Radio frequency identification: an introduction for library professionals.
Readers are welcome to make physical or electronic copies of this paper or quote or reference it. If you do so you must appropriately attribute the contents and authorship to Alan Butters, Sybis. While every precaution has been taken in the preparation of the paper, neither Sybis nor Alan Butters assume any responsibility for errors, omissions, or damages resulting from the use of the information and opinions contained within it. Products or corporate names may be registered trademarks of their companies and are used only for the purposes of explanation or illustration, with no intent to infringe. [c] 2006 Alan Butters, Sybis
Radio Frequency Identification (RFID) is a set of technologies that enables tracking and monitoring activities to be carried out using invisible radio waves over distances that range from less than a centimetre to many hundreds of metres. Most often it is employed to track items such as pallets or cartons within a supply chain or warehouse. RFID is also used to identify animals, hospital patients, shipping containers, laundry garments, airline baggage etc.
In many of these cases, barcodes have been used previously to identify the items concerned. One limitation of barcode technology is that the barcode scanner and the barcode must have a line of sight relationship. The barcode scanner must be able to 'see' the barcode to operate correctly. This requires items to be presented in a particular orientation to the scanner and every barcode must be visible on the outside of the item. If we imagine a pallet stacked with cartons, the cartons that are located at the centre of the stack would have their barcodes obscured by the cartons surrounding them. In this case, the cartons would have to be removed for the scanner to read each barcode.
In the RFID world, the barcode is replaced with an RFID tag and the barcode scanner is swapped for an RFID reader. The tags are essentially smart labels in most cases and have a chip and antenna as their main components. In the pallet stacked with cartons example, a suitably placed antenna would be able to read the RFID tags on every carton in the stack almost instantly and without requiring them to be visible. The radio waves generated during the reading process are able to penetrate many materials and so can be employed where tags are not visible to the eye.
RFID in libraries
RFID's property of non line of sight operation can be very useful within a library. If we replace the barcode used to identify library material with an RFID tag, we are immediately able to process multiple items simultaneously and the tag can be located inside the cover in the case of printed material. No longer does every item require individual handling. The benefits can be realised in improvements to productivity, service, materials handling and collection management.
RFID tags for libraries come in a range of sizes with the most common measuring approximately 50mm x 50mm and having a paper backing on one side and an adhesive layer on the other side. Special RFID tags are available from some vendors for cds, dvds and videocassettes. The tags can usually be overprinted with the name and logo of the owning institution if required. The RFID tag may be used not only as a barcode replacement but may also contain additional information such as the library and branch identifier, material type and even the title. In many cases the RFID tag is also used for item security and so takes the place of the electromagnetic security strip as well.
RFID in the world outside libraries The history of RFID
Many histories of RFID trace the roots of the technology to a seminal paper in 1948 by Harry Stockman entitled Communications by means of reflected power. (1) Much pioneering work was done in the late 1930s and early 1940s which resulted in the technology we now know as radar. While it was obviously useful to be able to detect an object at distance using a primitive radar system, it was even more useful to be able to determine the identity of the object. This requirement, given added impetus by World War Two, led to the development of Identification--Friend or Foe (IFF) systems. These two technologies, radar and IFF, conceptually form the beginnings of what we know today as RFID.
The point for library professionals is that the idea or principles of RFID have been around for some time. The application of the modern technological expression of these ideas is relatively recent, particularly in the library community.
RFID systems today
One of the most confusing things about RFID is that it is not one single technology. This fact is often the cause of misconceptions and misinformation carried by the popular press. The reality is that the way RFID readers communicate with RFID tags varies from application to application, as does the frequency at which they communicate. The good news is that within the library world almost all systems operate at the same frequency and use the same or very similar principles.
Different types of RFID systems in use today Broadly, RFID systems employed across current applications differ in three ways
* the way the chip on the RFID tag is powered
* the method of communication between the tag and the RFID reader
* the frequency at which the two communicate
Why do these differences exist? Because differences along these three dimensions provide a range of performance characteristics that enable RFID systems to be employed across a broad range of applications--each with its own requirements. As the pictures illustrate, RFID tags come in many different shapes and sizes. The pictures show examples of three types of tags; passive, semi passive, and active. The most meaningful thing to be said about the differences between these tags is the way that they are powered.
