This week saw the appearance of one of the most eagerly awaited documents (by me at any rate) in library RFID. Having spent the last four years of my life trying to persuade librarians (through publications, conferences and even official guidance – all available on my “resources” blog) that there really are some important issues that they need to consider before spending too heavily on the technology it comes as some relief to see a body as august as NISO endorsing that view.
US libraries have been much slower (in my opinion wisely) to adopt RFID on the scale we have seen in the UK and Australia. Whilst I’d like to think that this is a result of natural caution and a desire to wait for standards to be developed to regulate an industry quite capable of moving in several different directions at once I fear that it may have had more to do with (in my opinion largely misplaced) fears about privacy violation.
The NISO document dwells on the privacy issue at some length – and offers guidance to help libraries protect themselves against unwitting violation of the law. With the EU having recently invited expert opinion to advise it on the same issues – in libraries – this is a topic we are likely to hear more about very soon.
This post will however focus on the question of data standards.
Only this morning I posted the most recent results from this year’s global survey of RFID use in libraries. Sadly they showed that there are still very few libraries using any kind of data standard – and that even those that believe they are – probably aren’t.
So why should you consider doing so?
The NISO guidelines speak very concisely to that issue. In its opening chapters it states unequivocally that,
“Early RFID implementers are at considerable risk because of the lack of interoperability of proprietary vendor systems. As RFID providers and libraries adopt tags with the data model recommended in this recommended practice, true interoperability that allows libraries to procure the tags, hardware, and software from independent providers and distributors to use with all tags can become a reality.
The data model outlined in this document is an essential first step. This model is a key precursor to a world in which a library can procure tags from different vendors, merge collections containing tags from different vendors, and, for the purposes of interlibrary loan, read the tags on items belonging to other libraries.
Even with a data model, there are other barriers to interoperability and plug-and-play capabilities. They include:
- Vendor-specific encrypting and encoding of the data.
- Proprietary security functions, which are an advantage when considering hackers, thieves,
etc., but are a detriment to interoperability
- Software or firmware that is system dependent and can only be used with specific tags.
The ideal is that RFID tags compliant with the data model can be usable by other RFID vendors. With standards recommended in this document, interoperability and the ability to embed tags into books at manufacture is within reach.”
(emphases are mine)
None of this should be news to UK librarians. Guidance issued as early as 2003 made the same points about interoperability and even UK suppliers themselves endorsed the need for interoperability between different RFID suppliers in their 2009 statement.
Yet only a handful of UK libraries are even aware that the standard exists – and even fewer have deployed it in their libraries. Many believe that they have nothing to worry about – but NISO challenges such complacency,
“Libraries with existing RFID installations are unlikely to be compliant with all the data requirements of the new standard. While most tags purchased today and in recent years are in compliance with ISO 18000-3 Mode 1, vendors have historically used proprietary approaches to encoding data on the tags. In other words, even if two installations use ISO 18000-3 Mode 1-compatible tags and readers, they may not be able to interoperate since the data at each respective site may be encoded differently.
Think about going to a bookstore and selecting a book in a language you cannot read. While you may be able to recognize the letters and see words and paragraphs, you cannot decipher the actual meaning of the text.
Defining a standard data model for encoding is the problem that ISO 28560 intends to address. The question now is whether a site with existing tags and equipment can successfully migrate to an ISO 28560 compliant solution. Theoretically, if a library uses ISO 18000-3 Mode 1-compliant tags, readers, and other hardware, such a migration may be possible. However, issues like whether the data on the existing tags are locked and what sort of model is in place for security will affect migration.
While the prospect of migrating from proprietary to standardized systems can be daunting, with some careful planning and a good understanding of an organization’s goals and existing components, the labor and disruption involved can be minimized. Over time, thanks to a uniform standard, libraries will be able to operate with equipment from multiple vendors and suppliers.”
This is an important document. Happily for those of us (like BIC and NAG) who have been working hard to create the level playing field that NISO now seeks to create in the USA almost every word in the document coincides precisely with industry aims on this side of the pond. That means that as RFID now finally looks set to take off in the USA those libraries that have heeded the warnings will be ready to take advantage of the changes likely to sweep across the industry. The NISO document lists some of the areas of library activity that are likely to benefit from the adoption of the standard.
It also lists the pros and cons of migration – starting with the pros:
- Interoperability between libraries – This is a primary goal of the standardization activities. Libraries want to be able to read tags that are affixed to items owned by other libraries and that in many cases were programmed by systems produced by other vendors. By upgrading systems to support standards and by migrating tag data into standard formats, this kind of library-to-library interoperability can be achieved.
- Supply stream interoperability – A standard data model will make it much less expensive for distributors who offer pre-processing and will be an easier option for publishers and other supply stream elements to consider.
- Good citizenship – The AFI on an RFID tag is a means of ensuring that the application of RFID in the library industry does not interfere with RFID uses in other areas, and vice versa. To be good electronic citizens, all implementers should make sure that they are ISO compliant and using an officially assigned AFI code. The codes are referred to in Section 3.2.
- Equipment replacement – This is the other benefit of interoperability. Libraries are concerned about the future value of their investments and they are resistant to the concept of being locked to a particular vendor based on past choices. By migrating tag data formats to a standard, when equipment upgrades and expansions are considered in the future, the library may select and even mix and match components from standards-compliant system vendors. This is a benefit because it encourages competition, drives innovation, and reduces the need for compromises by the library.
And some of the cons:
- Information about tag formatting is public – Standards and data models are, by their nature, public documents. Other sections of this document describe how a sufficiently informed or clever vandal or thief might use an RFID reader to vandalize tags or to steal an item from a library. There is a finite risk in migrating to a standard and therefore to a publicly available data format. However, this is only an incremental difference from the proprietary data formats, which, unless truly encrypted, are generally not difficult to decipher for a technically oriented individual with an RFID reader.
- Labor requirements – There will be some labor effort for a migration to any new data model, such as the one described here. There will be work involved in equipment upgrades and staff time will be consumed where personnel are employed to reprogram tags, as well. The labor required for a migration can be minimized through thoughtful planning and many vendors offer services to assist with this process.
- System performance – During a migration period, when systems will need to deal with two tag data formats, there will be some small but perhaps noticeable performance reductions in different pieces of equipment. For example, if a security gate must run two security protocols in alternating fashion, perhaps for a second or so (probably less) for each protocol, the overall rate of detection will be reduced.
- Upgrade costs – There is always a cost associated with changing equipment and software. How this cost is absorbed by the industry will probably vary from vendor to vendor and library to library.
There’s much more in the document that’s well worth reading from migration strategies to supply chain modeling. Don’t be deterred by the lengthy sections on actual encoding, they’re from the standard itself and are the same everywhere.
Download it here.