceabbe
F20101204_AABEUI dagdelenuysal_d_Page_08.jp2
c075dacf0b918714069ebf7a06995e91
61fa04734297e2b2aa095222f5fae83487ce53c0
61186 F20101204_AABETV dagdelenuysal_d_Page_81.jpg
95c45db9ba95214c5a013caab1598ba4
e2f3b47f549465224bb77a268c9ffcc6c8f5a651
27427 F20101204_AABFAE dagdelenuysal_d_Page_55.pro
74870932b13757d72467f427dd8846fc
a5d57c014b453f9ccf34d399bd45c8ef83049d21
1024641 F20101204_AABEUJ dagdelenuysal_d_Page_10.jp2
a8d8c272cc49361d5ce64d6413064fc3
81a1d15cd7eda440416aebb3c53abfa9ff23827d
84325 F20101204_AABETW dagdelenuysal_d_Page_83.jpg
40474abe078bcb7ad87b8566ac7e1122
51d1d6d1b5988e50bb3c67399b9e957d01cccf0f
F20101204_AABEUK dagdelenuysal_d_Page_13.jp2
22c465eebffed34a2b5dd455afb0b4a5
e5a904622fe099e9a78d8b8b688e730e271351cf
42678 F20101204_AABETX dagdelenuysal_d_Page_85.jpg
a91b96aad5d995e85bd9752f3392f7e6
22c3a56922acd87fb674c15667b99fe6842411ac
14470 F20101204_AABFAF dagdelenuysal_d_Page_56.pro
cc7aaad7a9e5d01355922c7bdcd5bf04
1a8c55f762536ec5df9e5e6c72930e3ab6a1a824
1043462 F20101204_AABEUL dagdelenuysal_d_Page_14.jp2
1794bb6fddf73d4dff663e9c49e3c357
ca2db9a06aff3d3a78571bf7176563475e87cb5c
41029 F20101204_AABETY dagdelenuysal_d_Page_87.jpg
f4b32bdd36827b9c463a982bdc5c95e3
197c868fe079f7913ab4f2bedfb528fe797aca96
51848 F20101204_AABFAG dagdelenuysal_d_Page_57.pro
3bdca1a67a86848508e7eb5a442fa1b7
3b18554b378a3f9da1f14ca9c0dceca62b69c120
637192 F20101204_AABEVA dagdelenuysal_d_Page_39.jp2
3d810cee8fdb6cee5f79df94ea2e03af
ec4dad0d7055e187b3b06a7983a07ca2406e4bc0
1051931 F20101204_AABEUM dagdelenuysal_d_Page_15.jp2
41a9e99f6fd2efff89e63d46551ca579
a3730f40d926d6c4d50fad480f51df302e60373e
52862 F20101204_AABFAH dagdelenuysal_d_Page_58.pro
8217cc22c9bb6fa16125d62e50b1ccc2
128ac814aff856de57af71de534079af234c5df4
692046 F20101204_AABEVB dagdelenuysal_d_Page_40.jp2
314c3b354a469c42a859cdac343cf2e6
052bec7270f9ce798df83d0cbfef829a9a94dfb1
1051964 F20101204_AABEUN dagdelenuysal_d_Page_17.jp2
fe8e9d36dde1188845f5dffcdd566a32
9186175295b87c0266fe903249558236f57a3262
40387 F20101204_AABETZ dagdelenuysal_d_Page_88.jpg
5ff16981d160e06c35af9e66faec9a1b
dbb00b2f571ae713439a483177a1ff69406fca2f
35634 F20101204_AABFAI dagdelenuysal_d_Page_59.pro
b409f4b593fd02899f7536c6e81e7420
09af387261a905ec2c881983a54f9c9238fb6759
1051010 F20101204_AABEVC dagdelenuysal_d_Page_41.jp2
26c4c67306fee3302b9ed9cd0b086415
53ee617f1a60c5ee76434973bc1f2456df988969
1051944 F20101204_AABEUO dagdelenuysal_d_Page_18.jp2
63d09385f8fa0111289e88cfcfc4e4d0
5a0abc6f68f5d90f373857c7c981b9dff38c31e7
50460 F20101204_AABFAJ dagdelenuysal_d_Page_60.pro
6dc90c3db616c05471c1c30636e58f43
5b6ae234f106769d51a9ee1daf08bf844efef0aa
450996 F20101204_AABEVD dagdelenuysal_d_Page_43.jp2
38b316081049dd5be537b6874812fa46
9217e5f893eaf6103e86906325ad617db2912f61
1051968 F20101204_AABEUP dagdelenuysal_d_Page_19.jp2
9e000b10d341e20ca70baac32704d962
38c94f66f262467d08857829f05ec5e5e7c8767a
26536 F20101204_AABFAK dagdelenuysal_d_Page_62.pro
f7cdf2ff51234b3d62b4138a15f0c65b
5a9fd1fa41cb2fbe6ba2a884ae10d652840a6421
807951 F20101204_AABEVE dagdelenuysal_d_Page_46.jp2
14cebf05ca1a9bc32e867ec0b3e52d33
a07d7fbf46f4b2a030e591d60b747c3f3182002a
999407 F20101204_AABEUQ dagdelenuysal_d_Page_21.jp2
fbbdd540a90ab2ee46632f60419c2937
54b4a5e12e173ef4128c80914e5ef0069c05c584
50785 F20101204_AABFAL dagdelenuysal_d_Page_63.pro
fc254c586ef2b087fed4954f659052a5
7553ed8761a63dc20692b4688d2081b0f0404d15
794653 F20101204_AABEVF dagdelenuysal_d_Page_47.jp2
97f48c4308d73bd2ad3422e2b8ffd685
8bace380bb36bf6514d74c8031d8a599e7c3914e
F20101204_AABEUR dagdelenuysal_d_Page_24.jp2
39cb55f534709a3024367567a9d7d22d
7088b7fed61eb035c226e48b77eeba6712870662
30550 F20101204_AABFBA dagdelenuysal_d_Page_81.pro
08884d70467da97f456923dd5139c522
09d0ff723c85b5afd1731407d82cfac6da8b9b0c
32924 F20101204_AABFAM dagdelenuysal_d_Page_64.pro
e544b2c14b86593f1bc03287343d9269
1f18c527a577165af4d86e2bc1fc3e1fcaabbabf
1015735 F20101204_AABEVG dagdelenuysal_d_Page_49.jp2
c4a95ce07acebe0fab950ce600d411d4
17fc7ca3c06ee0a9a029b796dd712b1acf2f6984
F20101204_AABEUS dagdelenuysal_d_Page_25.jp2
ca3d908a511cff4ad7d99ba0d690656d
fd9f466f0cd63795a0d9e6cc21a565d42f22a207
53333 F20101204_AABFBB dagdelenuysal_d_Page_83.pro
cda83ed5990d9c2a5e498b4b3c3669d6
93ffdfe86ecfa5b2b3e524227b56c6da4da5739e
44945 F20101204_AABFAN dagdelenuysal_d_Page_65.pro
8a32f9808e978eb8c5bea3f7ed74987c
93327cc72f3104f253dee4192349ad44d5f144cf
F20101204_AABEVH dagdelenuysal_d_Page_51.jp2
7080e6865f6619e2a40decc84cedd688
5cae71acd64525a0dc03e98640a6f52206538b87
1016158 F20101204_AABEUT dagdelenuysal_d_Page_26.jp2
dec775a3dca5bb93ff0495049b2c5dc4
7f581ff49d6c88c7751a435d14210612bc6eeafa
47662 F20101204_AABFBC dagdelenuysal_d_Page_84.pro
ea1b499118936a51b5129a2f61a19707
1816219610845a41e10691180f40d2b0939b8176
40382 F20101204_AABFAO dagdelenuysal_d_Page_66.pro
2a1b659bb585a771705d8c72f36791a3
46731b7b60f5e348bedd8c6b7c386f117d91d6ae