The passive tags have no battery to power them (2) and so they draw their power from the electromagnetic field generated by the RFID reader--quite a neat trick. It will be obvious from the pictures that, for the library application at least, a tag without a bulky battery will be essential if the tag is to be placed inside a book.
The other two tags shown in the pictures both have batteries) The semi passive tag uses its battery to power its internal chip whenever it is interrogated by an RFID reader. These sorts of tags are used in vehicle tolling systems such as the Citylink system in Melbourne Australia. (4) The active tag uses a larger battery to transmit its signal over a greater distance and so permits monitoring of objects hundreds of metres away. Therefore the first thing we can say is that the physical size of the tag may be important in some applications, such as libraries, whereas the read distance may be more important in others.
We also said that RFID tags differ in the frequency range in which they operate. This concept should not be too foreign to us as we are all familiar with devices that operate at different frequencies. We know, for example, that if we wish to listen to a radio station we need to tune our receiver to a specific frequency. In this way the radio stations are kept separate and do not interfere with each other.
Similarly, while we might not worry that FM radio waves are affecting our health, we are more concerned about using a mobile phone on a regular basis and most of us are very familiar with the effect that microwaves produce in anything containing moisture. The point here is that these devices operate at different frequencies and at these frequencies the electromagnetic energy produced has different characteristics or behaviors. (5) This fact is used to good effect by RFID manufacturers who split operating frequencies into several ranges such as low frequency, high frequency, ultra high frequency, microwave frequencies. Frequency range is chosen to suit the requirements of each application. Two examples will illustrate this.
Here is a tag that is routinely injected under the skin or into the muscle of companion animals such as cats and dogs. Sometimes they are even injected into humans. What would be the set of requirements for an RFID system used in this application? First, the tag must be small and we cannot realistically remove it to swap batteries so it must derive its power over the air from the reader. So, a passive tag is needed. Secondly, the nature of the application means that we will always have to read this tag through a layer of tissue--skin, fat, muscle. Because of this, we need to choose our frequency carefully. Why? Remember that we said that different frequencies have different characteristics. Within the RFID world, one of the important differences between frequencies is how each one performs in the presence of water.
As a general rule the higher the frequency, the less penetration the electromagnetic field achieves through water. In fact we depend on this characteristic to heat food in a microwave oven which operates at a much higher frequency than many of the other devices with which we are familiar such as radios and mobile phones. If the microwaves simply passed through food, the heating effect would be minimal. Because the microwave energy is absorbed by the moisture within food, the food is heated. In the animal example, the various layers of animal tissue that we need to penetrate in order to read the tiny RFID tag are made up mostly of water. So if the field from the RFID reader is absorbed by the water component of the animal's tissue instead of passing cleanly through, it might make the animal warm (or worse) but there certainly will not be enough power left by the time the field reached the tag to power up the chip for normal operation. Therefore, in this application we need to select a low frequency system which permits the RFID field to easily penetrate to the tag. Because the operator has the animal in one hand and the reader in the other, we also do not need great read range and so a low frequency passive tag system fits the bill.
The point to keep in mind as we proceed is that the useful characteristics of the low frequency system are unique to this system, not generic qualities of all RFID systems. This point will be significant when we discuss issues of privacy later.
Vehicle tolling systems
The next example is the application of vehicle tolling systems. These systems employ an RFID tag within each vehicle, usually mounted somewhere near or attached to the front windscreen. The data on the tag is linked to an account on a remote computer and each time the vehicle passes a fixed point, the tag is read and the attached account is debited for the trip. The picture shows a system in Singapore but such systems are in use worldwide. What might be the requirements for an RFID solution for this application?
Obviously, the factors considered in the animal tagging application are not the ones important here. Probably the first concern would be how quickly we might need to read the tag for the system to work correctly. After all, the vehicles are usually travelling at speed--often freeway speeds. The next problem is that we have to read the tag successfully from quite a distance--the overhead gantry on which the RFID readers are mounted has to be high enough to allow the passage of large trucks. Also, at this distance it will be almost impossible to power the tag from the reader the reader's power diminishes very quickly with distance. So, read range, read speed, and power will be the three most serious concerns.
