Description

Loss rates

Despite the fact that, thanks to the invention of moving type by Gutenberg, books have become affordable for almost everyone, library books still remain an attractive target for thieves. In fact, the uninvited appropriation of these desirable goods from public and private libraries has, it seems, been en vogue since libraries first existed. Hamlin[1] wrote that the “most common and the most distressing enemy of books [is] homo sapiens.” Shuman[2] even went as far as to draw up a typology of book thieves.

Libraries, unlike retailers, do not undertake regular inventories of their stock, and precise figures for loss rates are therefore comparatively rare. Nevertheless, libraries estimate an annual “perceived” loss rate of between 3 and 5 %.[3] Some estimates and analyses, such as that undertaken at the University Library in Magdeburg, Germany,[4] or at a branch library of the ETH Federal Institute of Technology in Zurich,[5] revealed much lower values in the thousandths of a %. By comparison, figures from German retail suggest an annual loss rate of around 1.1 %.[6] In 2010 and 2011, the theft rate for the sector “books/magazines/stationery” was given as 1.7 % of turnover, significantly higher than the overall mean value of 1.3 % for all sectors.[7]

Spectacular incidences of theft remain the proverbial tip of the iceberg.[8] One such example is the so-called Shinn Lists in the USA, an inventory of the book losses attributed to a certain James Richard Coffman, alias James Shinn. In 1982, a book was even written about this “most accomplished book thief” as was noted on the cover.[9] In Germany, too, there are anecdotes and cases aplenty that from time to time attract the attention of the media. A few years ago, the case of a caretaker at the Erlangen-Nuremberg University Library made the headlines in Germany. Over a period of 20 years, he stole and sold books with a value totaling over Euro 400,000.[10] No less spectacular is the case of a civil servant at the Federal State of Hessen Ministry of Culture who “procured” more than 13,000 volumes from different libraries with an estimated total value in the millions. Aside from the obvious financial motive, some perpetrators are motivated by a passion for collecting, as was the case with a geologist from Darmstadt.[11]

Aside from these special cases, which do however illustrate the potential extent of the problem, there are a few, mostly older studies of loss rates that were determined following an audit of library stocks. One reason for this, aside from the significant time and staffing resources they consume, is that – unlike in the retail sector – there is no obligation to conduct inventories. Bahr quotes examples from university libraries in the USA that registered loss rates in excess of 10 %, and an inventory undertaken at the New York Public Library in 1975 determined that some 15,000 books had been stolen.[12] Wegner[13] quotes another study undertaken in 1996 of a college library with a loss rate of over 8 %. The subject library for economics at the Justus Liebig University in Gießen, Germany, which has a stock of 50,000–100,000 media units, was able to determine a loss rate of approximately ten books every month by undertaking regular inventories between 1993 and 2002.[14] After the introduction of a book security system for some of the media, the mean loss rate fell from 10.3 to 2.5 books per month. Such loss rates nevertheless vary significantly from library to library.

An area in which little data is available is the loss rate of audio-visual media (DVDs and CDs), which are often especially desirable items in library collections. Kahn noted that, “Today, DVDs seem to be the item of choice for theft and damage. In the first two years that DVDs were introduced to libraries, more than 50 % of them were checked out and never returned.”[15] This is one of the reasons why many information-­provision institutions put covers, dummies or empty cases in their public areas, and only hand out the respective audio-visual media at the loans counter.

The Scope of Book Security

Most people’s idea of book security is an inked stamp in the pages of an actual book printed on paper that indicates that the book may not be removed from the library without permission. The item may be used in the specified part of the library, usually the reading room, and sometimes also borrowed for use at home once it has been checked out. In Germany, this aspect of book security is also detailed in the currently valid DIN Technical Report 13.[16]

But book security also encompasses three further aspects: firstly, various other non-book materials also need securing such as CDs, DVDs, games, maps, USB sticks or microfiches in academic libraries. As such, the term “media security” would be a more accurate description. The second aspect concerns the considerate use of media. The condition of loaned media is generally checked when they are returned at the end of the loan period, but it is not as easy to monitor how media are used within the library. In the TV episode “Mr. Bean at the Library”, Mr. Bean manages, in his inimitable slapstick manner, to completely mutilate an antique book in the reading room. While humorous and exaggerated, it does portray a real problem.[17] An obvious way of preventing damage is to have staff on hand to supervise. Technical means such as camera surveillance[18] or closed-circuit television, as used in the British Library,[19] have been shown to reduce incidences of the mishandling of media in libraries. The third aspect is the inability to locate media, i.e. not the loss of the item itself, but of access to it. Media are usually organized by subject so that publications that belong together can be found in the same vicinity. In book stacks, which in some cases are made accessible to the public at a later date, media are often organized according to the numerus currens principle, i.e. consecutively numbered in the sequence of their acquisition. It is not uncommon for books on entirely different subjects to stand next to each other.