895175 F20101204_AABEVI dagdelenuysal_d_Page_52.jp2
c573c2a249b1b60de0b8fd9d42f73680
c5d872e068c060b4cba9e880ea8d5ffabd5e34ac
1051942 F20101204_AABEUU dagdelenuysal_d_Page_27.jp2
4d5df0293b6d9298e2e45297ec3e1262
9b1ceafdc4b7c0dcbb4d7fb0b643fb8fc3fe483e
2087 F20101204_AABFBD dagdelenuysal_d_Page_85.pro
a8bde9895b2b8d91ccbbf7681e9ce6c0
dccd259b82a2e88023367b8d5e23a8c7ea8819a6
41048 F20101204_AABFAP dagdelenuysal_d_Page_67.pro
77deb5b8d6f77bdc94cbaa9ab9760dff
da719ec22f2c4e76e0a4e9e9216d1b90cf421b2c
1051975 F20101204_AABEVJ dagdelenuysal_d_Page_53.jp2
4855801d31136e46e41df7d3783c932d
e268c1cd3c5c09f7bf0fef4bcb18de429107c623
1031807 F20101204_AABEUV dagdelenuysal_d_Page_29.jp2
7c4124295287cd55735f0320f6d31d1f
8e7935e9abbdb476a3f1a018564d1065534a8b56
1125 F20101204_AABFBE dagdelenuysal_d_Page_86.pro
58f664d79b5446c1177c20436fdb0f18
aac7997fafd6a104d655a65b3ad492f1bb9a96f2
49752 F20101204_AABFAQ dagdelenuysal_d_Page_68.pro
e8b7fa4c7803b77f92246bf3b16a73f8
3b6e4a12e25c80cbd2f3cceb3b5ae073ee3397a3
786002 F20101204_AABEVK dagdelenuysal_d_Page_54.jp2
49b7628d43b8d95a09b90186acad9bc7
b72e32c3e67ea132e54d8795f376f647ef44a1dd
1051953 F20101204_AABEUW dagdelenuysal_d_Page_30.jp2
8773ce7da6fbc3b96bee5305b9dff7c0
f230b1b898b98ad5ce3723d5c7f7ad06f21a7aca
F20101204_AABFBF dagdelenuysal_d_Page_87.pro
0471a1c7ddcd418353820753df62abb8
a2a14cec14c4961c0204d468bc6ac1ef936a7111
34275 F20101204_AABFAR dagdelenuysal_d_Page_69.pro
8cf92eacb6f2b82dd397047a3fa8170d
ba97e9cf9e7b363afb827e533f2569e45b50b7f2
F20101204_AABEVL dagdelenuysal_d_Page_55.jp2
3ec47daba8ecff5347b01a32767c3602
31da1c5977a070ae1744e5a3ab416bc62d8aadc6
804957 F20101204_AABEUX dagdelenuysal_d_Page_31.jp2
dca255f18dcc58c7345db1ab11545bd6
533d421b3f8b606933a9ba720a30937052bcc9d0
1051925 F20101204_AABEWA dagdelenuysal_d_Page_74.jp2
5b98d04d13e4231c091b3814a3622b21
4e513c2c10e2704fc75418e5fbdce12cf333693f
33283 F20101204_AABFAS dagdelenuysal_d_Page_70.pro
63a074c80111c1a129f58fec55c8e62d
9438c565340f5ac4f8fab1613940c0d1f3b7caad
1023992 F20101204_AABEVM dagdelenuysal_d_Page_56.jp2
a7e300fa8caaa98a3cf6fab42b28ee20
b4ef6e29b7ab183aab4ca5aa8cc8b541a97404a7
579969 F20101204_AABEUY dagdelenuysal_d_Page_33.jp2
f42a802f166d6e451080888904e932c3
5f5399f0bfb4f46f7238f30a757748762d4301c6
2221 F20101204_AABFBG dagdelenuysal_d_Page_88.pro
00200b9b2626dc01a1aa7a8bd72ec4da
74a5f897ae709fc68b68345c19cd0d5401f43a2c
596570 F20101204_AABEWB dagdelenuysal_d_Page_75.jp2
955870164f81e46a4e928bc6697f670c
2612cd0dfd5a5ee22f024f4aee47ffa6e07e01dd
44170 F20101204_AABFAT dagdelenuysal_d_Page_71.pro
049aca49af14af13bc28824d6de949ef
808f9115408fa1525c3c7f4de21233296230b1c6
F20101204_AABEVN dagdelenuysal_d_Page_58.jp2
aee6e70a74ac5a07de3e67a83634419e
f9f7ce1c86fe120006c865795c79b4ccf758cfa5
897050 F20101204_AABEUZ dagdelenuysal_d_Page_34.jp2
eb4e828ec5ac69e463b7d0e4f263a659
8c14a910456c601c08bbd27cd9ac002337996813
12618 F20101204_AABFBH dagdelenuysal_d_Page_89.pro
1449970bcf9c3c150892018105b2647e
be79ba5efce4496c73378f2434e7e1ef4ef95a80
1051979 F20101204_AABEWC dagdelenuysal_d_Page_76.jp2
974e8d682569bedff34d8294fad3130c
acb76acdd677f093a68aa91a735362cf8712a7af
38145 F20101204_AABFAU dagdelenuysal_d_Page_72.pro
cbb9b010a1dcc0404cd2a75ca3bd4170
2b717027ade5c60abd3f7fb106618f832d6867c0
1051960 F20101204_AABEVO dagdelenuysal_d_Page_61.jp2
2e8277dd5afb02d869dfa500d9ab2b14
38a19b0b12d7b21e3fef5938a40ef72713a73853
7696 F20101204_AABFBI dagdelenuysal_d_Page_90.pro
3df762f95742612f0c80acb8559eab47
7c66077090af2db8275228e6588f8be3f3b72cb6
54555 F20101204_AABFAV dagdelenuysal_d_Page_73.pro
66f4a2632038ce64befe4174a68e3f4c
c70a5edb66463116347dc3c8fb85ad75f22014a6
885937 F20101204_AABEVP dagdelenuysal_d_Page_62.jp2
c3fbfb60a1efbd64c742901145b0622f
a83c6681657150b705679c333f271f433a1a2923
50243 F20101204_AABFBJ dagdelenuysal_d_Page_91.pro
271e2320abacfdb351b48577a24a8c26
81d9086fe70f0b6845f30d75acf9393f2081899b
358332 F20101204_AABEWD dagdelenuysal_d_Page_77.jp2
5aae285f3a501d68b1c9fdb92fb47ecd
6c7ace207dbb2f8256d3faffd5f2cc41eccb1254
53232 F20101204_AABFAW dagdelenuysal_d_Page_74.pro
416227584aa4dd97d8f4b1dbbc40d6a4
9dcc382a044001112d692a1de16c2d6c50ba989f
992380 F20101204_AABEVQ dagdelenuysal_d_Page_63.jp2
f861cc50a663c7b7ff40ded94754c5ad
9bcadeacd159d9bc3c8e81a2689368f1f841f1d8
50285 F20101204_AABFBK dagdelenuysal_d_Page_92.pro
f78b27642ba83ba18b8ff70202f06645
b07b32eb42b7a46ce0a73cb1b8c6999be5cc4f95
1031797 F20101204_AABEWE dagdelenuysal_d_Page_83.jp2
aaf4eb8c887f34c5bdb3bbbcde54a560
9b0654dda546578d9955e74c76db268d711970ce
50690 F20101204_AABFAX dagdelenuysal_d_Page_76.pro
58eacc6ac8571969ac8fcd2f5f9b3677
82db0a1b471a10323e9be14fb2dc16f3bbd7ed48
980601 F20101204_AABEVR dagdelenuysal_d_Page_64.jp2
edc14282af89e10fa86e373eb8082db5
d7e15617bb7e78222eeafa2d9a445037b1d4d813
2157 F20101204_AABFCA dagdelenuysal_d_Page_19.txt
88d188f88c4e2d0ddab282bfb5d36178
23061c4681305215dc8e58d66882aec8dfa27182
55349 F20101204_AABFBL dagdelenuysal_d_Page_93.pro