Fortunately we have enough space inside each vehicle to allow a bigger tag than could be injected into the companion animal so we can choose a battery powered tag. A feasible scheme can also be designed to replace the batteries when they are exhausted. We will choose either a semi passive tag or an active tag, both of which have batteries. Which to choose? Semi passive tags are used in the systems within Australia. (6) These sort of tags signal to the reader by reflecting some of the reader's power back, similar to a radar system, but use their internal battery to power the chip when it is needed. The tags usually operate at frequencies in the microwave range. This suits the application due to another variable along the frequency dimension; the higher the frequency, the greater the data transmission speed that can be achieved. The semi passive tag gives the read speed, the read range, and the internal power needed.
For the sake of completeness, when would we choose an active tag? As a general rule, these are required when even greater read ranges are required--perhaps hundreds of metres. Having larger batteries, they are able to power not only their internal chip but also to boost the transmission power and achieve distances far beyond their semi passive cousins.
Once again, the point to remember for later discussions about library RFID is that the characteristics of the RFID tolling system are entirely different from the animal tagging system. No single RFID technology platform possesses all of the things needed in all applications. There is no one size fits all in RFID.
RFID in libraries
RFID systems have been in existence within libraries for about a decade. Various vendors claim to have installed the first RFID library system but we can say that somewhere in the mid 1990s, RFID appeared in a library somewhere in the world. During this current decade usage has accelerated but penetration remains relatively low. Over the last few years RFID tag prices have continued to fall steadily to the point now where, in many system quotes, the total cost of the tags no longer represents the biggest single line item on the page.
A question often asked about RFID is 'Should we buy now or should we wait for things to get cheaper/better/more standardised?' There is no universal answer to this question but, once again, the situation reminds us of the first decade of library computer systems. At that time, as now, everyone knew that systems would improve over time. They would, and did, become cheaper, faster, smaller, more functional, and more standardised. So why did not libraries wait to buy? Some did but many did not.
The real question for those libraries in the 1980s was this--does the purchase of a library computer system give my library demonstrable service/productivity/management benefits right now? If the answer was yes, regardless of where in the constant path of evolution these library systems existed, many libraries signed the purchase order and reaped the benefits. Of course they knew that things would change and in a few years their system would be upgraded. The point was that the system gave their organisation real benefits in the meantime.
RFID is in the same situation. Will the tags get cheaper? Yes. Will standards become more clear over time? Yes. Will we end up with total interoperability between systems? Probably. All of these things could reasonably be expected to happen as the future unfolds. The important questions are these
* do we really know what we want RFID to deliver in our library or can we devise a process to find out?
* are there systems on the market that can deliver what we want?
* can we construct a positive and realistic business case to demonstrate the benefits?
* do we have the skills and experience, or access to them, to make the right system evaluation/selection?
If the answers to these questions are all positive, it makes sense to proceed in a structured way. Many libraries have done the analysis and proceeded to implement RFID. Many have achieved their expectations. Some have been disappointed.
After the previous discussion about various RFID technologies and frequencies, we could perform a similar analysis on the library application of RFID. Of course, this decision has largely already been made for us but it might be useful to briefly explore the reasoning that was used.
Obviously the first challenge in selecting a library RFID system is the size and the cost of the tag. We essentially need a smart label application that can be attached to library material in a relatively unobtrusive way--a passive tag of some sort without a battery. The tag needs to be flexible and durable. We need to read the tag over a distance of at least half a metre. We need the field to penetrate the tissue of borrowers with books under their arms (as well as the moisture contained within the pages of the books) if we use the RFID tag for security. We would also like to be able to read a stack of tags in a stack of books very quickly so that we can improve productivity for our staff and the self serve experience of our borrowers.
The best compromise fit with all of these factors was deemed at the outset to be high-frequency passive tags operating at 13.56 MHz. This frequency offered a good balance between data transfer rate and moisture immunity. The additional good news regarding this frequency was that it was available for use with RFID in just about every country in the world. (7) Library vendors adopted it and it has become almost the only tag/frequency combination used in libraries around the world today.
