THESIS

' LIBRARY
W Michigan State
University

This is to certify that the
thesis entitled

THE EFFECT OF ANTENNAE CONFIGURATION, PRODUCT
AND TAG TYPE ON THE READABILITY OF PASSIVE UHF
RFID TRANSPONDERS

presented by

THOMAS JOHN SILVER CRAWFORTH

has been accepted towards fulfillment
of the requirements for the

MS. degree in Packaging

 

 

fly Major PW Signature
(29:; / g 010 of

Date

 

MSU is an Affirmative Action/Equal Opportunity Institution

—. _. _. a. .- n- - _ _

 

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

 

DATE DUE

DATE DUE

DATE DUE

 

APR 2 12007

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 m/CIRC/DateDuejndd-p. 1 5

 

THE EFFECT OF ANTENNAE CONFIGURATION, PRODUCT AND TAG TYPE ON
READABILITY OF PASSIVE UHF RFID TRANSPONDERS

By

Thomas John Silver Crawforth

A THESIS

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

MASTER OF SCIENCE
School of Packaging

2005

ABSTRACT

THE EFFECT OF ANTENNAE CONFIGURATION, PRODUCT AND TAG TYPE ON
READABILITY OF PASSIVE UHF RFID TRANSPONDERS

By

Thomas John Silver Crawforth

Many advances are being made in the supply chain world at this time. One of the
more prevalent additions is the use of radio frequency identification (RF ID). With
mandates being issued by powerful retailers and government agencies, numerous
companies are being forced into implementing this technology. The use of 915 MHz
RFID systems is rapidly growing and there is a need to better understand its use in certain
applications.

The purpose of this research is to determine if the number of read antennae used
and the location of these antennae has an effect on the readability of passive UHF RFID
tagged cases on a stretch wrapper. The variables for this testing were number of read
antennae, location of read antennae, case contents (empty, rice and water), and tag type
(Alien Class 1 and Symbol Class O).

The research found that there was a significant difference in the read rate of reads
for both number and location of antennae. There was also a significant difference in the
read rate of reads for different case contents with both rice and water decreasing reads,
though water was shown to have the greatest impact. The tag type was not found to have
a significant difference with respect to total reads; however there were differences seen

within specific combinations of variables.

Copyfight by
Thomas John Silver Crawforth
2005

ACKNOWELDGEMENTS

There are many people who aided in the completion of this thesis that I would like
to thank. First of all the members of my committee, this would not have been possible
without your support and guidance. To Dr. Ken Boyer from the Eli Broad College of
Business, and Dr.’s Paul Singh and Robb Clarke from the School of Packaging, thank
you very much. A special token of gratitude is due to Dr. Robb Clarke for taking the
time out of his constantly busy schedule to be there to answer my questions and guide my
research.

I would also like to thank the members of the statistics department that helped
validate the significance of my findings. To Dr. Dennis Gilliland, and to an especially
dedicated graduate student Jing Wang, many thanks.

I need to also thank all my fellow packaging students who aided in my testing,
and building of test equipment. Thanks a lot Jon, Tom, Cortney, Nav, and Farren you
guys were always supportive and great to bounce ideas off of.

Lastly I would like to thank the members of my family for being loving and
understanding throughout the entire process. Mom and Dad thanks for your support in
every possible way, I would not be where I am without you. Libby thank you so much

for encouraging me and being understanding, I love you.

iv

TABLE OF CONTENTS

List of Tables ......................................................................................... vi
List of Figures ....................................................................................... vii
Introduction ............................................................................................ 1
Literature Review ...................................................................................... 6
Methodology ......................................................................................... 1 7
Equipment ................................................................................... 1 7
Procedure .................................................................................... 18
Results ................................................................................................ 28
Case Content — Total Reads ............................................................... 28
Tag Type — Total Reads ............. ' ...................................................... 3O
Antennae Configuration — Total Reads .................................................. 32
Antennae Configuration — Location of Antennae ...................................... 37
Antennae Configuration — Number of Antennae... .. .................................. 40
Conclusions .......................................................................................... 45
References ............................................................................................ 50

LIST OF TABLES

Table 1. Configuration Identification ........................................................... 23
Table 2. Legend for Side View Diagram ..................................................... 23
Table 3. Data Collection Sheet .— Example 1 ................................................. 26
Table 4. Data Collection Sheet — Example 2 ................................................. 27
Table 5. Variations in Configurations with 2 antennae ..................................... 43
Table 6. Results for Empty Cases ............................................................. 44
Table 7. Results for Rice Filled Cases ........................................................ 44
Table 8. Results for Water Filled Cases ...................................................... 44
Table 9. Results for Case Contents ............................................................ 45
Table 10. Results for Tag Type ................................................................ 45
Table 11. Results for Number of Antennae ................................................... 46
Table 12. Results for Location of Antennae .................................................. 46

vi

Figure 1.

Figure 2.
Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.
Figure 8.

Figure 9.

Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.
Figure 15.
Figure 16.
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.

Figure 22.

LIST OF FIGURES

Wooden frame that antennae were mounted on .................................... 18
Alien squiggle tags on cases .......................................................... 19
Tag position on pallet, by tier ......................................................... 20
Pallet load on stretch wrapper ......................................................... 21
Symbol double di-pole tags on cases ................................................ 24
Water bottles in cases .................................................................. 25
Rice jars in cases ........................................................................ 25
Percentage of Total Reads on Empty Cases ........................................ 29
Percentage of Total Reads on Rice Filled Cases ................................... 29
Percentage of Total Reads on Water Filled Cases ................................ 29
Percentage of Total Reads on Alien Squiggle Tagged Cases ................... 31
Percentage of Total Reads on Symbol Double Di-pole Tagged Cases... .. ....31
Percentage of Reads for Configuration 1 .......................................... 34
Percentage of Reads for Configuration 2 .......................................... 34
Percentage of Reads for Configuration 3 .......................................... 34
Percentage of Reads for Configuration 4 .......................................... 35
Percentage of Reads for Configuration 5 .......................................... 35
Percentage of Reads for Configuration 6 .......................................... 35
Percentage of Reads for Configuration 7 .......................................... 36
Percentage of Reads for Configuration 8 .......................................... 36
Percentage of Reads for Configuration 9 .......................................... 36
Percentage of Reads for Top Antennae Configurations ........................ 39

vii

Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.

Figure 28.

Percentage of Reads for Side Antennae Configurations ......................... 39

Percentage of Reads for Combo Antennae Configurations ..................... 39
Percentage of Reads for Configurations with 4 Antennae ....................... 41
Percentage of Reads for Configurations with 3 Antennae ....................... 41
Percentage of Reads for Configurations with 2 Antennae ....................... 42

Percentage of Reads for Configurations with 1 Antenna ............... . ......... 42

viii

CHAPTER 1 - INTRODUCTION

It is generally accepted that Radio Frequency Identification (RFID) has taken
great strides in finding a place in everyday life. The technology is being successfully
implemented in many areas, and some are taken for granted. Certain gas stations are
using it as a contact-less method of payment; it is being used on highways as a method of
high-speed toll collection, in buildings and parking lots as access control, and even in
sports applications as a method of race timing. While these applications are in place,
there is an ongoing push for the technology to be used for asset tracking within the supply
chain. Before it is considered a success in this area, there is still much developing to do.

The benefits of implementing RFID into the supply chain are numerous.
Inventory management, accurate order picking and delivery, reduced labor, decreased
theft, smaller and more focused recalls, and a streamlined supply chain are all areas that
RFID can potentially help. Improvement in any of these areas, let alone multiples, can
create a significant competitive advantage for a company.