7eb639db126999fa7a9d353b5a74571f
99e1f057e3e502798bb6c7c4925000fcb4af76fd
834120 F20101204_AABEWF dagdelenuysal_d_Page_84.jp2
a54388dafdf60ba1349d0fe0079ab951
c1a1fc9503b12f577a4803974e6c724aa38e39d8
15931 F20101204_AABFAY dagdelenuysal_d_Page_77.pro
1248d8ce47fd40bd85eb3322a373b99d
adcfcd6a861b9d50436d15c5ce7cb2ddd35d99ef
88619 F20101204_AABEVS dagdelenuysal_d_Page_66.jp2
a8506f3adac753594f23f1b33728f8e8
071876ca4891769f6204d7ed5fc922ccae7a7c7a
2005 F20101204_AABFCB dagdelenuysal_d_Page_20.txt
53fecb3262aa4c1335cc3aba6ff5fe37
8c69249da31c39fb92ebff0f5e867f7552f14636
53521 F20101204_AABFBM dagdelenuysal_d_Page_94.pro
95727858cf842b489df02877a137a87c
b78fa886a468ca448b877b8091eae18cd24e7ca3
1051955 F20101204_AABEWG dagdelenuysal_d_Page_86.jp2
cd17d199ea26037306dc37cf1e12ff0e
2f2ff3c29ad0b983cab0296d46c9b00ea70b9f45
29125 F20101204_AABFAZ dagdelenuysal_d_Page_80.pro
4f803d4bb8d76d69a9339017b25968c5
7b1185e05f6935d1933caa4f41cb89a0505e20f3
802335 F20101204_AABEVT dagdelenuysal_d_Page_67.jp2
4df0e468de392e484179b3e218f06118
a5df5e192b1e58c0ebd8109402d7a038badcd546
2095 F20101204_AABFCC dagdelenuysal_d_Page_22.txt
c7f0f872feb3eb85b0c9af662e629242
3cc0f2e29a91e9c499304f8abe614be92f465573
15270 F20101204_AABFBN dagdelenuysal_d_Page_95.pro
562eced5b30fcae5858423cd6ce6c3bb
38d40148922c9e6a73bbfc8c00eba8b80c9ff2f2
F20101204_AABEWH dagdelenuysal_d_Page_87.jp2
64fe2d5ca4ba064c84689300a41ee7ca
a97e02a2c3a53961386d33559f140965592fe170
1051945 F20101204_AABEVU dagdelenuysal_d_Page_68.jp2
f9044c65a17ec9917f2a1ca2109be0da
56e4d306623c93f888581ecdeed2c7b075476393
334 F20101204_AABFCD dagdelenuysal_d_Page_23.txt
43584ea24720a51c9efa1a7a17ec3b24
4897d50f8570deac362940c08c07cfbafac8e341
94 F20101204_AABFBO dagdelenuysal_d_Page_02.txt
24ac950f5d0cfd5c1dc712d76aef8ab2
0ac7da3b0f49c1d0cf33e03ec04383816cb06322
1051978 F20101204_AABEWI dagdelenuysal_d_Page_92.jp2
d7008fdd93d87129cd71c74e0aa36b1e
fbc08a88617b3d9ddc4673e7566599001141371a
823025 F20101204_AABEVV dagdelenuysal_d_Page_69.jp2
b82b0078dab3315f3f072e2f452a0734
d644d3d0fd5d339cb3f618b8a4056d53278a919a
2120 F20101204_AABFCE dagdelenuysal_d_Page_25.txt
13dcb3ef49ba1f0b9f6cabd191169e32
b409521a3582f9c5ce7df788064730e0b9f2dc17
91 F20101204_AABFBP dagdelenuysal_d_Page_03.txt
7f9a72fd68b407a27e87ae9ed7c58616
911d09b3331ddf70e15ac509f84bb211adf38926
1051986 F20101204_AABEWJ dagdelenuysal_d_Page_93.jp2
1ab0f14756dc9849830e23016b230e3a
91cceb7a6d0d85111dd3762bd6ef13479d4ed8de
797315 F20101204_AABEVW dagdelenuysal_d_Page_70.jp2
66afd4c253f9a50e64d385a44735933d
c8fda29c2c165d1d675bfcfda49667848e399c10
1558 F20101204_AABFCF dagdelenuysal_d_Page_28.txt
ad2fa4c756ad402b780987e932958f7c
646631db68ca058da118543e761b66c5561f9cee
1388 F20101204_AABFBQ dagdelenuysal_d_Page_04.txt
7c9cd6680edde985cb35dd1ff41c3894
d270db7a8259fc243c279d684e0a150ea06ed3fb
F20101204_AABEWK dagdelenuysal_d_Page_94.jp2
622c80b0d39b6086c3b150e03cb1a7d9
d60964b944f61136a2adef95f723aee1fbec01aa
980366 F20101204_AABEVX dagdelenuysal_d_Page_71.jp2
25dde61e4c764a086609d70d77568383
dea45b4e5ebc3864f09bcd6c2c08c56895d4bb5f
2059 F20101204_AABFCG dagdelenuysal_d_Page_30.txt
c0caa005f43727e1733fee533de402d3
7ad336a65ac8ed8690889cafb50f018d5c13e218
3958 F20101204_AABFBR dagdelenuysal_d_Page_05.txt
2e30360a75394637320f14a450ce9791
70d6a6863b07a4c3f46d06754e61e61fd9e7472a
366530 F20101204_AABEWL dagdelenuysal_d_Page_95.jp2
191afe955901edea680b364553b4f465
c7fc2b34679dc81072cfe7a68c6dd2525c7f15c9
894216 F20101204_AABEVY dagdelenuysal_d_Page_72.jp2
1592ad745823693b8997a241cee70fc8
a336f786cc6f688c4f85d6754ebc2f246fb4f7e6
F20101204_AABEXA dagdelenuysal_d_Page_20.tif
b0048425f3da703a765767caf235484e
d4c073aa2ee9291ef8e3ad03f3f0fd2571830f46
4289 F20101204_AABFBS dagdelenuysal_d_Page_06.txt
3af4b4435efcf9b51f9509893d6f0fa9
16c6ba110f21aebb9e5aed10fdae356c40d55c73
F20101204_AABEWM dagdelenuysal_d_Page_02.tif
5b73f177c7dec1e1b44530cefd803375
cd1955a273cc290a2808a9fad9daaf4ca91da720
1051961 F20101204_AABEVZ dagdelenuysal_d_Page_73.jp2
0ceeb171da8f8c207d62d8aaa175e0ac
de6951b03321c0f325e8fa6f31ee53c5b03dd1e9
1013 F20101204_AABFCH dagdelenuysal_d_Page_31.txt
dd164d11aac5f2cbbb838c7ecff9e32d
47869b9ecd90936e6f7f05ecb3bc0f9eba59666d
F20101204_AABEXB dagdelenuysal_d_Page_21.tif
716c3d0b07536aa893aae90875274f34
9a71982cbcfdb9c7e8fec5174e5d569da757560e
760 F20101204_AABFBT dagdelenuysal_d_Page_07.txt
dcd6974d2dff9720e09ad0bd8428a234
c0b92fd800d1798eca6e704f77ab290f834fa595
F20101204_AABEWN dagdelenuysal_d_Page_03.tif
9071d1f54ef6b46f9df7c3875ee2fa30
9b9acbd0f384f00eb707ee824896753154a67f51
1938 F20101204_AABFCI dagdelenuysal_d_Page_32.txt
d926d62c32534524d6434f51001e81d8
c38b45d3d4e831d715aafe7ebacea560fd8ec929
F20101204_AABEXC dagdelenuysal_d_Page_22.tif
abf9c63d83ee727be705cfc46a4d24f7
f4bbd2ed524eadffddf1ce013c56afd2a35e60a4
2932 F20101204_AABFBU dagdelenuysal_d_Page_09.txt
f6735d1acd1bb8cabf01df4c85015aab
9201f49426fb52f253530589c3707c30834cf71e