More recently, there has been interest shown in using ultra high frequency tags which operate at close to mobile phone frequencies (920 MHz in Australia) within the library application. These tags are usually found in supply chain management and logistics applications. In Australia, Civica has introduced such a system. Time will tell whether these UHF systems become more popular than the currently used 13.56 MHz systems.
Useful things to know about library RFID Privacy and data security
The comment was made previously that because RFID is based on multiple technology platforms and frequencies, confusion often exists regarding what can and cannot be done with the technology by a malicious adversary. (8) This confusion is seen particularly in the area of privacy and concern is strongest in the US where privacy issues for libraries are hotly debated. (9,10,11) When analysing what an adversary might or might not be able to accomplish, we can see that such a person or agency would be limited to some extent by the restrictions of the RFID technology chosen for a particular application. Unfortunately, this fact is often overlooked in the debate. Sometimes it is assumed that because an RFID tagged shipping container might be trackable at a distance of 100 metres, a person's library books might be scanned by someone hovering over their house in a helicopter.
Let us put aside for a moment the issue of who this helicopter equipped adversary might be or what their interest in reading habits is that prompts them to go to such lengths. The question to be asked is whether this or a similar scenario is possible. For reasons already touched on, the answer is no. The library application employs passive tags. An explanation of exactly how a passive tag derives its power from the RFID reader and communicates is beyond the scope of this paper but, suffice is to say, the physics involved in the process set an upper limit on the distance over which communications may take place. An important point to note is that this maximum distance exists regardless of the power of the reader. For tags operating at 13.56 MHz, as in the majority of libraries, this distance is approximately 3.5 metres. So, no adversary will read RFID tagged library material from a helicopter, or from the roof of a building across the street, or from any position outside a distance of 3.5 metres. In reality, even achieving a 3.5 metre range would be a serious challenge.
We can say, therefore, that many of the extreme scenarios that sometimes arise are simply rendered impossible by the physics involved with the type of RFID technology employed by libraries. An analogy could centre on a being from another planet performing a review of the transport systems on the earth. A superficial report to the alien's superior might indicate that, on earth, motor cars are wonderful devices--they can carry seven passengers, they can travel at 300 kilometres per hour, can climb steep hills and ford rivers, can haul large boats and travel 1000 kilometres on 60 litres of fuel.
The reality is that each of these things is true in the case of individual cars but no one car possesses all of these attributes. The car chosen to haul large boats is unlikely to travel 1000 kilometres on 60 litres of fuel, for example. So it is with RFID. One cannot generalise about possibilities across multiple applications as the technology used in each application may well be different.
So if the extreme examples are unrealistic, what is possible for a well equipped adversary? Certainly a covert RFID reader could be installed in a doorway, or as part of an airport screening system. Possibly a covert reader could be carried by an adversary and, should he or she get sufficiently close, could read the tags on library books carried by another. Most, but not all, of these sorts of privacy attacks assume that the adversary has access to additional information not carried by the RFID library tag. This is particularly true where the tag has only the item number--the number that is, or was, represented by the barcode.
To derive the title from the barcode requires access to the library's computer system. If the adversary has the titles and tag numbers correlated then this adversary could possibly use that information to screen passengers with specific materials as they attempt to board an aircraft, for example. From a technical perspective this would be entirely possible.
Many other forms of potential attacks have been described such as tracking, profiling and hotlisting. (12,13,14) A detailed analysis of all of these falls outside of the scope of this paper, as does the increased theoretical potential for digital vandalism when employing RFID within the library. (15) Libraries having specific concerns or who want more detailed information should seek expert assistance.
Many libraries take a pragmatic approach to the issue of borrower privacy with RFID. There are many lists available containing tips for circumstances where security is an issue. This is mine
* do not use RFID tags for borrower cards--if an RFID borrower card is required, employ specific, purpose built smart card technology, not a laminated RFID tag.
* reduce the amount of data on each RFID tag--putting only the primary item ID on the tag may assist in the customer relations effort where privacy is expected to be an issue.
* do not allow bibliographic searches by the public using the barcode--this reduces the likelihood that the item identifier and the title can be linked.
* ensure that transparency exists with regard to library RFID projects--this serves to help manage expectations and fears within the population of library users. Several high profile cases of covert RFID testing have been exposed in the retail area and have generated much negative publicity for the organisations involved.