Initially, these benefits may be most easily realized in a large supply chain
situation. There are two examples of extensive supply chains that would benefit greatly
from implementing RFID. The first is the United States Department of Defense (DoD).
The DoD has over 43,000 suppliers [1], shipping to locations all across the world.
Organizing and streamlining a supply chain such as this is an ideal application for RFID.
The second example of an extensive supply chain is the retail giant Wal-Mart. Wal-Mart
is the largest retailer in the world, with sales for 2004 surpassing $250 billionm, and has a
supply chain to match. Implementing RFID could dramatically alter these supply chains

for the better. For this reason Wal-Mart has issued a mandate that its suppliers begin

implementing RFID, and the DoD issued a similar mandate shortly after. Wal-Mart
undoubtedly knows that the requirements in the mandate are difficult at best, yet they
also know that the industry would be less active in this area if not given some strong
motivation. Certainly Wal-Mart is a large enough company to put pressure on the
suppliers to accept a requirement for RFID adoptionm As expected, Wal-Mart’s
competitors in the retail industry, and other companies from different industries are
following suit and have issued mandates of their own, not wanting to be left behind.
With these mandates in place, the technology is being advanced much more rapidly than
it would have if the retail giants had not stepped in and forced the issue.

Implementing RFID technology is not a simple thing. That is likely a reason why
Wal-Mart felt the need to push this onto its suppliers because, while the success of the
technology will benefit both sides, its difficulties are a strong deterrent. Issues such as
cost of equipment, availability of tags, standardization of information, and the physics of
the technology will be difficult to overcome. With of the nature of the technology, there
will not be a simple implementation procedure that can be used across industries, or even
across some companies. Each implementer will have to look at the situation individually
and make several decisions specific to their own requirements and limitations. For this
reason, there has been much testing done, at the industry level and at the university level,
looking at the effects of many different variables. While this testing has been done, there
are a few areas that have grown in importance as a result of this testing. The purpose of
this research is to cover, in depth, one of these resultant areas.

Prior research at the Michigan State University School of Packaging has led to

this specific study. Since RF ID study began at MSU in 1999, many areas have been

studied; material, environmental, and product issues have all been assessed; tags, readers,
antenna read patterns, tagged cases and case loads, pallets and pallet loads have all been
evaluated. A unit load is defined as “assembly of goods on a pallet for handling, moving,
storing and stacking as a single entity” and may include homogenous and mixed loads,

. and may be made up of a multitude of smaller items or may be very few. An evaluation
of a unit or pallet load is relevant for several reasons. On the surface it may appear to be
irrelevant because of the Wal-Mart mandate and the details of Wal-Mart’s supply chain.
Wal-Mart typically breaks a pallet apart upon arrival to one of its storage distribution
centers, and the individual cases are loaded onto a high-speed conveyor, where they are
read by RFID equipment. Reading a unit load in this case is not necessary, (although it is
read as a unit load as it is unloaded fiom the truck anyway) however this system is unique
to Wal-Mart. Many other supply chains do not implement this same read point because
they store their product as unitized loads, and therefore would be highly interested in
reading a unit load. Another reason that unit loads are relevant is that in many
distribution centers mixed pallets are constructed and shipped on an order-by-order basis.
Since each order and pallet make up differs, it becomes very important to read the entire
pallet load to know what is included therein.

Previous Master’s theses at MSU have looked at many of the aforementioned
areasl3’ 4’ 5]. Different test methods have resulted in seemingly varied outcomes, in which
many variables have been found to have an impact. Some of these variables are tag
location and orientation, product content, and read point. With respect to unit loads,
studies have been conducted in which a pallet load, with varying tag locations and

products, is walked through an RFID portal with constant antennae location. This

research will examine if the same (or better) results can be achieved by altering the
number and location of read antennae situated around a stretch wrapper.

Due to the importance of reading an entire unit load, and the poor results obtained
in portal walk through testing” 1, a different read point was selected to increase read rates.
A stretch wrapper has many applications in which RFID could be utilized. Unit loads of
mixed product cases coming out of a warehouse headed to an individual store are stretch
wrapped, homogeneous loads of product to be shipped and homogeneous loads that
orders are picked from are also stretch wrapped. The stretch wrapper used throughout the
testing consists of a vertically moving wrap carriage, and a flat turntable on which the
pallet load sits. The difficulties of reading a unit load are combated using the stretch
wrapping method because the method allows the tagged product to be in the read field for
an extended time period, and multiple views from the rotations allow for maximum
penetration of the RF signal to the interior cases. Based on this new set-up, the
hypotheses for this research are:

0 The case contents will have a similar effect on the read rates of RFID tags at the
stretch wrapper as they did at the portal.
0 The tag type will not have an effect on the read rates of RF ID tags.
0 The number of read antennae utilized will not affect the read rates of RFID tags.
0 The location of the read antennae utilized will not affect the read rates of RF ID
tags.
The significance of this research lies in the fact that while much has been made of the
product interactions, tag location effects, tag orientation effects, package system

interactions, frequency effects, antenna type effects, etc, there has not been a focus on the

location and number of antennae needed to obtain ideal reads. The results show how
significant this variable is when implementing an RF ID system. Instead of relying on
guesswork or a generic configuration, companies will now have some solid answers to

utilize when configuring RF ID into their business.

CHAPTER 2 — LITERATURE REVIEW

In today’s highly competitive world of retail and consumer packaged goods
(CPG), every company is looking to get a leg up on the competition. Increased revenue,
higher market share, and larger profits are all goals of this rapidly growing industry.
There is no specific business model for each and every company to follow in order to do
this, but there is a lot of follow-the-leader going on. In this particular industry, that
leader is Wal-Mart. “Wal-Mart Stores Inc. is the world’s largest retailer, with $285.2
billion in sales in the fiscal year ending Jan. 31, 2005.”[2] One of the ways that Wal-Mart
is gaining a competitive advantage is through increased supply chain efficiency. An
efficient supply chain can help decrease manual labor, shipping costs, and inventory
holding costs, while increasing inventory Visibility and turnover. An in-depth analysis of
this is found in Chapter 9 of “The Wal-Mart Way” by Don Soderquist.[6]

While it is easy to say that the supply chain needs to be more efficient, this is not
the easiest thing to do. One way that this issue is being addressed is through the use of
information technology (IT). Virtually every major company now has an IT department
for the sole reason that as a business grows, there is more and more information to be
gathered and an electronic means of obtaining, sorting, storing and retrieving is
necessary. Automatic Identification and Data Capture (AIDC) technologies are one of
the main ways that this is accomplished. There are many AIDC technologies in use today
that are a part of our every day life; magnetic strips and barcodes are two of the simplest
and most common. Others include biometrics, optical character scanning and RFID. The
main goals of these technologies are the elimination of human error associated with

manual data entry, and increased speed of data capturem.

One of the most rapidly growing AIDC technology is Radio Frequency
Identification (RF ID). RF ID is defined as “A wireless data collection technology that
uses electronic tags for storing data. Like bar codes, they are used to identify items.
Unlike bar codes, which must be brought close to the scanner for reading, RFID tags are
read when they are within the proximity of a transmitted radio signal.”[8] The trend in
CPG and retail is moving toward this technology as an answer to the problem of how to
increase supply chain efficiency. “ ‘Supply chain event management is the largest
application of RFID technology in industry today,’ says John Fontanella, senior VP of
supply chain services at Boston-based AberdeenGroup”.[9] A Master’s thesis from the
University of Texas at Arlington provides a further justification for using RFID
technology in mobile asset tracking. It provides the fi'amework for how to implement the
technology into the supply chain.“01 Further examples of how the technology is gaining a
foothold can be found at virtually every turn. As companies look to break away from
established business models and revolutionize the way their business operates, RFID is
coming to the forefront as illustrated in a Master’s thesis from the Massachusetts Institute
of Technology. “RFID in the supply chain represents an enabling technology that will
allow warehouse operations to break away from traditional methodologies and adopt
revolutionary techniques, such as location-relaxed storage.”[l '1 Location-relaxed storage
is made possible through the use of RFID by allowing for a less rigid organizational
system. RF ID tagged goods can be stored anywhere in the warehouse, and when a
worker needs to locate a specific item, RFID readers will notify them of the location.