F20101204_AABEWO dagdelenuysal_d_Page_04.tif
97408cac20dcd20cc6debcd73ac10764
e7a340c7064d16a43b19495b7a3004efb87ae978
1086 F20101204_AABFCJ dagdelenuysal_d_Page_34.txt
08bc2a4d427940c446258cf909bb9aa2
da9921cdcd1a196f7e38aab41c4e05db6555bcf2
F20101204_AABEXD dagdelenuysal_d_Page_23.tif
0ed46099cc67f65ea557c058016167ab
4f5f779f643d7e5908b6475012ee11f596a0cfc9
2063 F20101204_AABFBV dagdelenuysal_d_Page_11.txt
a5db3f95ef7b86a4f539a2c5094294ee
9f2178a78d6c065c95f7e57dcd965f559d935a80
F20101204_AABEWP dagdelenuysal_d_Page_05.tif
7f67fc575875731a6ed16c5dae548be9
e732928d2b608b606f36041f72693242023e79ad
1329 F20101204_AABFCK dagdelenuysal_d_Page_35.txt
66d4215b1b30d5bbe67291efea3870e6
e6ce1df98b523ef949d94ef242e140bee35bcbdc
F20101204_AABEXE dagdelenuysal_d_Page_25.tif
844324dccd3dcd049b4bbad3bc179914
9105d6afb0d9fa587e8840d3793659a21aab8f8d
1955 F20101204_AABFBW dagdelenuysal_d_Page_15.txt
6490e73508686228eebee92d76b4b344
a3dfd94d551c960a2f969cee321ccaf622b5587d
F20101204_AABEWQ dagdelenuysal_d_Page_06.tif
eaa13bc75e3e74715d5d190a8096fb67
f6dc728d7f9e61bf37a6bb57734b52a4a3eee46b
2885 F20101204_AABFDA dagdelenuysal_d_Page_63.txt
ea28f262687d5bfee1ad31de7a3dc8fe
d96310276315bae3fab5530b6a0bc487eedd34e7
2228 F20101204_AABFCL dagdelenuysal_d_Page_38.txt
3e0bf8673b98e9500a4f246639661f9e
8a4136c6637b23e1bfd89979bdb93513a03d56c9
F20101204_AABEXF dagdelenuysal_d_Page_26.tif
64b20a00dad8bf3284eaf2062b663d6b
a2db1196603aac6c79349fc2bf9c095abb260d14
1946 F20101204_AABFBX dagdelenuysal_d_Page_16.txt
bbb1f1094ad46290846ba3f897183f3a
0ba7d0ebfac844639918e3d1fae9cac6ec39c6b4
F20101204_AABEWR dagdelenuysal_d_Page_07.tif
480cbf0055ddcfbd66c73591275bee70
0d81f96830e0e0026ea6136f66239921855c8b0f
1289 F20101204_AABFCM dagdelenuysal_d_Page_39.txt
e1e119ed0e854fdab88a6b42b5d0cff3
383fe9fba22084fb6fe24abc430c807807d6a727
F20101204_AABEXG dagdelenuysal_d_Page_27.tif
1649ac2b7a0412cb332d6e5544442106
288efe7eae007ba9f435332002aedad18a919938
1919 F20101204_AABFBY dagdelenuysal_d_Page_17.txt
6af71d38878d30aefbc37412dff7dbeb
0c23f4996a464f19545f13ba43f0466b351ccfbc
F20101204_AABEWS dagdelenuysal_d_Page_08.tif
d3f286ef5f5a2d474e73b187f30bc214
d4955e8acbde1175a22f2ddd3b84f40a13141652
1362 F20101204_AABFDB dagdelenuysal_d_Page_64.txt
c5549b39375cc91d6f6e9942d222cf6b
19f37071df20a9adf589a50fa0cfcf343914e59e
1264 F20101204_AABFCN dagdelenuysal_d_Page_40.txt
b92ba9ed113976ba3c135e4ebfcb4b35
2051de5a688e5e5dd8a8cd4a307ddb30d27cf617
F20101204_AABEXH dagdelenuysal_d_Page_30.tif
fd5fbc36caa42209c66e8619bfa8f3f4
0228c1664b26bae05edcad1d645ff68e5637d30d
1093 F20101204_AABFBZ dagdelenuysal_d_Page_18.txt
8101008ef1f13c7c66c7a5a58bf2fa39
7c44ef6719102e47738397c8fc956ec38e562e17
F20101204_AABEWT dagdelenuysal_d_Page_09.tif
a7a7acd40b0b1bb462dc1da664e98a29
3a734245ace89bf631d64e836d2839530d99e92c
1796 F20101204_AABFDC dagdelenuysal_d_Page_65.txt
df32cca72794907c842a10e9048d90db
cc41c3322ab16294cb29bb3530734ec13595629f
874 F20101204_AABFCO dagdelenuysal_d_Page_42.txt
5d65d5a8d97990af5f3c9bb16ee3f904
68650f510ea74949efae9a37f3f36617ade88f7b
F20101204_AABEXI dagdelenuysal_d_Page_31.tif
c73a251b25bb930b9db22ae3899a3a9c
36b08a3f9e355e9f8a4c989f561ffbec5ab4af23
F20101204_AABEWU dagdelenuysal_d_Page_13.tif
aae37aed46ae56e96c3d506a3d811763
8f0dc842bab98d0ccad85989567cfcd1cba68c68
2090 F20101204_AABFDD dagdelenuysal_d_Page_67.txt
eb3caedd8371b58f5282fd14243efdbc
e5e91f1db5f5aa7128e9b092b4f32fa0582e6f41
156 F20101204_AABFCP dagdelenuysal_d_Page_43.txt
7d4bcce2f9afbccee3e241f82206f72f
17811e60cc61186ec304820142fd78f96aaa4f7a
F20101204_AABEXJ dagdelenuysal_d_Page_35.tif
f86d1ec9ba7ae5b3d5ed38511db63250
fbe9d39d72178f3d34511f44f89fb084f9d74067
F20101204_AABEWV dagdelenuysal_d_Page_14.tif
1134abe6d07e8bfede331fc8b0778281
fb164ded4487c09bf347a6107484b85acb4ce7ff
2252 F20101204_AABFDE dagdelenuysal_d_Page_68.txt
6354b4033f8f2de3b58ed4b5de554c3d
574bb037ddf309851c78c1adb9f0d28cf1985a7b
2658 F20101204_AABFCQ dagdelenuysal_d_Page_48.txt
e867d0fd721d49217ae3aba1cedeb571
c49c3495ce84cecc702ebf330b3a28be8cab84e0
F20101204_AABEXK dagdelenuysal_d_Page_37.tif
11bbb27bd766479bca07357e2bbe09f4
26690c5075b85396b98b95d072c325c7ee02edeb
F20101204_AABEWW dagdelenuysal_d_Page_15.tif
d62adae3f415ff9144e2c11d6a1df1e2
d88fc3b066780926bf268a40f19e5fc2c114742b
1569 F20101204_AABFDF dagdelenuysal_d_Page_70.txt
c4e42161d5926d66375adb8b23c46ccb
84e0ce09d4fefe9cd5fad3bcdee46950a8593e60
3519 F20101204_AABFCR dagdelenuysal_d_Page_49.txt
ada541d8da49e17408ca0563d2899101
5d0040abd8a738966873f31c503c50bd1789b9bb
F20101204_AABEXL dagdelenuysal_d_Page_39.tif
617d3957a9c24a2d436cc3b5683cf4ba
c5762e70e6a8765b4e84fc7b1dba145087ced767
F20101204_AABEWX dagdelenuysal_d_Page_16.tif
566b7faf2b0b48434957cbb7beb8b4fd
5b2de5fcc63dd15fd4bc3b61005536a13050db13
2003 F20101204_AABFDG dagdelenuysal_d_Page_71.txt
3d99862a01ab04c32a5d5f67ec08dfc3
a81b71874340b659f459dada9b2005a6f242eb70
F20101204_AABEYA dagdelenuysal_d_Page_64.tif
