2014 Library RFID Survey – Part Two


Day two of this year’s analysis looks in more detail at how libraries are using RFID to manage their assets, which frequencies and standards are most common and how RFID systems communicate with library management systems (usually referred to as “LMS” or “ILS”).

Managing assets

The survey reveals that although RFID is used predominantly to manage books there are many other assets for which solutions have been found.

CD/DVDs 318
Journals 203
Music Scores 109
Laptops/tablets   51

The relatively high number of libraries using RFID to manage laptops and tablets may surprise some (indeed one respondent quite rightly commented that RFID doesn’t work near metal) but the inventiveness of both suppliers and librarians knows no limits and solutions range from tags attached by Kensington lock, oversize tags (expensive according to one respondent), to what appears to be the most popular method – tagging the locations in which devices are stored rather than the devices themselves.

But the range of items being managed doesn’t end there. In addition to these more obvious uses librarians reported using to manage all sorts of objects, including:

  • Audio Books.
  • Jigsaws.
  • Developmental aids (puzzles, toys, etc.).
  • Empty book covers (dummies) for QR-Codes
  • PC-mice.
  • Book bins and baskets
  • Game consoles.
  • Models, bags of bones and other objects used by medical students;
  • Videos, cassettes and LPs used by music and drama students.
  • Toys; games; kindles; comics; graphic novels;
  • Ukuleles.

Given the current transformation of library services to a more digital and technological platform we can only expect this list to grow, even perhaps seeing more use of RFID to “blur” the distinction between the physical and virtual worlds as in this example:

Smart shelf in use at Oslo Public Library

Buying the tags

RFID can be expensive, and one of the items often overlooked by librarians when making their purchasing decisions is the ongoing cost of the tags themselves.

Before 2011 countries that had not developed a national data model were at the mercy of the market when it came to buying tags. Since tags were programmed only to work with the equipment supplied by the RFID vendor it was the vendors that generally continued to supply tags in the long term. At least that was the safest option for librarians since most suppliers would not warrant tags supplied by anyone else.

Since 2011 the opportunity has existed for libraries to buy the same tags from the same source as the RFID supplier – the manufacturer. Yet few appear to be taking advantage of this option. Most libraries still buy their tags from the company that supplied the bulk of their hardware (relatively few libraries buy from more than one supplier).

Respondents were asked where they sourced their RFID tags:

RFID solution provider 343
Tag manufacturer 50
From a jobber/book/media supplier 45

19 libraries reported buying tags from all three sources, while 27 buy directly only from the tag manufacturer. 17 still buy tags only from their book/media supplier.

I have been hearing recently about the possibility of libraries forming consortia to buy tags (something I have long advocated) – the price points from the manufacturer being set to encourage large volume sales – but this was not reflected in the survey.

Data Standards

The next section of the survey examined the somewhat confused world of data standards. As mentioned earlier prior to 2011 there were only a few examples of national standards in existence – the most famous probably being that developed by the Danes as early as 2006. Following revisions in 2009 the Danish Data Model is now almost indistinguishable from ISO 28560-1 and ISO 28560-3.

Since 2011 there has been more consolidation of data standards. The USA, UK, Australia and New Zealand among others determined to use ISO 28560-2 in the belief that it offers more flexibility than 28560-3 – which is based more closely on the earlier models developed by the Danes and others. Both versions use almost the same data elements and values – codified in ISO 28560-1.

So the picture is less confused than it used to be – and most librarians now seem to have understood the significance of the “data” standard – i.e. what goes on the tag as opposed to a “communication” (or “air interface”) standard – i.e. how the data is carried over the air compared to earlier surveys.

Nevertheless there are still a few who aren’t clear about data standards – as the answers to the next two questions showed. First question, “Do you use a data standard on your tags?” The replies were very encouraging for a standards enthusiast like me:

Yes 295
Don’t know 63
Did Not Answer 33
No 28

295 libraries represents 70% of those that completed the survey! Back in 2009 only a handful had even heard of a data model. The next question was, “Which one do you use?”:

ISO 28560-2 103
ISO 28560-3 49
Danish Data Model 48
Don’t know 46
ISO 15693 and ISO 18000-3 6
ISO 28560 6
3M 5
ISO 15693 4
idrabib01 3
FR01 (French recommendation for RFID tags) 2
ISO 28560-1 2
Dutch 1
ISO 18000-3 1
ITG model 16 1
Swedish 1
Various proprietary 1

Just a couple of comments to make here.

The version of the international data standard present on tags will either be ISO 28560-2 or 28560-3. ISO 28560-1 lists the elements and values that can be encoded by using either version -2 or -3. ISO 28560 on its own doesn’t mean much without the supplementary digit. But it does indicate that a data standard is probably in use on the site.

ISO 15693 and ISO 18000-3 are not in fact data standards at all but relate to the way in which data (any data) is transmitted and received. 3M do not appear to have a single data standard having modified the data they store on their tags many times over the years – and have also varied the data they choose to store from site to site. That’s NOT a criticism – 3M have simply been doing this for a very long time now and many of their solutions pre-date ANY data standards.