RFID technology utilizes radio frequency waves to energize tags in order for them

to release their information, also in the form of radio band energy. The radio bandwidth

portion of the electromagnetic spectrum is extremely wide, ranging from 3 kilohertz
(kHz) to 300 Gigahertz (GHz), and the Federal Communications Commission (FCC) has
allocated all the bandwidths across this spectrum for different uses in the United States.
RFID has been allocated a few separate frequencies. The major fiequency allocations
for RFID in the US. are 134 kHz, 13.56 MHz, 433 MHz, 915 MHz, and 2.4 GHz.“21 134
kHz is considered low frequency (LF), 433 MHz and 13.56 MHz are considered high
frequency (HF), 915 MHz is considered ultra high fi'equency (UHF) and 2.4 GHz is
considered to be microwave. Each of these frequencies demonstrates differing
performance characteristics in regards to read range, product interaction and tag collision.
In selecting one frequency over the other, the industry has taken all these aspects into
account. There are many uses for the different fi'equencies, but as a whole, the retail and
CPG industry in the US. has moved towards UHF, or 915 MHz. “UHF (915 MHz in the
US) has been focused on for the recent retail supply-chain mandates and investment due
to its distance and cost attributes“l3 1.
As a retail industry leader, Wal-Mart has taken on the role of a driver for RFID
technology. They made this happen through issuance of a mandate for their suppliers.
The mandate has been segmented into 2 phases. To date, the first phase, which was
effected January 2005, stated that the top 100 suppliers to Wal-Mart must tag all cases
and pallets shipped to Wal-Mart distribution centers and stores with UHF RFID tags,
readable by the Wal-Mart equipment. For this mandate, items such as returnable
containers, shrink wrapped bundles, bags, and other larger sized items are considered

cases.““ The second phase states that the next 200 suppliers do the same by January of

2006. Neither phase of the mandate specifies how this is to be done, what manufacturer

to obtain equipment from, or where to place the tags. The mandate only says that Wal-
Mart must be able to read the tags upon receiving the shipment. This mandate leaves the
supplier with 2 options, comply with the mandate, or lose Wal-Mart as a customer.
Compliance at this point is a very expensive proposition, with a lot of investment
involved and very little identified return on this investment. This expense, however
mighty, pales in comparison to the potential loss of Wal-Mart as a customer, for most
companies. For this reason, many suppliers are implementing a ‘slap and ship’ policy for
shipments to Wal-Mart. Slap and ship is when a tag is manually applied to a case or
pallet, either at the manufacturing plant, or at a distribution center, on its way to Wal-
Mart. In some cases an entire truckload is unloaded, depalletized, tagged, and reloaded
as the final step before heading to Wal-Mart. This is a very bad situation for the supplier,
as there is a large increase in costs associated with tags, equipment, labor for unloading
and reloading the truck, and no increase in revenue. This is illustrated in a Chief
Information Officers (CIO) journal article. “Many suppliers are applying the tag on
pallets to be shipped to warehouses in the region where the Wal-Mart distribution center
is located. Using the product that is destined to be shipped the next day (which is already
palletized and stretch-wrapped), they are taking off the stretch-wrap, depalletizing,
applying the tag, writing the tag, repalletizing, rewrapping and sending. And even then,
suppliers are afraid the tags might not function on anival because of mechanical
problems.”“5] Simple compliance with the mandate using slap and ship will not show any
return on investment (ROI) for a supplier, and aides only Wal-Mart in its’ goal for a
competitive advantage. In order for there to be a ROI, the suppliers must implement the

technology firrther upstream in their own processing. The tire manufacturer Michelin is

one company that has already seen an ROI by tagging their product further up the

manufacturing process. “But [Michelin] already is using RFID in connection with large,
industrial tires, including those involving re-treads, where the chips are usefiil in tracking
the product's life cycle”“6].

There have been many additional mandates issued by Wal-Mart’s competitors

. such as Target, Albertsons and Best Buy, further increasing the need for suppliers to find
a way to make RF ID work for them.

Target Stores has issued a mandate similar to Wal-Mart, in that they want all their
suppliers using RF ID. “Target, the fourth largest retailer in the United States, has told its
top suppliers that they will be required to apply RF ID tags on pallets and cases sent to
"select" regional distribution facilities beginning late spring 2005. The company wants all
suppliers to tag pallets and cases by the spring of 2007”“71 Something that both Wall-
Mart and Target mandates have in common is that they want the RFID to be used as a
complement to barcodes, not as a replacement for barcodes.

Another large grocery retailer, Albertson’s, has followed suit buy issuing a
mandate of their own. “Albertson’s, the nation’s second largest food and drug retailer,
has launched its first RFID pilot and announced that it will require its top 100 suppliers to
tag pallets and cartons by April 2005.”“81 This mandate is similar to others in that it
requires pallets and cases from suppliers be tagged, but is similarly vague in how this is
to be accomplished.

A more recent mandate that follows similar guidelines is from Best Buy. “The

Minneapolis-based consumer electronics retail chain will require its major suppliers to

begin applying EPC-compliant tags to product cases and pallets by Jan. 2, 2006.”“91 A

10

couple advantages that Best Buy has are that they have a later starting date, January 2006,
and a lot of the cases shipped in are single items, so they will see more accurate item
level tracking.

In order to gain ROI, suppliers affected by these mandates are attempting to
implement RF ID technology into their internal processing. In doing so they are

[20], some of which has to do with the nature of the technology.

encountering many issues
It is not a system that will react identically in every situation; in fact it will most likely
have drastically different issues with each different implementation. “RFID is not plug-
and-play, and it may never become that. ‘RF ID is a complex recipe: it's equal parts
physics, process changes, supply chain synchronization, and software integration’ states
Deon Nel of Avatar Partners”.[2” As the article states, many of the issues that
implementers are facing have to do with physics of the technology, while others deal with
the equipment, supply chain and software.

Another big area that is garnering a lot of attention is standardization. With the
ratification of the Gen 2 standard, some of these issues are being resolved, but it is a slow
process. Essentially the problem is that the first generation of tags and readers could not
communicate with tags and readers produced by a different company. One can imagine
the headaches this would cause, as Wal-Mart refuses to publicly endorse one
manufacturer over another. The Gen 2 standard, and presumably the following
generations, is designed to eliminate the problem of each of Wal-Mart’s suppliers using
tags from a different manufacturer, and thus forcing Wal-Mart to have that many

different readers. This would be impossible for many reasons including cost and physical

space limitations. Fortunately the Gen 2 standard, which was ratified December 16,

11

2004,92] is slowly gaining a foothold. It is a slow process because of a vicious cycle
involving price and availability. Many companies are waiting to place orders until the
prices come down, and manufacturers cannot make the prices come down until they are
receiving bulk orders. Currently inlays are available for 7.9 cents and full self-adhesive
tags are available for 12.9 cents for orders of 1 million or morem‘ 24]
There has been a fair amount of research done on the physical limitations of the
technology, with the goal being more efficient and accurate use. One area covered