680233ae34558e5299630a8b91954d0b
67171d121c903efb9f990614a1656328ab8560b4
2396 F20101204_AABFCS dagdelenuysal_d_Page_50.txt
9963ce6f63f1350fa913733917b6fb12
336c8f4f71fbb3a81353bb0c9ae0791fb34d8aa3
F20101204_AABEXM dagdelenuysal_d_Page_40.tif
88f81de7e8f6e8939789545d03ef0a54
6d2d7b1add1f69dea1aa864479e81f3b1d7b1382
F20101204_AABEWY dagdelenuysal_d_Page_17.tif
7b67271804c4efec3d7ac73b8e819109
73d8b692d1e64277181dc41859016a47ac001747
1819 F20101204_AABFDH dagdelenuysal_d_Page_72.txt
d19681103f50716c6b5df0f69b2ae307
7ffad88f410260a6cdc6257b27cf3a7056932ad3
F20101204_AABEYB dagdelenuysal_d_Page_65.tif
946f08cf85de1ceaa459d0fdb63fc116
b2a49daea93efb0aa248eb56f8ceb0ff56c9d75e
3369 F20101204_AABFCT dagdelenuysal_d_Page_51.txt
c718269cff4897cceec2a41d3915786b
c7b3b5c691826566a9dc368ce5fb38e250198ba7
F20101204_AABEXN dagdelenuysal_d_Page_41.tif
0cd5df0b4354367919053fab8ca2ab5f
cbfc68991c88f129ef95838894f33c90e6a46145
F20101204_AABEWZ dagdelenuysal_d_Page_18.tif
f4233f8672a3fabc2894d7d7c6eee00f
4a62c6b8706eea704b7cee36f49fb700d4ce53eb
F20101204_AABEYC dagdelenuysal_d_Page_68.tif
d93783417e380d72ad3b1ee91afb2594
fa17e153da93f233f9a6e4a40fb4265cec055467
1798 F20101204_AABFCU dagdelenuysal_d_Page_52.txt
8bce907a4c664bc43f046c3a9b0bca15
35409ffba7fb35bc2e60cc4c66845a726c9f92f5
F20101204_AABEXO dagdelenuysal_d_Page_46.tif
4bba27017e8885d7d534198a9d94f196
b1581e165888c1b0c0b30324481155c96c3739fa
1127 F20101204_AABFDI dagdelenuysal_d_Page_75.txt
e8f78effd07cc626fe12764c9bfc6a39
7cd828155feaa054e109d5d115cc5c72d670feae
F20101204_AABEYD dagdelenuysal_d_Page_70.tif
f0d4b9c9504e920fafb26ac83b500b8d
6eed8717ad0ed59b86254dde874a5281680cf4a2
3699 F20101204_AABFCV dagdelenuysal_d_Page_53.txt
dbf9282e05d1b8c9f4b8eb9e6dbe6ac0
781bc18a4250bfb888e972fa0593a9e2035202fb
F20101204_AABEXP dagdelenuysal_d_Page_47.tif
566a9de7b462ff10e114201307c01033
de8152a7ca9a0e796dde661311accd76b0a92f30
F20101204_AABFDJ dagdelenuysal_d_Page_76.txt
04e33a33feecdde9d427256930fe921d
95b146df1fa5d1e2e2f2fe5fd84a4041ccb7f5c7
F20101204_AABEYE dagdelenuysal_d_Page_71.tif
fce5a9a6eba53015ee6044a53dd45b16
910afb86433b90f083c2ab57957d6599e4676e55
F20101204_AABFCW dagdelenuysal_d_Page_57.txt
f046a76afa3baefe2bf96f30cdad9f99
1ae96f2b778f3572ab1a3c722b755904b2c517ed
F20101204_AABEXQ dagdelenuysal_d_Page_48.tif
9b157c86c2d76032cceacf373832c588
7f0613d37177f8b97776da24e9a01f25ce5294eb
1476 F20101204_AABFDK dagdelenuysal_d_Page_78.txt
b861c2b891d2889328c466dacb6955c7
758b31ea6b34d515edba16fe37e4fabc3bc7d898
F20101204_AABEYF dagdelenuysal_d_Page_72.tif
705f8da8b9771c98e17164ca563e3fce
a8aff2b4c71d0d909d45d6769951cf36025be609
2082 F20101204_AABFCX dagdelenuysal_d_Page_58.txt
16961162b4e85678a24916c549160fdd
298fd34eebd253d1c9420d605b33ba612482956a
F20101204_AABEXR dagdelenuysal_d_Page_49.tif
d9ea060a9e153627bcf853f5e4328347
68ebcb9c581b294bec4b955f0c44af2aef345fb6
21520 F20101204_AABFEA dagdelenuysal_d_Page_05.QC.jpg
1957fabe1c69ed65626460ac3177e101
967792a05fdd2236cfd610a6d89b769a20d6650a
1378 F20101204_AABFDL dagdelenuysal_d_Page_80.txt
d3da865232cae1d091ed9d59c1d00b7a
ab48f2c7ffee08c9ff1af175d45af9e6208da261
F20101204_AABEYG dagdelenuysal_d_Page_73.tif
5df69d938b0dbfb81550462fe60ca74d
063e8d9e560d4ff7539e7630fdc68a6184c7d67f
1477 F20101204_AABFCY dagdelenuysal_d_Page_59.txt
e9a8e0aa02bea3ac0984392abf298509
f66341ddfb30f5600d554ac594f34a79b5039ad6
F20101204_AABEXS dagdelenuysal_d_Page_50.tif
3d954bc0eb6c8fd9ddc0d05611ccee4b
1f1e827f222f928276fdc4ede2eacd907b1f9847
24325 F20101204_AABFEB dagdelenuysal_d_Page_06.QC.jpg
73a6c63c70d57420e97274a2a67c4e1e
f5b33840a574db53b4b174ea48e573f3ecca60af
1380 F20101204_AABFDM dagdelenuysal_d_Page_81.txt
1081a802211db39d8bd2dce1f22e5aeb
2bfa09c99abb924149671e548b1c0dfc20bf1ec6
F20101204_AABEYH dagdelenuysal_d_Page_74.tif
28f9956589782cf5eb5f155a8b09ae8e
34783304cc44c0ebe3200c83db1783f777a87469
1112 F20101204_AABFCZ dagdelenuysal_d_Page_62.txt
9ee676897a65f4efa6d1a4dfc6352f11
454754566fde9ccecedd5239140ffb75939cf55c
F20101204_AABEXT dagdelenuysal_d_Page_51.tif
25433ce1f83ca143c6997c41421cc3b1
af91a3ebc6a691ac9936e94f2ce228c461d75447
5989 F20101204_AABFEC dagdelenuysal_d_Page_06thm.jpg
82f48a5a67ea6b98dfdcf019126c0564
a52d77c52bb335d6cc78a58ab76707b08179430e
1163 F20101204_AABFDN dagdelenuysal_d_Page_82.txt
f09a864dfaa3fc026f621f71a7ba376c
7629d0bc7e04caaa007a99c140b38f174076222f
F20101204_AABEYI dagdelenuysal_d_Page_76.tif
719ffdbce1e8b5eefccca239a322e59a
199461347b6a0ae6c728c0a63a02fe6487b2b10d
F20101204_AABEXU dagdelenuysal_d_Page_52.tif
21ee6cf7d7db6fd7cb712aff3a8fa25f
89c2b35d43740ec20e65da7d1a9249086b38fbe1
23545 F20101204_AABFED dagdelenuysal_d_Page_09.QC.jpg
e5a46a9c064c16d5571f7f36cbfe56f2
042e5d156dc9183b67f5642b73475fccd2c67caa
2789 F20101204_AABFDO dagdelenuysal_d_Page_83.txt
40988b274fa6f3efe5b4168c75efdf72
a7214de0563ec78dfd383677a69a31b6a2ac7b7a
F20101204_AABEYJ dagdelenuysal_d_Page_79.tif
cfb10b21001627dd84070016ea6e2b5a