EAS and AFI refer to security bits not a data standard. Normally you would use one or the other and not both – but it happens.

All that said – and these are small numbers – the picture shows a market that has become much more homogeneous in the past 3 years and which now presents a much more attractive market for both new providers – who have fewer local differences to take into account when developing their offer, and for libraries seeking to innovate, migrate or co-operate existing systems.

Scottish NEC reserves space for library related data.

Scottish NEC reserves space for library related data.


Smartcards – as opposed to book (and other asset) tags – currently have no data standard for use in libraries although there are examples of data modelling for local authority and national ID cards that specify parameters for library data such as the Scottish National Entitlement Card. In British Columbia the “One Card” uses a data model that is used by all of the libraries in the province. Given current concerns about RFID privacy (in the US and the EU) this is an issue that may yet prove significant. Smartcard developers in the EU for example will need to be cautious when adding personal data to smartcards or risk prosecution under new legislation likely to be introduced in the next few years.


Identifying the Owning Library

Whilst we’re on the subject of standards it’s worth mentioning ISIL (International Standard Identifier for Libraries).

As more and more libraries join consortia or interlending schemes the need to identify the institution that owns the asset becomes increasingly important. Library Management Systems (LMS or ILS) developed ways to manage this scenario in the days when only a barcode number (which may be duplicated between libraries) was available as a means of identifying items but RFID allows the storage of much more data than a barcode and ISO 28560 (all versions) and most national data models allow or mandate the use of the ISIL code to declare ownership. Many countries now have a national agency to manage and issue ISIL codes to libraries using RFID systems and much has been written on the subject – not least by me.

So I was interested to know how many libraries now use this code in their RFID solutions.

The survey asked if the library used an identifier of any kind to indicate ownership:

Don’t know 176
Did Not Answer 78
None 48
Other 26

Whilst the 91 libraries using the ISIL code is encouraging it shows that libraries aren’t exactly excited about identifying ownership of their assets. Understandably there are many libraries that will not see any need to do so since they don’t presently loan items to other libraries. I would argue that this is a rather narrow view on several counts.

  • It costs almost nothing to do it
  • It’s another way to identify stolen assets
  • Your library may change its loan policy one day
  • Your library may join a consortium
  • You may be required by legislation to share resources
  • The nature of the assets you manage is changing – learn good habits now!
  • It’s much easier to do this now than have to do it later.

But I’m happy to debate the subject!


Currently the library world uses two different frequencies to run their RFID solutions – High Frequency (HF) operating at 13.56MHz and Ultra High Frequency (UHF) operating at a range of frequencies dependent on supplier.

Each of the frequencies has its advocates and each has both advantages and disadvantages. Much has been written on the subject elsewhere.

My concern with the survey was to attempt to take a snapshot of the current distribution of the two frequencies. New standards are being developed for UHF – to take account of the increased data capacity of the latest UHF tags and standards bodies are eager to ascertain the relative size of the markets.

Sadly for me the largest market for UHF systems appears to be China and Japan both markets effectively closed to the survey by my lack of the necessary language skills for either market. The figures obtained are therefore unrepresentative of the situation existing outside of the markets I defined earlier.

Only 6 libraries reported using UHF systems – 3 in Australia, 2 in the USA and one in Malaysia. A number of libraries reported that they thought they were using UHF but consultation with their supplier revealed that they did not supply anything other than HF solutions so the figures have been adjusted to reflect this.

334 reported using HF – leaving 79 that either did not know or did not answer. The dominance of HF in the market seems fairly well established.

Communicating with the LMS/ILS

Finally, in today’s post, let’s look at communications. Most RFID systems can only deliver the functionality that the LMS/ILS can support and for years now that has been limited mostly to circulation supported by 3M® ’s Standard Interchange Protocol (aka SIP).

Many ILS/LMS developers have worked with individual RFID suppliers to create Application Programming Interfaces (APIs) to overcome these limitations but of necessity these APIs generally don’t migrate so well. For many of the university and academic libraries this is not an insurmountable problem since they have the skills in-house to make the necessary modifications – or indeed write new APIs if things change.

A new initiative – called the Library Communications Framework – seeks to mitigate these limitations by creating an open framework within which new services and functions can be supported to better exploit RFID technology. If you’re interested you can read more about it here.

In the meantime asking about the means by which library RFID and ILS/LMS communicate is a useful indicator of the degree to which systems are being used to support more than self-service.

The survey asked “By what means does your RFID system communicate with your library management system?”

SIP 299
API 22
Don’t know 4


SIP remains dominant. Admittedly SIP does allow for the use of extensions to facilitate additional functionality so by no means all of the 299 libraries using it are necessarily limited to using RFID for circulation, however since these “extensions” are not regulated in any way (standardised) and may vary from library to library they cannot really be considered a “standard” protocol.

Very few libraries reported using an API – which was a little surprising. I plan to further analyse these data in a future post when I will be concentrating on regional differences.

But the next post will be looking at supplier performance.

Have a view? Please share!