[3]. The location of the RF ID tag on a case of goods

extensively is the tag orientation
matters a great deal in obtaining reads of that case. This has to do with the product and
packaging materials in the case and the location of the tag itself. An example of this from
one of Wal-Mart’s top suppliers stated, as follows, “Del Monte quickly realized the
biggest problem it would face: where to place the tags. Cases of tuna are only a few
inches tall -— the height of a can — and the cans are packed closely together. Del Monte
knew it needed 4-inch by 2-inch tags to fit on the cases, and the tags needed to be
carefully placed for the best read.”[20] In this case, the metal can was causing the tag not
to be read, but a slight adjustment of the tag location eliminated this problem. In each
situation the location must be carefully assessed in order to obtain the best reads. Jeffery
Tazelaar, researching tag orientations and product interactions, did a highly
comprehensive study for a Master’s thesis at Michigan State University. He utilized a
pallet load of 48 corrugated cases, each tagged with a UHF Class 0 RF ID tag. He

compared five different tag locations and 5 different products, walking the pallet load

through a simulated warehouse RFID portal. The hypotheses for his research were:

12

o “The orientation of the RFID tag on the package will have no effect on the
readability of the transponders,
o The product contained in the package will not have an effect on the readability of
the RFID transponders,
0 Cases containing products made up of water will have lower read rates than those
cases containing waterless products,
0 All waterless products will have the same rate of readability.”
He found that the first 2 hypotheses were false, the third was true, and the fourth was also
falsem

Another Master’s thesis at Michigan State University looked at the difference
between refrigerated beef loin and frozen beef loin. Individually packaged cuts of beef
loin were tagged and subjected to either refiigerated or frozen environments. An attempt
was made to read the tagged packages individually, and stacked atop one another. Beef
loin, being a product with high water content, was found to have difficulties reading
when the beef was refiigerated. When it was fiozen however, the crystallized water
molecules did not prevent the tags, even several deep in a stack, fi‘om being read.[4] These
studies all illustrate that the product can alter the tag readability, as does the tag
orientation.

The durability of RFID tags is another issue to be considered, as most tags will be
subjected to the harsh environment of a distribution system. A Master’s thesis was
devoted to this topic at Michigan State University. A range of tests, implementing
American Society for Testing of Materials (ASTM) and International Safe Transit

Association (ISTA) standards were completed for UHF Class 0 RF ID tags. Drop, shock,

13

vibration, compression and direct impact tests were all attempted. There was virtually no
damage done to the Class 0 passive UHF tags, with the exception of when a direct impact
was applied to the microchip. If the impact were not centered over the chip, no effect
was recorded”). These results may vary with differing tag constructs, however can be
assumed true for other similar tags.

Another issue that a company must face is choosing a read point. There are a lot
of options in this area, and each presents its own set of problems. Within a warehouse
alone the read points can vary between racking systems, conveyor belts, hand held
readers, around portals/doorways, stretch wrappers, or fork trucks/pallet jacks. In fact

[25] are selling equipment to be used in all these

some RFID equipment manufacturers
locations, as there will be a market for each. A manufacturing plant has some of these
same options, as well as some unique to that specific plant. An example of a company
using two of these read points is Wal-Mart. “Wal-Mart aims to read 100 percent of all
tagged pallets coming through the dock doors at the DC and stores equipped with readers
and 100 percent of all tagged cases on conveyors within the distribution center”.[26] Just
because Wal-Mart is an industry driver in RF ID does not mean that the way they do
things works for everyone. Other companies have found other solutions to the
implementing headaches they have come across. “Kleenex-maker Kimberly-Clark Corp.
accidentally discovered a simple workaround... During normal assembly-line operations,
all pallets are fully shrink-wrapped. The machinery that does this slowly spins the pallet
around. The company discovered that an RFID scan taken during that rotation delivered

much more accurate reads, Das said, apparently because it allowed the reader to "see" the

tags from different angles and grab the best view”.[27] In talks with employees at

14

Kimberly-Clark, it has been determined that the machine discussed is actually a stretch
wrapper, and this type of solution, one that fits into a companies current process, is ideal,

however many companies may discover that they are not so luckym].

Among the research being done across industries and at universities, there appears
to be a noticeable gap in published literature. This gap is the read antennae
configuration, consisting of the location of the antennae and the number of antennae
used. Even in comprehensive books on RFID technology antennae are discussed only in

[29' 30]. A significant issue lies in

terms of how they work or what different types there are
the antennae configuration around whichever read point it is that has been chosen. This
is an issue because results may vary with different configurations and significant cost
reductions may be observed if fewer antennae can be used, by configuring them properly.

There were no found published works showing that research has been done in this area,

or to indicate that anyone has done anything beyond guesswork.

There are two aspects to antennae configuration studied in this research. The first
is the number of read antennae utilized, and the second is the location of those antennae.
If one can obtain equivalent reads from using less antennae and locating them properly, a
significant amount of money might be saved. For example, in a warehousing or
distribution center situation, where potentially hundreds of dock doors are to be equipped
with read RFID antennae, this savings is especially significant considering the average
antenna cost is around $200 each[3 I]. While there has been plenty of research that shows
that product characteristics can alter a tag’s readability, none has been done to see if this

product will alter an antenna’s performance, in different configurations. Another area

15

that has had some research performed is the difference that a tag from a different

manufacturer makes. Individual companies are faced with an increasing choice as more
and more tag manufacturers enter the market, and while they have undoubtedly assessed
different tags based on price and performance, there is no published research on the tags

interactions with differing antennae configurations.

Antennae selection for this research was based on a number of factors.
Availability, and cost were assessed, but the more important aspect was functionality.
Compatibility with the system used was a requirement and the polarization of the
antennae was also evaluated. Antennae are polarized in one of two ways; linear or
circular. This terminology refers to the way in which the radio waves are propagated. A
linear antenna propagates the energy in a concentrated planar fashion. A circular antenna
propagates the energy in a rotating circular pattern. Linear antennae are best utilized
when reading tags from a long distance is required and the tag orientation is controlled in
such a way that it aligns with the planar propagation of the radio waves. Circular
antennae are best utilized for varying tag orientations and shorter read distances. This
phenomenon can be observed in lab exercises for PKG 491: RFID and Packaging, a
course taught at the Michigan State University School of Packagingm]. For this research,
a short read distance was required and the tag orientations change as the pallet load spins,

thus circular antennae were used throughout the testing.

16

CHAPTER 3 - METHODOLOGY

Equipment

The reader utilized throughout the testing was the Sensormatic Agile2 Reader,
powered by ThingMagic (Boca Raton, F 1). It was a Gen 2 reader, meaning it was able to
read and send data from Electronic Product Code (EPC) Class 0, Class 0+, Class 1 and
Gen 2 tags, to the computer for analysis using a crossover Ethernet cable. Gen 2 readers
are different from first generation readers in that they are able to read tags from different
manufacturers and of different classes at once. This capability made this testing possible.
The Agile2 was an eight port reader, capable connecting to eight ultra high fi'equency
(UHF) antennae, however for this research no more than four antennae were used at any

given time.

The antennae used were Sensorle OmniWave circular polarized antennae. The
circular polarization antennae were used because they are less sensitive to tag orientation,
and the read distance required for this testing was not large enough to require linear
antennae. Differing combinations of 1-4 of these antennae were connected to the reader
using 25’ coaxial cables. The antennae were mounted in various locations around a
wooden frame using small metal brackets and screws. Wood was the material chosen
because it is relatively radiolucent compared to metal or polymer based structures.
Constructed around a stretch wrapper (Synergy 3, Highlight Inc. Grand Rapids, Mi), the
frame consisted of 4 upright posts (84”), 2 top crossbars (92”), and 2 connecting bars

(36”)(Figure 1). For most configurations an antenna was screwed into the center of the

17

36” span. The reader rested upon one of the top crossbars such that all antennae cords

could reach it.