e4938355617442b776c5632c9ea1d3bbfd6c06fa
F20101204_AABEXV dagdelenuysal_d_Page_55.tif
74f5039dd139c614a812bae70cd2da0c
bba6613c4f871c143fc077cc9c5d13e46720d0bf
11671 F20101204_AABFEE dagdelenuysal_d_Page_10.QC.jpg
7b455c048dc0d29bb90db54d1f745347
fdc48d0396128979fa2dc8a5d7cd516504127f08
92 F20101204_AABFDP dagdelenuysal_d_Page_85.txt
2966d1484252aad9741e7e5daa2cdcd8
eba07658e23774ef0e8a6dc515c6f2741725ca30
F20101204_AABEYK dagdelenuysal_d_Page_80.tif
25efb91a55b0941d97598f309b07ba74
f671ada8084453fd38dbf4d7be6bfb0f47d253d0
F20101204_AABEXW dagdelenuysal_d_Page_57.tif
830a8029c7123708c18da9664f4e0023
ac898c82794a2bf1c373cfc137a7b9bee40d7c6d
3497 F20101204_AABFEF dagdelenuysal_d_Page_10thm.jpg
e95d56cbea7a71363e6bf37ce1f931fb
ced64267324536495c20bb8ff669331d08a23a5e
143 F20101204_AABFDQ dagdelenuysal_d_Page_88.txt
f802ea246adcff022bedb747febababd
7b56ebeea1aa39a38c4f1b916f82aa5d749929e1
F20101204_AABEYL dagdelenuysal_d_Page_82.tif
3f8fbb7f86234a9c605e608a1ba8ed5e
70d2588069051522a29f249b163d369d42ab14f6
F20101204_AABEXX dagdelenuysal_d_Page_59.tif
7dd0180ee296d701aa6ea1249a5416d0
de5d0c7eabcb412c43a818309afa0c361f4cbce5
24672 F20101204_AABFEG dagdelenuysal_d_Page_11.QC.jpg
1ba40a24fafb05b07f12f6de30829901
631d034a843ed9bc5f3e1f17d616af6aceaeab32
530 F20101204_AABFDR dagdelenuysal_d_Page_89.txt
92c64551debbf3d58f4ecc823e1bebbd
ac250d5d857c55bef23556f1b7a181eee552f6d2
F20101204_AABEYM dagdelenuysal_d_Page_85.tif
74eef795c565ce9a098771f5ae5bcae9
e61eea4b595a58c2fc841b2d5113e2b6f0eec993
F20101204_AABEXY dagdelenuysal_d_Page_61.tif
77b83c0205118eb8fced54c88d9bd404
0f9f062e61ec2572b0f42dbeb4414abefc36371c
6913 F20101204_AABFEH dagdelenuysal_d_Page_11thm.jpg
30ac4af375b6de9094ca3e633f2ec5af
f4d858c20692f16ef6ed1af2905499e4fdd6dccf
73490 F20101204_AABEZA dagdelenuysal_d_Page_09.pro
3357e6d7d9d22edace2cd45b009d8923
17f04ff2ec0e2bab649b66eb99cb6b4463c9d91e
307 F20101204_AABFDS dagdelenuysal_d_Page_90.txt
152132829701d69b1979619b6ca95dab
c945ba1a4623a8cfd55e44a6a9e001ba8ac9957f
F20101204_AABEYN dagdelenuysal_d_Page_88.tif
ee26da27c569ddcd16019ff9db23bf27
0c26aabd3fd27a8e270e7a270e919e2f6b4ddbc3
F20101204_AABEXZ dagdelenuysal_d_Page_62.tif
955436ed931ceef90839dbc93aadf512
40942048ffcdb4df3208123e6004dc6546a252b0
5785 F20101204_AABFEI dagdelenuysal_d_Page_12thm.jpg
2eddc89040b0e20b91d728db89822c68
9f851bf6e9dc43557e58e771cfad0c3ccb5b9bd2
47624 F20101204_AABEZB dagdelenuysal_d_Page_11.pro
19b025ae97bf6546acb3c640c5f3fb97
7f1a81c37fd1f5420ce038a1ef050ba7ccac5d1b
2074 F20101204_AABFDT dagdelenuysal_d_Page_91.txt
9e9f07f65c39f906df3a682cf1b17dba
4d8fa44d7116189c0c61ba20e02263a2fb6578e0
F20101204_AABEYO dagdelenuysal_d_Page_89.tif
0b0a0b74d3aa8eab06d9d934b5022454
3f19bdfd7fa338c1e82053661b3ac54be789001c
38165 F20101204_AABEZC dagdelenuysal_d_Page_12.pro
3e502f200321e185857e8ea819742420
82966d08a58d5d979e6ace66c4332b70c4cf177c
2294 F20101204_AABFDU dagdelenuysal_d_Page_93.txt
249a7991cee48f7f8f45fa3303433af1
cc4ef44fa09bb2b526f07f123df64758a03187f4
F20101204_AABEYP dagdelenuysal_d_Page_90.tif
48696fdd0deb17fd35f7158db969f480
804ed01772574192eb4f519a8bb5e378640efdac
26013 F20101204_AABFEJ dagdelenuysal_d_Page_13.QC.jpg
96db3b7aa466fd74e1e8c6233e9ffa3c
945bc835b88f41ce1ae7009bd14b807eabd7228a
46408 F20101204_AABEZD dagdelenuysal_d_Page_14.pro
f9fff63a8e1d0bdcde60404ee34e2543
c2cf3d16d4a939575b9361f1f29f0032d4044047
2241 F20101204_AABFDV dagdelenuysal_d_Page_94.txt
808fda8f99f88effc3cf47a514f16640
a925734a18dc4989009b544c99209b17ca521ab2
F20101204_AABEYQ dagdelenuysal_d_Page_91.tif
17bec90f0dba46d03b323ea4f194528b
539a8f1687458848f8651d2dbe359a6105947511
6604 F20101204_AABFEK dagdelenuysal_d_Page_14thm.jpg
a00966f478a8b8c4a6f7c42c60106387
b8a9a785db0ac8b08bb3ac0f0b652d6763d2c761
46568 F20101204_AABEZE dagdelenuysal_d_Page_15.pro
84da45475063d10728827e941f0635fc
5a5544dcbc80689afa7dab848aa9cca333ebedef
1545823 F20101204_AABFDW dagdelenuysal_d.pdf
2dfbd462db45c1d1a12f7eb123c2489d
8310906ef6a48409ce5fb31b8d9203ddcb9da721
F20101204_AABEYR dagdelenuysal_d_Page_92.tif
7b41cab57c9317c6a6a7a572317d6aaa
1da08dbb46e1094d988c5e6b211007870e1adf87
6702 F20101204_AABFFA dagdelenuysal_d_Page_29thm.jpg
594b91bcfd59ae57b06d2b8eb810f5e7
0fe7a457fa9113ae7b58b828a2a17e65fdd8dbdb
24404 F20101204_AABFEL dagdelenuysal_d_Page_15.QC.jpg
9e0051ccc6d5d01c5e83f72580ab5d30
cf80a2451dc7f981681fe44173b4731c1e7b0729
F20101204_AABEZF dagdelenuysal_d_Page_17.pro
039bf424e0feb45378d59a2414c48977
1df8c2bd40b6a63e7bb8d8706459b145903b5217
8075 F20101204_AABFDX dagdelenuysal_d_Page_01.QC.jpg
07e62ddc9b84af57e038a32f42a231fb
75920aead59c85d83c42d97b3d586f636dda9e31
F20101204_AABEYS dagdelenuysal_d_Page_93.tif
e98fd7f14bafa1e9ccedd946796c0a48
2c40dc7adb06aff8856246bb0fcea8a85db3c0c1
7231 F20101204_AABFFB dagdelenuysal_d_Page_30thm.jpg
34fc1aed01b0572a37179a5489119b4b
785732e445797d52d33bdaafffe74ddde62166f5
6959 F20101204_AABFEM dagdelenuysal_d_Page_15thm.jpg