* review security procedures generally--as some of the attacks mentioned depend on matching external information with the data on the RFID tag, ensure that the library is secure in all its practices and policies, particularly in the area of information and communications technologies. Consider a specialist audit.
* lobby vendors for improved security solutions--while not of immediate benefit, much development work could be done to improve the security status of RFID systems based on current standards. These include
** development of anonymous ID systems using randomly generated numbers on issued tags paired with real accession numbers on the library database
** improved authentication of readers
** improved password protection
** solutions based on ISO / IEC 18000-3 Mode 2
* support research into library specific solutions for current RFID privacy concerns.
Libraries have been aware of the benefits of standardisation and open systems for many years now and so it is not unusual to be asked about the situation with regard to RFID standards for libraries. The bad news is that there is no library specific standard for RFID. However many RFID vendors have taken a smart card standard from outside of the library application area and used part of this standard in the library smart label application. While this standard does not meet the needs of libraries in all areas (data security and therefore privacy is one), nevertheless it is a start. When a library RFID vendor claims its solution to be 'ISO compatible' what it is saying is that it employs tags conforming to ISO 15693 parts two and three or ISO 18000-3 Mode 1 which is the latest incarnation of this standard (to be precise, the former is a perfect subset of the latter). These standards define things like command sets for communication, encoding methods, timing.
What does this really mean for libraries? It means that the same physical tag may be used in systems from two vendors that both support the standards mentioned. Unfortunately it does not mean that the two systems will have interoperability. The reason that this is the case is not obvious at first glance.
When an ISO compatible system uses an ISO compatible RFID tag, it first formats the tag to accept library data. This is analogous to the use of a floppy disk in a pc. A truly blank, unformatted, disk cannot be read by the computer--it must be formatted magnetically before data may be written to it. The same situation exists with an RFID tag. The problem is that the standards do not specify how the data will be formatted--that is how it will be arranged in the tag's memory. So RFID vendors are required to devise a formatting scheme of their own. They have all done this and in every case the scheme is unique. When an ISO tag from one system is read by a second system, the second system cannot interpret the data contained on the tag as it knows nothing about its format. It does not know for example, where individual data fields start and stop, how the data is encoded, and even how many data objects the tag contains. So despite the ISO compatibility of both systems, there exists no interoperability between the two libraries involved.
Clearly this is not what libraries expect from standardisation. Fortunately, there are several working groups around the world addressing the problem. Some, such as the Danish, Finnish, and Dutch groups have finished their work and have developed national standards for their own countries. (16,17,18) The decision to be made is which format will be used as a global standard and be accepted by ISO, the international standards body. Within Standards Australia, a working group (chaired by the writer) has been working on a proposal that it considers suitable for both Australian use as well as international acceptance. The formal proposal is expected to be completed during 2006 and will be offered to ISO for consideration along with proposals from other nations. Libraries wishing to get more detailed information regarding the Australian Standards project should contact the writer.
Many libraries choose to abandon their electromagnetic (EM) systems in favour of using the RFID tag for both identification and security. The advantages are obvious, in requiring the application of only one tag and the savings associated with material and processing cost. When it comes to security, there are several options on the market which may be broadly summarised as
1 two tag systems--one RFID tag and one EM tag in each item
2 one tag systems using part of the RFID tag's memory to indicate the security status of the item
3 one tag systems that employ server based security and do not rely on part of the tag memory for security
4 one tag systems that employ proprietary technology on the tag to perform the security function
Which of these systems is to be preferred? What advantage or disadvantage is conferred by each option? Obviously a detailed analysis lies outside the scope of this paper but some broadly based, and necessarily subjective, observations may be made regarding the various options.
The first two options essentially represent a choice between EM security and RFID security. First, there is no solution that suits all libraries--no single superior choice for everyone. Much depends on where item security sits in a library's hierarchy of requirements. The fact that there are libraries which have EM gates installed but not actually switched on, whereas others have two sets of security gates back to back, tells us that not all libraries feel the same way about security. It is true, however, that libraries in general, with the possible exception of many in the school library sector, have chosen EM security over RF security. There are many reasons for this, including the ability to circulate tagged material easily, but a key driver is the superior detection rate of EM compared to RF. EM systems are harder to shield and the security tags are often covert and hard to spot. Now, RFID and RF security are not the same thing but RFID tags used for security do share some of the failings of traditional RF security tags.