[—— 92” ——|
k/ Reader

l

84”

36” ‘J

Figure 1
Wooden frame that antennae were mounted on. The frame was
set up around the stretch wrapper turntable

 

 

 

 

Procedure

Forty-eight center special slotted containers (CSSCs), were assembled from
standard 42-26-42 C-flute corrugated board and sealed at the glue flap using
Surebonder® All Purpose StikTM glue sticks, to create inside dimensions of 13” x 12” x
10”. These case flaps were sealed with 2” clear, pressure-adhesive, packing tape and
stacked in a 4 x 3 pallet pattern, 4 tiers high on a standard Grocery Manufacturers
Association (GMA) 48” x 40” wooden pallet. This same pallet and pallet pattern were
used throughout the testing with the only variables being product, tag type and antennae

location.

18

The cases were securely tagged in the upper left hand comer with Alien squiggle
tags (Alien Technologies, Morgan Hill, Ca), using scotch tape (Figure 2). These tags
measure 4” X 1/2” and are EPC Class 1, meaning they have the capability of being
programmed on site. The tags have 16 digits of information that can be programmed by
the user. For this test the Alien tags were named 5555 5555 5555 0001, 5555 5555 5555
0002,. . .,5555 5555 5555 0048. Henceforth they will be referred to simply as tag 1, tag

2,...,tag 48.

 

Figure 2
Alien Squiggle tags on cases

Cases with tag numbers l-12 were positioned on the bottom tier of the pallet, 13-
24 on the second tier, 25-36 on the third tier and 37-48 on the fourth or top tier. The

placement of the tagged cases is shown in Figure 3.

 

Figure 3
Tag position on pallet,

by tier

20

The pallet load was placed on the stretch wrapper turntable (Highlight Industries),
with the wooden portal constructed around it (Figure 4), and set to a configuration where
the machine would wrap the pallet with one full band around the bottom, and 2 full bands
around the top. For these tests, however, the wrap was not attached to the pallet as the
stretch wrap being used was found to have no impact on the results and thus would have
been a waste of resources. Each trial consisted of 8.5 turns on the stretch wrapper lasting
50 seconds total, and the reader was manually started and stopped to synchronize with the
start and stop of the stretch wrap machine. After each trial the tags that read were
manually recorded into a Microsofi Excel document where a “1” indicated a read and a
“0” indicated a no-read.

For each of the antennae configurations/tag types/product combinations, 25 trials
were performed. Given nine antenna configurations and two tag types, the total number
of trials per product was 450. Three different products were used, empty cases, to serve
as a baseline, rice jars, and bottles of water. Two tag types were used, the Alien squiggle

and the Symbol double di-pole.

 

Figure 4
Pallet load on stretch wrapper

21

After 25 trials were completed, the antennae configuration was changed and the
process was repeated. There were nine different antennae configurations used per each

tag type and product. They are listed in Table 1.

22

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

32> 32> .l
0E9”. 582$ 50.. 5.5a E9“. m::2:< 2.5 3:25. W. “I‘m &
k a a
J A U
I _ e .1... w
w M m
8235 32> 02m com “Bowed .N 035.:
. :8: Eat .8 :oo: So: .8
5:3 9on :8: : 8% 5:3 - 93:25 25 .. 5::25
.8 .505. Ea: .5 £35 €35.65 Bo: 5:3 305 :oo: :8: cam 5:3
.8: =5: .8 a E822 3 .855 22 co :8: . ”Rum .8 5o: .8 .8 .595. .9. 2m .5 26am .8 . 28.2%.:
- 3:95 go... 3.62. we E0: 3: .8 t Rm on :0 E328 £35 .5de -
- 3:95 no... 03:25 no:
I I __ __ _ _ 85.555
32> 2.5
I I \ I I H I I
02:85
_. F N N N N n v we 5:832
\ cow—3525.0
2.6 no... no: no... 25 :8. cam on :80 8:50 5:30..
53:52
a a N o n v n F 25:33:00

 

 

 

 

5335252 553580 ._ 035.:

«.3
Ala

After the nine different antennae configurations were tested with 25 trials each,
the Alien squiggle tags were replaced with Symbol double di-pole tags (Figure 5). These
tags measure 6” X 4”, and are EPC Class 0, (prewritten). For this reason the tags were
individually read before application and the last three digits (both numbers and letters) of
the data already in place were recorded and sorted alphanumerically. The 3-digit code
was assigned a number (1-48) and applied to the cases in the same order as in the first
test. When all 450 trials were completed, the tags were removed and the entire process

was repeated using a new product.

 

Figure 5
Symbol double di-pole tags on cases

The first set of trials was completed with empty cases, which have proven to be a
good baseline for research as the corrugated alone has little to no effect on the readability
of tags. The second product tested was bottles of water. Water was used because it is
known that water has a significant effect on the readability of RFID tags. The bottles
used were made of polyethylene (PET), contained 32 fl. oz. of tap water and sealed with
twist caps made of polypropylene (PP). Each case held 12 of the 9.75” tall and 3.5”
diameter water bottles. The bottles left no headspace in the cases which were again

sealed with 2” clear pressure-adhesive packing tape (Figure 6).

24

 

Figure 6
Water bottles in cases

The third and final product for these tests consisted of rectangular PET jars of rice
pilaf. This product was selected because it was a dry, dense product, to contrast with the
water bottles, and thus was expected to have an impact on readability” 1. Each case held 9
jars in a 3 x 3 pattern (6” tall, base of 3.75” x 3.5”, 3.5” cap diameter), leaving a 4”

headspace afier being cased (Figure 7).

 

Figure 7
Rice jars in cases

Data from each test was collected and analyzed for statistically significant differences in
certain areas. The comparisons of read rates were made were for case content, tag type,
and antennae number and location(s). While different tags read differing amounts of
times, if a tag was read at least once for a trial, it was considered a read and recorded

accordingly. Samples of the data collection sheets are displayed in tables 3 and 4.

25

 

 

 

Table 3. Data Collection Sheet — Example 1

2 antennae: 2 Sid-21:3")

       
   
  
  

Alien Tags

Water Bottle Filled Casa Trial #

Tag Number 12 “:6 14 15 B 1? 113 "El 23 2’1 2 E 251- 25 '16 Read $6 No Read
1, ”" "’ ”’ ... a 100% 0%
2 100% 0%
3 50% 44%
4 100% 0%
5 0% 103%
5 0% 100%
7, 100% 0%
B . 0% 103%
g , 0% 100%

”D . 100% 0%
11 0% 100%
12 , 0% 100%
13 100% 0%
14 100% 0%
1‘5 100% 0%
13 100% 0%
17 0% 103%
’B 20% 83%
10 100% 0%
23 0% 103%
21 0% 1CD%
22 100% 0%
23 ~ 0% 103%
a 100% 0%
25 100% 0%
23 100% 0%
'2? 100% 0%
E, 100% 0%
2; 0% 111336
33 0% 100%
31 100% 0%
32 0% 103%
33 , 0% 100% 4
234 100% 0% ‘ _
35 0% 103%
33 0% 103%
3:..- . 1 ‘16- 0%-
E 100% 0%
33 . 100% 0%
43 100% 0%
41 0% 103%
42 0%- 103%
48 100% 0%
44 0% 103%
{:5 28% T236-
55 ,. 100% 0%
4r 4% 00%
£3 0% 103%

715.
AU

 

2

Example

lection Sheet —

l

e 4. Data C0
Trial #

1

Tab

a rrtenn are: 2 s ideniS’B'")

Z

 

M%%%%%%%%%%%%%%H%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
00080000000000000000000000000340400890000000$80m0
0
%
%mw%mMMM%%%M%%M%%%MM%%%%%M%%%MMMM%%%%%%%%%%%%%m%$
mamam0gnawmmammqmmmammmmmmammmmqwqummmmmmmmmmmzmm
€11 1111111111111111111111111 4.. .11 1.111111 Al 1
%

 

3'5

$4?