df3b6ec5feab5c696f3d5fdb05659086
06e7a99f96a2f1b0cd19fc7bd7ce912ea2dd5664
F20101204_AABEZG dagdelenuysal_d_Page_18.pro
63fe43d5995901e0d0596f0ce907e0c3
596a0485dd614d3b6e4a9b22f0cc092a4354851e
2399 F20101204_AABFDY dagdelenuysal_d_Page_01thm.jpg
5f1dbaaa5759991e1492828cee6e48e3
728459633e28cc828f0aefef0cd300ef9ee806b3
F20101204_AABEYT dagdelenuysal_d_Page_94.tif
dddc147aaaad814989e56a1a359ed24d
0e3b6953ce8cad1b76070f8cf21a06d5a6f6d67a
16318 F20101204_AABFFC dagdelenuysal_d_Page_31.QC.jpg
51e5e9fe14354429b9865ed6573d8b8b
b5d311b8171d375cc42cbba7aee716bf3bbd7563
24657 F20101204_AABFEN dagdelenuysal_d_Page_16.QC.jpg
a076d5420b395f632c0c432899ad862c
d0320140886e817183e8d53a267997887fc6f978
54652 F20101204_AABEZH dagdelenuysal_d_Page_19.pro
5fd21f64705f96546aed92f86532c25d
a6ca6924339a75ac5de1985f0ea5c9133c021286
17826 F20101204_AABFDZ dagdelenuysal_d_Page_04.QC.jpg
279394a9f5fb4c63fe096b369ba595a6
966fe0249a35d8dcdcb63c0a4016cf5422924500
F20101204_AABEYU dagdelenuysal_d_Page_95.tif
04e212ce46a69cce95b6d77af73f7218
b3a2be0dc00b238d443d41f90efaa3d6ee252850
23927 F20101204_AABFFD dagdelenuysal_d_Page_32.QC.jpg
3d5b2340a00a525a6904327c0463baf3
535005414a76d661a0c409736da9700694154294
6954 F20101204_AABFEO dagdelenuysal_d_Page_16thm.jpg
15e89591053213e3832e87b71cea612b
11a34f16514fc835d4f1a002f65fab74b3b2ed20
49960 F20101204_AABEZI dagdelenuysal_d_Page_20.pro
ea1c0f34f51d165f32e6516fe879f8a9
619446fc506c54911300854f65b93c1a729c4cb4
997 F20101204_AABEYV dagdelenuysal_d_Page_02.pro
9003d54148f5ab7d8a925961ea5206ff
f473bc341950d00dfafba256655f02921d2ba051
16898 F20101204_AABFFE dagdelenuysal_d_Page_34.QC.jpg
61d3a925cca92ef52c15631e4cddd30c
b80acac2bb165149fb382ca4d1a625bcd8c570f8
24734 F20101204_AABFEP dagdelenuysal_d_Page_17.QC.jpg
e66ca5e39919fedd3c69b2090491c614
c289923fbc6d20b338286f4fca94a64b5630e201
43875 F20101204_AABEZJ dagdelenuysal_d_Page_21.pro
c0ad84ee73b7861fcaacc4eefba3c64d
c0feca93586b2b8ccbc2326bfd54e0fe9d875ae2
917 F20101204_AABEYW dagdelenuysal_d_Page_03.pro
68a602e169515d5b53a4cedcc5eca8da
670e36c3d4acfb40d546a1555a3243d0fc8c4e33
5430 F20101204_AABFFF dagdelenuysal_d_Page_34thm.jpg
e7fbe362001bd28ec566109edecae907
b9575cd11e671e11de42e4ec9136ff306da6195e
6781 F20101204_AABFEQ dagdelenuysal_d_Page_17thm.jpg
794c9d99027b3d3d354679e1b2daa427
26bffccc3408e1c739a23000be736d58af2c93f5
8346 F20101204_AABEZK dagdelenuysal_d_Page_23.pro
682c1769091521454bbd8ca0f2f86cd3
3030313eb31abb2f4dd0afcec5e5a755ad4d07ac
33721 F20101204_AABEYX dagdelenuysal_d_Page_04.pro
041d91c33f267c5b4c8ed9497ec7ae8a
21d3938df7b24aab1e2fa02072bcc328f195a14c
17717 F20101204_AABFFG dagdelenuysal_d_Page_35.QC.jpg
b4291a3a87a8aff020cf07fe72a31688
d6c3fc9fe5449c811689a0090422061322b269f7
5456 F20101204_AABFER dagdelenuysal_d_Page_18thm.jpg
717777f4a241f15454eaa6fe1a5129be
eb5c82af9a571d019af3be3d6bf414362662dcfb
52802 F20101204_AABEZL dagdelenuysal_d_Page_24.pro
ecab9abe4c854599d4d442ad1b532111
fcc7e93b5882d22bd7e2ee250b7b69f97578b711
104765 F20101204_AABEYY dagdelenuysal_d_Page_06.pro
4cd2caa56d19b99d91f5fb21fafcd052
ff27ab72204743d69a72678053823c907c0bdf74
5559 F20101204_AABFFH dagdelenuysal_d_Page_35thm.jpg
81d75db602c1600fad94ce9175353154
7867a5feaccd97ccc15509a4d145afcfa305fe50
7523 F20101204_AABFES dagdelenuysal_d_Page_19thm.jpg
c2066a74fa657570652ad90dd4b98706
9bec730ccc04c03b0ddeb72b3bc088b8fe7509e7
53691 F20101204_AABEZM dagdelenuysal_d_Page_25.pro
b3cec9b3404fc40828e2988e879bfa6c
5b58333bee2b558ac602b397c9a385e95a52310d
55325 F20101204_AABEYZ dagdelenuysal_d_Page_08.pro
9bd3367dad181f27e478efea203ac612
076b91517123d629f1a1ad36d4fa59cd72dcb3c0
24465 F20101204_AABFFI dagdelenuysal_d_Page_36.QC.jpg
d86e4b38fb3d095c4e207ebcee471717
e5c86030e7a0060c60e726d02037ba6938c20667
23115 F20101204_AABFET dagdelenuysal_d_Page_21.QC.jpg
9fc2e762278d707050d3934b849e3102
45ea7520ad23c29d955228c2f1948ddd4edb5983
33949 F20101204_AABEZN dagdelenuysal_d_Page_28.pro
fd8a2d3f28bfd97525b5a5c229e26cc7
3ddef3b36cfed72c84c25b5aa23b11f556d742e2
7052 F20101204_AABFFJ dagdelenuysal_d_Page_36thm.jpg
bc1142ee147db826b95844ffe3294f43
e58723117829ac6838c525b88a4c742fb1615b03
7555 F20101204_AABFEU dagdelenuysal_d_Page_22thm.jpg
b75459143436d69e01a9ec3f26acddff
5b0e10d3013c23352fdd00fdbebcd15a84e192ac
24332 F20101204_AABEZO dagdelenuysal_d_Page_31.pro
65f0bbdaec2de3ae5b648b2de1fee974
892c53946fae70747002778b0257c10c5cd6a1f0
6770 F20101204_AABFEV dagdelenuysal_d_Page_23.QC.jpg
e0ffca91e143cf8f29cc3f5610f88708
eb5ef85c55af38bd5dc93c0583a14cf39f08046f
10080 F20101204_AABEZP dagdelenuysal_d_Page_33.pro
b392a9c738e9554ead7a6ed582db3ad1
74d99028650611d0176c1da29fde8e1c0ff75698
18939 F20101204_AABFFK dagdelenuysal_d_Page_37.QC.jpg
ab746c98e4593c0071e064c6baa27a86
fcdc562044c24c76c6d9c730a820b67fd50769c9
2311 F20101204_AABFEW dagdelenuysal_d_Page_23thm.jpg
9071f10f4aed13c0e946cf21a3ce095e
365278de2d9de32e29e02fe7a88333c1da6e05b6
49505 F20101204_AABEZQ dagdelenuysal_d_Page_36.pro