If a particular item is available but cannot be found on the shelf in the designated place according to the organizational system or the online catalogue, they are classed as lost, unless they happen to be in use at that moment. For example, of the 2.5 million books at Duisburg-Essen University Library, some 3,000 are classed as missing, but that does not necessarily mean they have been stolen.[20] Librarians are familiar with book hideaways, where individuals place books in a position known only to them to prevent other users from accessing and removing them. Few libraries have the resources to regularly undertake inventories or audits of their increasingly large collections. However, for the user misplaced media are just as unusable as stolen media. Some experts even maintain that the practice of deliberately misplacing books represents a greater problem than their theft.

The inventory or stocktaking of library materials is an area that librarians can expect to change in future. The comparatively recent introduction of RFID (radio frequency identification) systems can help reduce the effort considerably. The first libraries to be equipped with RFID are successfully using it to undertake inventories of their stock, for example the art library in Sitterwerk near St. Gallen in Switzerland. The location of the 16,000 items in their stock is polled several times a day using an automated RFID inventory tracking procedure.[21] Misplaced media are even tolerated for test purposes because the position of the item can be determined and visualized based on the automated inventory system.

An RFID-based inventory of over 61,000 media in the open-access area of the Technische Hochschule Wildau near Berlin, conducted in early 2013, took a tolerable 40 work days and resulted in a recognition rate of 99 % and a reading rate of 60 media per minute. A loss rate of 550 media was determined. The widespread use of this approach may make it possible to soon have more reliable data on loss rates in libraries.

The Introduction of Article and Media Security Systems

Electronic article surveillance systems (EAS) were first developed for the retail sector as a means of preventing the theft of articles for sale. This term encompasses various different technical systems, including electromagnetic security systems (EM) and radio frequency identification systems (RFID), both of which are used in libraries.

Such EAS systems have been used for several years in library environments. In most systems, a label or button-sized tag is attached to each individual item or article (the media item itself or the media box). Different sizes and designs are available. Using different methods of contactless information transfer, the status of the tag is read or written (bought, borrowed, returned and so on). This information is written at the point of sale or loans desk. This fairly cost-intensive technology only makes sense if the articles in question have a certain sales price. As such, these systems are most commonly used in the electronics and clothing sectors. In the book sector, some examples of application scenarios already exist. The Boekhandels Groep Nederland has used RFID since 2006 in the UHF frequency band to secure books for sale with a value of at least Euro 5. Similarly the publisher Wiley & Sons have started using UHF labels to label their 500,000 non-fiction books, though their primary motivation is to be able to better track the flow of goods.

Electromagnetic Security Systems

Developed in the USA in the mid–1960s,[22] electromagnetic security systems were first used more widely in libraries in the 1970s in the form of “Tattle Tapes”.[23] Shortly after, similar approaches also began to be used in Germany.[24] Each book or media is tagged with at least one strip of electromagnetic metal tape (EM), usually in an unobtrusive position near the spine between two pages. The Biblioteca Central at the Universidad Nacional Autónoma de México (UNAM) even secured each book or media with three such magnetic stripes. These stripes can switch between just two states: magnetized or non-magnetized. They cannot hold any other information. When active, the magnetic stripe sets off a visual or acoustic alarm signal as it passes through a suitably equipped barrier-free gateway or double-leaf doorway at the entrance and/or exit. To be able to pass through the gateway without triggering the alarm, the strip must be de-magnetized, usually at the loans counter or with a mobile scanner.[25]

EM systems are only used for security purposes. For other needs, such as self-service loan systems and semi-automated returns, a second parallel system is required, usually involving barcodes. EM strips are also available for audio-visual media and are sometimes used for securing a library’s entire collection. The German DIN Technical Report 13
[26] on book security notes that pathways in the library must direct users towards a central control desk that cannot be bypassed. This needs to be taken into consideration when positioning the entry and exit gateways, regardless of the type of technology used.

The experience of using EM strips seems to vary: on the one hand, there are many thousands of installations in libraries around the world, and on the other, it exhibits a number of security vulnerabilities in different areas. One example was a trial by students at Augsburg University in 2010 in which different test-thefts were undertaken at the library, revealing the system’s patchy detection rate.[27] There has been no definitive conclusion as to the effectiveness of this theft-prevention system.

RFID Systems

Towards the end of the 1990s, RFID (radio frequency identification) systems began to be used more widely in library environments. This system employs high-frequency radio waves (868 MHz instead of the usual HF band with 13.56 MHz)[28] to enable the contact-less airborne identification of objects according to the open ISO 18000–3 Mode 2 and ISO 15693 Standards and does not require a line of sight. Using a credit card-sized label, it can even identify individual items in a stack of books. An RFID system consists of a transponder attached to an object that contains an ID code as well as a reading device to capture the ID code.