24

43

$40474?

45$474745434343454347

$45

4540474544

4.14.1.1

iawmmaa%_Aaamamaammaameaqa

    

 

,12 3 4 5 ,0 .7 3 0,0 .12. 3.4.53 ,7 .
fil- Al Al. Al.

19
10
:53,
21
Z?
24

Tag H umber

 

CHAPTER 4 — RESULTS

Case Content - Total Reads .
The results listed below were collected for the 3 different case contents. The total

number of reads consists of 9 antennae configurations, 2 tag types, 48 tagged cases and
25 trials performed for each antenna configuration and tag type. This gives a total of

21,600 possible reads for each case content.

1. Empty cases, for all antennae configurations and both tag types, had a total of
21,214 reads out of a possible 21,600 for a read rate of 98.21%.
2. Rice filled cases, for all antennae configurations and both tag types, had a
total of 18,342 reads out of a possible 21,600 for a read rate of 84.92%.
3. Water bottle filled cases, for all antennae configurations and both tag types,
had a total of 12,611 reads out of a possible 21,600 for a read rate of 58.39%.
These results show that tag readability is drastically dependent on case content,
regardless of antennae configuration and tag type. The information above is illustrated in

Figures 8, 9 and 10.

28

 

 

 

Empty Cases - All Configurations and Tag Types

98.21%

 

Reads
I] No Reads

 

 

 

 

1 .79%

 

Figure 8. Percentage of Total Reads on Empty Cases

 

 

Rice Filled Cases - All Configurations and Tag Typos

Reads
El No Reads

 

15.08%

 

Figure 9. Percentage of Total Reads on Rice Filled Cases

 

 

Water Filled Cases - All Configuratlons and Tag Types

   

 

Figure 10. Percentage of Total Reads on Water Filled Cases

29

 

 

 

Tag Type - Total Reads
The results listed below were collected for the 2 different tag types. The total

number of reads consists of 9 antenna configurations, 3 case contents, 48 tagged cases
and 25 trials performed for each antennae configuration and case content. This gives a
total of 32,400 possible reads for each tag type.
1. Alien Squiggle tags, for all antennae configurations and case contents, had a
total of 25,786 out of a possible 32,400 for a read rate of 79.59%.
2. Symbol Double Di-pole tags, for all antennae configurations and case
contents, had a total of 26,381 out of a possible 32,400 for a read rate of
81.42%.
These results show that the readability of the different tag types is not statistically
different when altering antennae configurations and case contents. The information

above is illustrated in Figures 11 and 12.

.30

 

 

Alien Squiggle Tags - All Configurations and Case Contents

79.59%

 

Reads
E] No Reads

 

 

 

    

20.41%

 

 

Figure l 1. Percentage of Total Reads on Alien Squiggle Tagged Cases

 

 

Symbol Double Di-Pole Tags - All Configurations and Case
Contents

81.42%

 

121 Reads
[1 No Reads

 

 

 

    

18.58%

 

 

Figure 12. Percentage of Total Reads on Symbol Double Di-Pole Tagged Cases

31

Antennae Configuration — Total Reads
The results listed below were collected for the 9 different antennae configurations.

The total number of reads consists of 3 different case contents, 2 different tag types, 48
tagged cases and 25 trials performed for each case content and tag type. This gives a

total of 7,200 possible reads for each antennae configuration.

1. Configuration 1, for all case contents and both tag types, had a total of 6,556
reads out of a possible 7,200 for a read rate of 91 .06%.

2. Configuration 2, for all case contents and both tag types, had a total of 6,623
reads out of a possible 7,200 for a read rate of 91 .99%.

3. Configuration 3, for all case contents and both tag types, had a total of 6,680
reads out of a possible 7,200 for a read rate of 92.78%.

4. Configuration 4, for all case contents and both tag types, had a total of 6,064
reads out of a possible 7,200 for a read rate of 84.22%.

5. Configuration 5, for all case contents and both tag types, had a total of 5,41 6
reads out of a possible 7,200 for a read rate of 75.22%.

6. Configuration 6, for all case contents and both tag types, had a total of 6,028
reads out of a possible 7,200 for a read rate of 83.72%.

7. Configuration 7, for all case contents and both tag types, had a total of 4,680
reads out of a possible 7,200 for a read rate of 65.00%.

8. Configuration 8, for all case contents and both tag types, had a total of 4,681
reads out of a possible 7,200 for a read rate of 65.01%.

9. Configuration 9, for all case contents and both tag types, had a total of 5,439

reads out of a possible 7,200 for a read rate of 75.54%.

32

These results show that readability is changed when the antennae configuration is
changed, across all case contents and tag types. The information above is illustrated in

Figures 13-21.

33

 

 

D

 

Configuration 1- All Case Contents and Tag Types

91 .06%

 

 

 

Reads
[3 No Reads
8.94%
Figure 13. Percentage of Reads for Configuration 1
Configuration 2 - All Case Contents and Tag Types
Reads
[3 No Reads

 

8.01%

 

Figure 14. Percentage of Reads for Configuration. 2

 

 

D

 

Configuration 3 - All Case Contents and Tag Types

92.78%

E Reads
1:] No Reads

 

Figure 15. Percentage of Reads for Configuration 3

34

 

 

 

 

 

Configuration 4 - All Case Contents and Tag Types

 

 

 

 

 

 

Reads
D No Reads

 

 

 

 

 

Figure 16. Percentage of Reads for Configuration 4

 

 

Configuration 5 - All Case Contents and Tag Types

 

 

 

 

 

75.22% . ,

 

 

 

Reads
El No Reads

 

 

 

 

 

‘ 24.78%

 

Figure 17. Percentage of Reads for Configuration 5

 

 

Configuration 6 - All Case Contents and Tag Types

 

 

 

 

 

83.72%

   

 

 

Reads
El No Reads

 

 

 

 

 

Figure 18. Percentage of Reads for Configuration 6

35

 

 

 

I \ Configuration 7- All Case Contents and Tag Types

   
 

65.00%

 

35.00%

 

Figure 19. Percentage of Reads for Configuration 7

 

 

 

— Configuratlon 8 - All Case Contents and Tag Types

 

   

 

 

Figure 20. Percentage of Reads for Configuration 8

 

 

Configuration 9 - All Case Contents and Tag Types

E

   

 

Figure 21. Percentage of Reads for Configuration 9

36

 

 

There are two aspects of antennae configuration that are responsible for the
differences seen above, the location of the antennae, and the number of antennae used.
The location portion has been segmented into 3 groups for analysis; Top, Side, and
Combo. The Top segment consists solely of configurations in which the antennae are
above the pallet load reading down, the Side segment consists solely of configurations in
which the antennae are on the side of the wooden portal and the Combo segment consists
of those configurations in which antennae are both above the pallet load and on the side.
There are 3 configurations in the Top group, 4 configurations in the Side group and 2

configurations in the Combo group.