Most noticeable is their sensitivity to orientation within the detection area and their weakness to being masked. There are certain orientations, or angles of presentation within the detection area of the gates, which make it more difficult for the passive tag to be powered by the reader's field. In these specific orientations the tag might not be read at all and so the security status of the tag, or even the actual presence of a tag, cannot be determined. Also, two tags in close proximity to each other (overlapping wholly or partly) can prevent either tag from being read. For this reason, some vendors such as 3M offer dual EM security / RFID identification configurations as part of their solutions.
So is RFID tag based security a nonstarter? Not necessarily--it depends on the requirements. A question that is often asked is 'How much worse or less effective is RFID based security than EM security?' Obviously a definitive answer is not possible as it depends, among other things, which vendor systems we are comparing. My sense is that an average RFID based security solution will have about 80 to 85 per cent of the detection performance of an average EM solution.
That is a conclusion based on observation and not on empirical data. Some vendors may do better, others might be worse. That number may be good enough for many libraries. I have often been told over the years that simply having a set of security gates on display solves 80 per cent of the problem anyway. If you have sympathy with that view you are unlikely be troubled by the shortcomings of RFID tag based security. If not, you will need to evaluate each vendor's solution more critically in this area.
What about options three and four on the list? To the fourth option first: these are systems where additional technology is added to what otherwise might be a standard tag to provide an Electronic Article Surveillance (EAS) function --sometimes described as tag-talks-first solutions. Tagsys is one vendor that can supply this sort of system. Vendors claim that such systems offer increased performance when compared to systems using a part of the tag's memory for the security status. Observation suggests that this claim has some merit. If the detection performance is particularly important, tests should be carried out with systems under consideration using both single items and multiple items in the detection field with items presented in several orientations. An hour spent in this fashion should provide some comparative results which can then be weighed in the context of other factors for or against each system.
The third option is an interesting one. In this scenario the tag is not used for security in the traditional sense but a cache of recent transactions is held on a server attached to the local network and when the tag's identity is read, its eligibility to leave the building is checked against the cache of transactions. Thus eligibility to leave is determined directly as a result of an actual checkout transaction. The most well known implementation of this option is by Checkpoint. When this server based system is combined with a worm tag which is contact programmed, some significant security benefits are generated including immunity from many over the air digital vandalism and security bit denial of service scenarios. Of course, this is achieved at the cost of increased information technology complexity.
Each library will need to establish its own weighting for item security within their overall set of requirements. Libraries who want to gain more detailed information regarding item security should seek expert help.
RFID solutions for libraries
From the foregoing it will have started to become obvious that there are many differences between RFID library vendor solutions on the market.
These differences are seen, for example, along the following dimensions
* solution focus
* systems architecture
* functionality of individual components
* product range
* product maturity
* size, market penetration, representation
* vendor capabilities and experience
* attitude toward and compliance with standards
How to choose a system? This can be a challenging task and the next section discusses recommendations.
Beginning an RFID investigation
How not to begin
The writer's experience in speaking with libraries about technology over the past 25 years has been that in many cases, when information is required regarding a new technology, the first person to be called is the library's systems vendor. It can be no accident that many of the major systems vendors now have their own library RFID solutions. In this commercial world it would be unrealistic to expect totally impartial advice from a party which stands to benefit from your purchase. However, that libraries turn to a trusted technology supplier is not the real problem. It could be argued that at least these suppliers have a vested interest in maintaining a long term relationship with their customers and so would be unlikely to sell them a 'lemon'.
The real problem lies with at what stage libraries choose to approach the market. While it can be easy to pick up the phone, when an environment exists where there is great variability in the solutions offered--as is the case with RFID library systems--it is always risky to evaluate advice without really knowing what is important in one's own case. It is only when a very clear picture of local requirements, expectations and targets exists that a meaningful evaluation of features and benefits can be made.
A better way
It can be a nontrivial exercise to establish what specifically is required in order for an RFID implementation to be declared a success. The writer has known of libraries that have entered into projects with no real vision of the outcome other than generalities concerning productivity and service.