RFID transponders can be as small as a grain of rice and it is also possible, using a special printing method, to create stable circuitry made of polymers. The advantage of this technology is a combination of its small size, the inconspicuous retrieval of the RFID data and its low price (some cost in the range of a few cents).

In contrast to EM stripes, RFID tags, as the radio frequency labels are called, can hold more information, which can also be modified, making them predestined for other purposes in addition to security. The most common application is for self-service book loan systems.[29] The ability to combine media security with self-service loan and return systems, along with the added possibility of RFID inventory tracking, has contributed to its popularity, and RFID systems are used in more than 10,000 libraries around the world. The demand has in turn spurred development of RFID products, for example the 3D location sensing of RFID tags in security gates and open standards for the data models that determine the information stored on a tag and its structure.

Much has been written about RFID systems and their integration in library contexts,[30] for example on an internet blog by Mick Fortune,[31] who has become known as “Mr. RFID”.[32] RFID systems require two components. The first is the RFID transponder as coupling element that is affixed to or within the media. It consists of a ring-antenna and a chip that serves as the data carrying and identification device. The second is the RFID reader, which serves as a reader and writer and employs inductive coupling, using the principle of a load-modulated resonant circuit to provide both an energy supply for the transponder as well as for the data transfer using electromagnetic waves. The distance between the two components – the reader and the tag – varies according to the configuration used but must usually be within 1 m. Metallic objects – like many of the shelving systems used in libraries – can impact on the efficacy of the system. The media-security gates placed at the entrance and exit of libraries – typically barrier-free gateways of just over 1 m wide – are to all intents and purposes indistinguishable from EM security gates. A security gate consists of at least two antennas embedded in the walls that flank the gateway and a reader for reading and writing radio signals, likewise concealed behind the side walls or fascia of the gate.

The RFID tag, as the coupling element, can contain modifiable information, whether protected or not, depending on the book’s or media’s loan status. The majority of high-frequency (HF) installations use 13.56 MHz store data according to the ISO-standardized Application Family Identifier (AFI) values, which use pairs of hexadecimal 8-bit codes to record the status, protection status and information in plain text format, i.e. unencrypted. The latter is a potential security flaw, especially given the wider availability of NFC-capable smartphones (near field communication uses the same radio interface ISO 15963 as RFID), laying it open to risk of manipulation.[33]

An alternative form of electronic article surveillance using EAS bits has not yet seen adoption in library contexts. The EAS bit is not part of the ISO 18000 standard, leading to potential incompatibility problems between vendors of such RFID systems.

For recording the on-loan or in-stock status (item security status) of a book or other media item, three combinations of pairs of 8-bit hexadecimal AFI values are in use: initially 92_HEX (on loan) and 91_HEX (in stock), then from 1995 onwards, 9E_HEX and 9D_HEX, and in conjunction with the Danish data model (ISO 28560 Part 3), the combination 07_HEX and C2_HEX. The security gates of RFID systems scan for these three pairs of codes and respond with a visual and acoustic alarm signal should a media pass through it without having been checked out. It is also possible to identify the media passing through the gate using a procedure called gate tracking.

Compared with EM technology in which the magnetic strip can be hidden from casual view by placing it close to the spine, the typically credit card-size RFID tags have the disadvantage that they are plainly visible and can potentially be removed mechanically.[34] It is also possible to disable the security mechanism without great effort: aside from direct manipulation using NFC-capable smartphones, the RFID tag can be blocked from being read while passing through the gate by shielding it with metal packaging. This has not had an impact on the continuing adoption of RFID-based book and media security systems around the world. The full potential of the available possibilities has not yet been exploited; for example, many of the applications that currently use QR codes could be replaced in future by NFC-based applications. Such developments will depend on whether NFC technology becomes a standard feature of smartphones.

Conclusion

Specific aspects of RFID applications, such as the cost of implementation and the security systems, the legal context (privacy), physical criteria (building structure), quality standards and norms, the design of entrance areas (so that staff can intervene or gates coupled with turnstiles), as well as possible health risks must be assessed on an individual basis.[35]

In Germany procedures were developed for EAS systems in the early 1990s, and were defined in the Association of German Engineers VDI–4470 and VDI–4471 guidelines, but a basis for systematically comparing the performance of conventional EAS systems and RFID systems has only recently been outlined in VDI–4478. Performance criteria include the maximum number of RFID tags that can be detected at once as they pass through the gate, or the detection reliability at different speeds of passing through the gate, with different tag orientation and different carrier materials (e.g. metal and other materials that affect the resonating frequency). A systematic analysis of the respective strengths and weaknesses of both systems (EM or RFID) is still lacking and would provide a useful basis for libraries to assess their specific needs.

---

Originally published in: Nolan Lushington, Wolfgang Rudorf, Liliane Wong, Libraries: A Design Manual, Birkhäuser, 2016.