Antennae Configuration — Location of Antennae

The results listed below have been grouped into segments based on the location of
the antennae. All three segments include 3 case contents, 2 tag types, 48 tagged cases
and 25 trials. The Top segment consists of 3 configurations for a total of 21 ,600 possible
reads, the Side segment consists of 4 configurations for a total of 28,800 possible reads,

and the Combo segment consists of 2 configurations for a total of 14,400 possible reads.

1. Top antennae, for all case contents and both tag types, had a total of
14,777 out of a possible 21,600 reads for a read rate of 68.41 %.
2. Side antennae, for all case contents and both tag types, had a total of

24,154 out of a possible 28,800 reads for a read rate of 83.87%.

37

3. Combo antennae, for all case contents and both tag types, had a total of
13,236 out of a possible 14,400 reads for a read rate of 91 .92%.
These results show that the readability was significantly affected by the location

of the antennae. This information is illustrated in figures 22, 23, and 24.

38

 

 

Top Locations - All Case Contents and Tag Types

68.41%

    

31.59%

 

Figure 22. Percentage of Reads for Top Antennae Configurations

 

 

Side Locations - All Case Contents and Tag Types

83.87%

      

16.13%

 

Figure 23. Percentage of Reads for Side Antennae Configurations

 

Combo Locatlons - All Case Contents and Tag Types

121 Reads
El No Reads

 

8.08%

 

 

Figure 24. Percentage of Reads for Combo Antennae Configurations

39

 

 

Antennae Configuration - Number of Antennae
The results have also been segmented into number of antennae groups, 4

antennae, 3 antennae, 2 antennae and 1 antenna. The results listed below have been
grouped into segments based on these groups. All four segments include 3 case contents,
2 tag types, 48 tagged cases and 25 trials. There were 2 configurations utilizing 4
antennae, for a total of 14,400 possible reads, 1 configuration utilizing 3 antennae, for a
total of 7,200 possible reads, 4 configurations utilizing 2 antennae, for a total of 28,800
possible reads, and 2 configurations utilizing 1 antenna, for a total of 14,400 possible
reads.
1. 4 antennae, for all case contents and both tag types, had a total of 13,179 out
of a possible 14,400 reads for a read rate of 91 .52%.
2. 3 antennae, for all case contents and both tag types, had a total of 6,680 out of
a possible 7,200 reads for a read rate of 92.78%.
3. 2 antennae, for all case contents and both tag types, had a total of 22,1 88 out
of a possible 28,800 reads for a read rate of 77.04%.
4. 1 antenna, for all case contents and both tag types, had a total of 10,120 out of
a possible 14,400 reads for a read rate of 70.28%.
These results show that there is not a significant difference between 3 or 4 antennae, and
that there is also not a significant difference between 1 or 2 antennae. However there is a
significant difference when comparing 3\4 and 1\2 antennae.

This information is illustrated in Figures 25, 26, 27, and 28.

40

 

 

4 Antennae - All Case Contents and Tag Types

 

 

Reads
[3 No Reads

 

 

 

 

 

Figure 25. Percentage of Reads for Configurations with 4 Antennae

 

 

3 Antennae - All Case Contents and Tag Types

92.78%

 

 

 

   

 

 

Reads
El No Reads

 

 

 

 

 

Figure 26. Percentage of Reads for Configurations with 3 Antennae

41

 

 

 

2 Antennae - All Case Contents and Tag Types

 

77.04%

 

 

Reads
1:] No Reads

 

 

 

 

 

Figure 27. Percentage of Reads for Configurations with 2 Antennae

 

 

1 Antenna - All Case Contents and Tag Types

 

 

15:1 Reads
1:] No Reads

 

 

 

  

29.72%

 

Figure 28. Percentage of Reads for Configurations with 1 Antenna

42

 

There were variations within the configurations that utilized two antennae. The
results ranged from 65.00% reads to 84.22% reads and because the number of antennae
was constant it can be confirmed that the location of the antennae was responsible for this

difference. The variation in results can be seen in the table below.

Table 5. Variation in configurations with 2 antennae

 

I \

 

 

 

 

Diagram

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Configuration 4 5 6 7

 

 

 

 

43

Table 6. Results for Empty Cases

 

Table 7. Results for Rice Filled Cases

 

Table 8. Results for Water Filled Cases

 

CHAPTER 5 — CONCLUSIONS

The results of the testing described in previous chapter show that product (case
content) and antennae configuration have a significant impact on the readability of RF ID
tags in a pallet load being read on a stretch wrapper. The results also show that the type
of tag, in this situation, has little effect at all. The hypotheses outlined at the start of this
research were evaluated and determined to be true or false. The outcomes of this
evaluation are listed below.

Hypothesis 1. The case contents will have a similar effect on the readability of RF ID
tags at the stretch wrapper as they did at the portal.

This hypothesis was found to be true. Empty cases had no effect on the
readability of tags, rice filled jars decreased the readability, and water bottled filled cases

significantly decreased the readability.

Table 9. Results for Case Content

 

Case Content Empty Rice Water
Total Read % 98.21% 84.92% 58.39%

 

 

 

 

 

 

 

Hypothesis 2. The tag type will not have an effect on the readability of RF ID tags.

This hypothesis was found to be true. There was no significant difference in the
readability of Alien squiggle tags versus Symbol Double Di-Pole tags across all case
contents and antennae configurations.

Table 10. Results for Tag Type

 

TagType Alien Symbol
otaIRead% 79.59% 81.42%

 

 

 

 

 

 

45

Hypothesis 3. The number of antenna utilized will not affect the readability of RF ID
tags.

This hypothesis was found to be false. While there was not a significant
difference in comparing 3 and 4 antennae, or 1 and 2 antennae, the groups were
significantly different, showing that there is a dramatic increase in reads when adding the

3rd or 4th antenna.

Table 11. Results for Number of Antennae

 

Number of Antennae

 

1 2 3 4
Antenna Antennae Antennae Antennae

 

Empty 93.15% 99.41% 100.00% 100.00%
Rice 71 .47% 81 .20% 98.54% 99.00%
Water 46.23% 50.52% 79.79% 75.57%

Frotai 70.28% 77.04% ’ 92.78% ( ”91.52% “.1

 

 

 

 

 

 

 

 

 

 

Hypothesis 4. The location of the antenna utilized will not affect the readability of RFID
tags.

This hypothesis was found to be false. Antennae that were above the pallet load,
facing down had a lower read rate than antennae that were on the side of the pallet load,
and a significantly lower read rate than configurations that had a combination of both top

and side antennae.

Table 12. Results for Location of Antennae

 

Location of Antennae
Top Side Combo
Empty 94.64% 100.00% 100.00%

Rice 70.30% 89.04% 98.63%
Water 40.31 % 62.58% 77.1 3%

Tom. 68‘4170 . 83.870/0 . ‘ 91‘920/0 it

 

 

 

 

 

 

 

 

 

 

 

46

The effect of package content was quite clear for this situation. Empty cases
served as the baseline, and yielded by far the best reads with only 2 of the 9
configurations (numbers 7 and 8) had any misreads at all. Rice filled cases were the next
best, and while none of the configurations yielded 100% reads, 5 of the 9 (numbers
1,2,3,4, and 6) were over 90%. Water filled cases were by far the worst, with none of the
configurations over 80%, and some as low as 35%. This progression from empty cases to
dry product to water filled cases shows the same relationship to readability as was found
in the Master’s Thesis work by Jeff Tazelaarl3 1. A major difference however was found
in the water results, due to the test method. In Tazelaar’s research the pallet load was
walked through an RFID equipped portal, and in this research the pallet load was spun on
a stretch wrapper. This new test method yielded much higher results for the water cases.
The variable for the Tazelaar testing was tag orientation, and the highest read rate that
was obtained was 67% and the rest of the results were much lower, with 2 virtually
reading 0%. Contrasted with the stretch wrapping method, where the 6 of the antennae
configurations yielded read rates of over 50% and 35% was the lowest result observed.
The reason for this is twofold. First, the tags spend an extended period of time in the read
field as the stretch wrapper runs its program, allowing more tags to energize and release
their information. Secondly the rotation of the load allows for multiple tag orientations,
which also aid in the readability.