A list of the promised benefits to the library through an implementation of RFID would probably look something like this
* increased productivity outcomes
* improved customer service
* streamlining of materials handling--improvements to occupational health and safety
* improved collection management
* enabler for future functionality
* maximising self serve loans and self serve returns
Although lists may vary from person to person, almost all would include these things. If we took such a list to all the libraries in our area and asked the library management whether these things were important we would almost universally receive the same positive answer: However if we tried to insist that all of these benefits were of equal value then we would probably be met with an argument. A benefit that is very important to one library may be of little relative importance to another. For example, collection management might be the single benefit most sought by one library whereas another might be focused almost exclusively on productivity gains. Even when two libraries rank productivity gains as their primary target, they might actually want two different outcomes.
So, we have an environment where there are significant differences in the strengths and weaknesses of vendor RFID solutions and we have requirements that are not consistent across all libraries. This would suggest that it is possible to choose systems that could be both suitable or unsuitable for our specific requirements. This is not to say that a given system might bring no benefit to an organisation. Rather, what is required to maximise benefit is that the strengths within the system chosen must match the critical factors for success within the library. This means that the library must have a very clear vision of what those success factors are before approaching the market. To do otherwise is to risk having the success factors influenced by the pitch of a particular vendor. This is a particular risk when the knowledge gap between the systems supplier and the library is great.
A suggestion is to hold a workshop with key people to evaluate exactly what the priorities are for a successful migration to RFID. What does the word 'productivity' mean to this organisation? How would we describe a successful implementation of the technology? What would it mean for the staff?. For the users? What would change and what would remain the same? Can we be very specific about what we expect to gain?
These are important questions that deserve time to explore. The answers can sometimes be challenging, controversial, and surprising. Once armed with a clear picture of success however, a library is in a good position to weigh the benefits offered by current systems. Libraries that do not have access to the necessary experience and skills to reach this position should seek outside assistance at this crucial stage.
How to get more information
There are not too many good published sources of unbiased and independent information specifically for libraries. However, more libraries are embracing RFID and their experience should be considered as a minimum. The websites of the library RFID vendors are also a good place to begin to gather information for a comparative study. There are also consultants specialising in RFID within libraries but not many as yet. Most books on RFID, if they mention the library at all, do so only in passing. The author's website has a number of links which can be followed for more information on standards, privacy, vendors, and general RFID matters. For those interested in the technical detail, the book RFID handbook: fundamentals and applications in contactless smart cards and identification by Klaus Finkenzellar (Wiley) is probably the best general work. RFID and contactless smart card applications by Dominique Paret also has good technical information about systems based on ISO / IEC 15693.
For a discussion of RFID privacy issues, the book RFID, applications, security, and privacy edited by Garfinkel and Rosenberg (Addison Wesley) has a number of interesting chapters including one on privacy within libraries. The paper by Molner and Wagner listed in the references gives an overview of the technical challenges involved in creating a secure platform using current RFID standards.
This paper has attempted to touch on a number of areas whilst trying not to get bogged down in detail or technical jargon. Clearly, more attention may need to be given to specific areas before committing to a system. Libraries are well advised to consult broadly and to invest time developing critical success factors before commencing discussions with vendors. As with many projects, the time invested at the start is often reflected in the quality of the outcome.