a6222cd104f8e824c24a40890a1ca564
699471b29fdb225b75d0147839b97cba63018f1b
5459 F20101204_AABFFL dagdelenuysal_d_Page_37thm.jpg
5b4e382083a75aab43084006df959140
9b02b7ed51bcb94e16e760be22ba22924916338b
27791 F20101204_AABFEX dagdelenuysal_d_Page_24.QC.jpg
e4ca5f774515ff987ad9ecded626a400
ed1c7784546ce3eebfad35a4d7ecbda1c9d93369
24453 F20101204_AABEZR dagdelenuysal_d_Page_37.pro
36153c5e025055e611f775a5f5a447b7
4337de2a7f1c3313cf08ec2e4e8764eb1768d0e4
22764 F20101204_AABFGA dagdelenuysal_d_Page_49.QC.jpg
958d1559e152074b65c38cd3b127b585
34f701a4a70c0c6f18ce1a4434427c0851a91acd
26085 F20101204_AABFFM dagdelenuysal_d_Page_38.QC.jpg
78534eccccdceca47a4623aa3f665d2b
e91390266ebf7ba9e538b6509820f20f53c781e7
19733 F20101204_AABFEY dagdelenuysal_d_Page_28.QC.jpg
596227ec56d01bb2a39b5cff92b0b5d4
6bff21870369079b70df62d64134f82c75bc0baf
49500 F20101204_AABEZS dagdelenuysal_d_Page_38.pro
4015fcf6b2409feb7d5ecafa59857574
730416f36ebc80e8b684f376d21b0f49fce11ad9
6727 F20101204_AABFGB dagdelenuysal_d_Page_49thm.jpg
633c79710f996b2fa60ed41f813cd43d
e728612337f4fec6a1c12e84157979b0d82d401f
6921 F20101204_AABFFN dagdelenuysal_d_Page_38thm.jpg
9cad42994c805a14a75356e21990d373
fdafc7461b7bf6571554c89d8c23c47e17e657fc
5698 F20101204_AABFEZ dagdelenuysal_d_Page_28thm.jpg
f09d5282255bc675f52a3c9a9cc8020c
54518fe5f55229f5e7666671ddc7ff3e05c15e98
30248 F20101204_AABEZT dagdelenuysal_d_Page_40.pro
92573ed89806adee4269e36f4ee402f6
2907af31a0a7ae58f959972ed7e67d236c67b431
21067 F20101204_AABFGC dagdelenuysal_d_Page_51.QC.jpg
f8333f4407cada8fb43d5fd5120aa212
4c74920cec5697b7528958754e26ad50b33cafb7
15900 F20101204_AABFFO dagdelenuysal_d_Page_39.QC.jpg
8de0999de0e898c1feb25485371127ff
db97a4a9ef943d19c1860a3da0e31ce3b1696e06
46732 F20101204_AABEZU dagdelenuysal_d_Page_41.pro
e12fd41ee2139ec45245975934602ef5
1cec0d08ff573074edab425d2c23547537c52688
6178 F20101204_AABFGD dagdelenuysal_d_Page_51thm.jpg
a8ad78970d03147c5d13bf86a25bf0f4
1e0c9b9db4474f047a60964531d4d5a232a14229
4852 F20101204_AABFFP dagdelenuysal_d_Page_40thm.jpg
c1d3d494be93d2e8ba02fd4c0560ed1a
6627e83f95f6ffb210b22fde082ba47689992865
21122 F20101204_AABEZV dagdelenuysal_d_Page_42.pro
8a400eb7badb4e5c7686d416c2b9ea9a
f37ba319d216de4f9355f3cbe39b86bd85ea7734
5870 F20101204_AABFGE dagdelenuysal_d_Page_52thm.jpg
86e7f8f12c7469efccc12bcccf8eae45
0eb6752d8ade1a55b2ce3515008a9d7181f0ab5a
6669 F20101204_AABFFQ dagdelenuysal_d_Page_41thm.jpg
ec83a8c17997e1c6e6069ef2337ca13e
2969084611c97f90877b7fdfc0626bf094bef4ce
2174 F20101204_AABEZW dagdelenuysal_d_Page_43.pro
ef97b37d9d90310ac7fc713efe4a8850
a68cf079c2cce71392765b1e8137d2dd6fa990cc
21592 F20101204_AABFGF dagdelenuysal_d_Page_53.QC.jpg
0fe8db6d2511593210b1474c1592bbf0
94ae1ba5024156cdc92d18bbd6f380786f86428f
5412 F20101204_AABFFR dagdelenuysal_d_Page_42thm.jpg
0a0bac0ef4e886389279ef8ccff0990b
69db05c2b034c1f7a378a58b7f583b2e7a6d7f40
27487 F20101204_AABEZX dagdelenuysal_d_Page_44.pro
8fd2a4ecc536f4200c1e9c818872c025
ad9a182b1d94257ab95a6766d66a6276c071d32d
6627 F20101204_AABFGG dagdelenuysal_d_Page_53thm.jpg
7d060996bd662fb1909a532f891861e8
14df7d216eed61a39c8ee3e0a7b60e56a5fa0ae4
8770 F20101204_AABFFS dagdelenuysal_d_Page_43.QC.jpg
db314f19f8593ee2463405882127f4ce
4882a576fe561bf30cc776a0d77d57450230077d
12846 F20101204_AABEZY dagdelenuysal_d_Page_47.pro
47d8c27e7dc92b60da55d0305813bc19
eaae93c9ef9cfe7f09fd7dcada1b4722c70bee43
16713 F20101204_AABFGH dagdelenuysal_d_Page_54.QC.jpg
fa5634181765e9f4150edcd9f5d473ad
756fb49a184f7e2633534c77ac6bbcf298ab7161
19721 F20101204_AABFFT dagdelenuysal_d_Page_44.QC.jpg
bbc6d134e769de506a06838f4d73e313
861e39b25eec26f53d34747816e6f3826c5e9d23
13251 F20101204_AABEZZ dagdelenuysal_d_Page_48.pro
09f5b94c652352f715b316b26e7466ec
983f553e7c2f30f1bea74356cfc6a35ebe0331ce
23273 F20101204_AABFGI dagdelenuysal_d_Page_55.QC.jpg
07d86e6d7b32f04bd8e4f42bd67dd31b
82257c2f89fcf11a735fbce73419efbfe7db38f3
5997 F20101204_AABFFU dagdelenuysal_d_Page_44thm.jpg
8de0a618e3a4b5873270a64a48ee40d3
d484fc54c0361c57b7e59f3f41248eace1284726
6487 F20101204_AABFGJ dagdelenuysal_d_Page_55thm.jpg
9c8ddfa88aa0f0283446a8f2cdde1070
1497863d5a6aaa47abcee833a63c0c3973844330
17981 F20101204_AABFFV dagdelenuysal_d_Page_45.QC.jpg
ebc23e0c50d5307e728c258bae457d08
f81943a6532c5661a041dc6c6a2ef4b911626f26
17932 F20101204_AABFGK dagdelenuysal_d_Page_56.QC.jpg
7cef680216e971f3d156184e877eeece
dddc05d73692ffe9082b6728fa9063185b84b792
5599 F20101204_AABFFW dagdelenuysal_d_Page_45thm.jpg
d1a0dffc9acfcc3d82ee8e5f6ff310b6
94fb6c51546c33a84603ca7d54440233df1eaeb8
16576 F20101204_AABFFX dagdelenuysal_d_Page_47.QC.jpg
f4bcd1423cf2fbbf6fb0549da6d01ee3
ca063426e4f4a40b5b84002c81e0d35cc2564628
20934 F20101204_AABFHA dagdelenuysal_d_Page_67.QC.jpg
0bb43b7a75925d302d7f4363605e54c0
2d8cb809ce0ff8dc21720b79413ad9b57d219ba8
5499 F20101204_AABFGL dagdelenuysal_d_Page_56thm.jpg
eb1aaea4eb41bb658329ec8a6e774c46
acfcdf710e1f2e1b54fff851755c15ec0cd8882e