As previously mentioned, the tag type for this research did not have a significant
effect on readability. While there were some differences between the tag manufacturers
for certain case content/antennae configuration combinations, the results were proven to

be statistically insignificant.

47

The antennae configuration was found to have an impact on readability. Both
components of the configuration, number and location of antennae, played a role in this.
As shown in tables 11 and 12, the configurations which only had antennae above the
pallet load, reading down, performed the worst, configurations that only had antennae on
the side performed next best, and configurations that utilized antennae in both of these
locations performed the best. As far as the number of antennae is concerned, some
interesting results were observed. There was no significant difference in switching from
1 to 2 antennae, and there was also no significant difference in switching from 3 to 4
antennae. There was however a large statistical difference between the 2 groups (1 and 2
vs. 3 and 4). This is important to know because if you are getting the reads necessary
with 2 antennae, it may be possible to drop down to 1. Also if you are not getting the
required reads with 2 antennae, it can be observed that adding a third will significantly
increase your readability.

Now the question becomes at what point is the percentage of reads that are being
obtained acceptable for use in industry. As shown by these tests, the results are nowhere
near 100%, even with steps being taken to optimize the situation. The preliminary
thought on RFID was that in order for it to be used on a widespread basis, the results
needed to be flawless. The nature of the technology prevents this from being a
possibility, so implementers are altering their stance. Now it appears that 100% reads are
not necessary, as long as each item is found to be present. This means that the use of a
backup plan is necessary, or at least some redundancy in the system. For instance, if a
tag is read at one point along the distribution chain, and then again further along, it is not

necessary for the tag to be read at all points in between. This is an example of a

48

redundant system. An example of a backup system would be in the use of RF ID as a toll
collection on a highway. As a car passes the toll booth, a tag is read and the fee is
deducted from a predetermined account number. If the car is traveling too fast, or it is
raining out or for some other reason the tag does fails to read, a snapshot of the car’s
license plate is taken and a bill is mailed to the person the car is registered to. In this case
100% reads are not required, as long as 100% of the fees are still collected in one manner

or another.

49

10.

ll.

12.

13.

REFERENCES

Navas Deb, World 's biggest supply network, 2005, http://www.scs-
mag.com/index.phpjoption=com_content&task=view&id=940&Itemid=87,
11/16/2005

Wal-Mart, Wal—Mart Facts Sheet, http://www.wa1martfacts.com/newsdesk/wal-
mart-fact-sheets.aspx#a125. 10/29/05

Tazelaar Jeffery, Master's Thesis, The eflect of tag orientation and package
content on the readability of radio frequency identification (RFID) transponders, p.
2004, Michigan State University, School of Packaging

'i—~_.
.I

Onderko John, Master's Thesis, Radio fi'equency identification transponder
performance on refrigerated and frozen beef loin muscle packages, 2004,
Michigan State University, School of Packaging

Jonson Michael, Master's Thesis, Developing T est Methods for Failure Modes of
Class 0 RFID Tags, 2005, Michigan State University, School of Packaging

Soderquist Don, The Wal-Mart Way. 2005, Nashville, Tenn: Nelson Business.
Personal Conversation with Clarke Robb, 10/5/05
Answers.com, RFID, http://www.answers.cothQpic/rfid, 10/29/05

Provia, 'RFID will find first home in CPG and Pharmaceuticals ', 2005,
http://www.provia.com/News/Headlines/ZOOS1001 .htm. 10/29/05

De Pradip, Master's Thesis, An RFID based ubiquitous fi'amework for mobile
object tracking, 2004, University of Texas, Computer Science

Ho Stephen, Master's Thesis, Intentional fragmentation for material storage,
2004, Massachusetts Institute of Technology,

National Telecommunications and Information Administration, The Radio
Spectrum, 2003, http://www.rm.docgov/osmhome/allochrtpdf

Intel, RFID: The Real and Integrated Story. Intel Technology Journal, 8/3/2005,

http://www.intel.com/technologv/itj/ZDO5/volume09issue03/art09_rfid/p03_ecosv
stem.htm

50

 

14.

15.

l6.

l7.

l8.

19.

20.

21.

22.

23.

24.

25.

26.

27.

Matrics, Understanding the Wal-Mart Initiative, 2004,
http://rfidbluestarinc.corn/resources/Understaflling_ThetWal-
fiartlnitiativefipdf. 1 1/17/2005

Wailgum Thomas, Tag, You're Late. CIO,
http://www.cio.com/archive/1 1 1504/rfid.html

Murphy Chris, Real World RFID: Wal-Mart, Gillette, and others. Information
Week, 5/25/2005,
http:// informationwrgkcom/ storv/ showArticle.ihtml?2_1rticleID=1 63 70095 5 &_100

pback=1

RFID Journal, Target Issues RFID Mandate. RFID Journal, 2/20/2004, F ~«
hm)://www.rfidiournal;com/article/articleview/ 802/ 1 / 1/ 1

1‘7——__” ' ‘

Collins Jonathon, Albertsons Announces Mandate. RFID Journal, 3/5/2004,
http:wawrfidioumal.com/article/articleview/819/1/20/

Roberti Mark, Best Buy to Deploy RFID. RFID Journal, 8/31/2004,
hfipJ/wwwrfidiournal.com/article/articleview/ 1 104/1/1/

Chapin Khristen, RFID Compliance: Trial and Error, 2005,
http://www.integratedsolutionsmag.com/Articles/2005 08/050803 .htrn, 10/3 1/ 05

Brooke Marlo, RFID Strategy - RFID in the Supply Chain. Industry Week,
8/23/05, http://www.industrvweek.com/ReadArticle.aspx?ArticleID=10657

Roberti Mark, EPCglobal Ratifies Gen 2 Standard. RFID Journal, 12/16/2004,
http://wwwrfidioumal.com/article/articleview/ 1293/ 1/ 1/

RFID Update, Avery's 7. 9-Cent Inlay Excites Industry. RFID Update, 9/22/2005,
http://rfidupdate.com/articles/index.php?id=960

RF ID Journal, Alien Drops Tag Price to 12.9 Cents. RFID Journal, 9/15/2005,
http://www.rfich:ournIall.coni/article/articleview/1 870/1/1/

Barcoding Incorporated,
http://www.barcoding.com/RFID/rfidireader_options_bv_application.shtrnl,
1 1/1/05 .

Roberti Mark, Wal-Mart begins RFID Rollout. RFID Journal, 4/3 0/2004,
http://www.rfidiournal.com/article/articleview/926/ 1/ 1/

Schuman Evan, Companies find workarounds for RFID roadblocks. Extreme
Tech, 4/28/2005, http://www.extremetech.com/article2/0,1 558,1790504,00.asp

 

51

28.

29.

30.

31.

32.

Personal Conversation with Gary Madsen, 10/5/2005

Finkerzeller, RFID Handbook. 2003, West Sussex: John Wiley and Sons.
Sweeney Patrick, RFID for Dummies. 2005, Indianapolis: Wiley Publishing.
Hutton Communications, 2005,
https://www.huttononline.com/HuttonOnline/advancedsearch.gpoh=2,

11/17/2005

Crawforth Thomas, Lab Exercise # 8. 2005, Michigan State University School of
Packaging.

52