(1) Shepard, S Radio frequency identification McGraw Hill 2005 p42
(2) Lahari, S RFID sourcebook IBM Press 2006 p9-10
(3) Glover, B and Bhatt, H RFID essentials 2006 O'Reilly p58
(4) Transurban Electronic toll collection roadside systems--familiarisation course Transurban, Edition 1 2001
(5) Finkenzeller, K RFID handbook: fundamentals and applications in contactless smart cards and identification Wiley 2004
(6) Transurban op cit
(7) Paret, D RFID and contactless smart card applications Wiley 2005 pp91-93
(8) Mather, K and Wiebell, H RFID should be restricted by the state Information age December 2005/January 2006
(9) O'Connor, M San Francisco Library denied funds for RFID RFID journal 2005 www.rfid journal.com/article/articleview/1708/1/1/
(10) Electronic Frontier Foundation Keep RFIDs out of Californian public libraries! 2005 www.eff.org/Privacy/Surveillance/RFID/
(11) Ayre, L Wireless tracking in the library: benefits, threats, and responsibilities in RFID applications, security and privacy Addison Wesley 2005 pp229-243
(12) Molnar, D and Wagner, D Privacy and security in library RFID--issues, practices and architectures CCS'04 October 25-29 2004 Washington DC, USA
(13) Albrecht, K RFID: the doomsday scenario in RFID applications, security and privacy Addison Wesley 2006 pp259-273
(14) Givens, B Activists: communication with consumers, speaking truth to policy makers, in RFID applications, security and privacy Addison Wesley 2006 p432
(15) Tanenbaum, A, Crispo, B and Rieback, M Is your cat infected with a computer virus? 2006 www.rfidvirus.org/papers/percom.06.pdf
(16) RFID Data model for libraries working group RFID data model for libraries--proposal for a data model 2005 www.bibliothecarfid.com/docs/ dokumente/Datenmodell.pdf
(17) Finnish Libraries RFID working group Finnish data model 2005 www.lib.helsinki.fi/katve/toim inta/docs/RFID-DataModel-FI-20051124.pdf
(18) NBLC Netherlands Association of Public Libraries Generic set of requirements RFID for public libraries, version 0300 2004 www.dbc.dk/medier/intspec02-10.pdf
Generally used term for an RFID tag that contains a battery powering both the tag's transmission as well as the onboard chip.
The wireless free space through which the RFID reader communicates with the RFID tag via a specific communication protocol.
A conductive device designed to receive or radiate electromagnetic energy. In RFID systems, antennas may be used to both receive and transmit data modulated electromagnetic energy.
The silicon component containing the memory and processing capacity of the RFID tag. The chip is very small when compared with the attached antenna. Memory capacity varies between chip types.
The range of electromagnetic radiation characterised in terms of frequency or wavelength.
The number of cycles a periodic signal (such as a radio wave) executes within a given time. Usually expressed in Hertz (cycles per second). RFID systems operate at thousands of Hertz (kilohertz), millions of Hertz (megahertz) or billions of Hertz (gigahertz).
The process by which power is transferred from the RFID reader to the tag using a magnetic field without contact. Method of powering passive tags.
The ability to use tags, components and modules from different RFID vendors within a single system.
A method of storing data in an electronic form. Within RFID tags, memory capacity is expressed in Bits.
In the world of RFID, the ability to capture, store, handle and communicate data according to defined standards and between systems.
An RFID tag without a battery that is powered by inductively coupling to the reader. Power is transferred by the electromagnetic field of the reader.
The ability to enter and to change data stored in an RFID tag's memory.
A defined set of rules governing how (for example) an RFID reader communicates with an RFID tag.
The maximum distance between the RFID reader and the tag at which the tag may be effectively read. This distance is affected by the power output of the reader, the orientation of the tag in the field, the method of communication, and the electrical environment.
The maximum rate that data can be communicated between the RFID reader and the tag. Expressed in bits per second.
Generally used term for an RFID transponder.
An electronic device that can both transmit and respond--an RFID tag.
Write Once--Read Many. A tag that is programmed once and then can subsequently only be read.
Alan Butters Principal consultant Sybis Victoria
Alan Butters is principal consultant at Sybis, a Melbourne based technology consultancy focused on the needs of Australian libraries. He has a Masters in Digital Communications from Monash University where he also teaches in the School of Information Management and Systems. Alan specialises in library RFID systems and process automation technologies such as self serve loans and returns and materials sorting. He has almost 25 years experience within the library sector including roles as technical manager for 3M Australia and Raeco International where he managed product development laboratories generating innovative solutions for the global library market. Alan chairs a working group within Standards Australia committed to developing a standards proposal for an RFID tag data model for Australian libraries. Address: 6/180 Central Road Nunawading Vic 3131 Tel(03)98786447 firstname.lastname@example.org
|Printer friendly Cite/link Email Feedback|
|Publication:||Australasian Public Libraries and Information Services|
|Date:||Dec 1, 2006|
|Previous Article:||Emerging technologies changing public library service delivery models.|
|Next Article:||Reader development in the UK: an Australian perspective.|