Chapter 31: Application 3 --- Library Systems (1990s) |
A Chapter from *The Silent Network: How RFID and Barcodes Together Map the Physical World* |
Summary |
The 1990s witnessed the unlikely transformation of the library from a quiet repository of barcoded volumes into a laboratory for some of the most ambitious civilian applications of radio frequency identification technology. While barcodes had served libraries adequately for decades, their fundamental limitation---the requirement of direct line-of-sight for every single item---created bottlenecks that grew increasingly intolerable as public demand for library services expanded. The National Library Board of Singapore, facing a mandate to triple its lending volume without tripling its staff, made a decision in the late 1990s that would reverberate through the global library community for decades: it would deploy RFID not merely as a security measure but as the foundational infrastructure for self-checkout, automated sorting, and inventory management. This chapter examines how that decision unfolded, why barcodes could not solve the problems Singapore faced, and how the library became one of the first civilian sectors to demonstrate RFID's transformative potential at scale. |
|
1. The Library Before RFID: A World of Lines and Light |
To understand why the events of the late 1990s mattered, one must first appreciate the technological constraints under which libraries operated for most of the twentieth century. The barcode, that ubiquitous pattern of black lines and white spaces, had become the standard method for identifying library materials by the 1970s and 1980s. It was cheap, reliable, and well-understood. A barcode label could be printed for fractions of a cent. A scanner could read it with near-perfect accuracy, provided the operator aligned the beam correctly. |
That last qualifier---*provided the operator aligned the beam correctly*---was the source of nearly every operational headache in library circulation. |
Barcodes are optical technology. They require light. They require a clear line of sight between the scanner and the printed symbol. They require the symbol to be at the right angle, at the right distance, and sufficiently clean to be decipherable. A barcode scanner cannot read through a book cover, cannot read a stack of books simultaneously, and cannot read a barcode that has been scuffed by years of handling or partially obscured by a due-date slip. Every transaction---every single checkout, every single return, every single inventory count---required a human being to perform the small but cumulative labor of locating the barcode, orienting it toward the scanner, and waiting for the confirming beep . |
For a library circulating a few thousand items a day, these constraints were manageable. Staff could process returns at a steady pace, develop muscle memory for the motion, and keep queues moving. But as library usage grew, the barcode's limitations became structural bottlenecks. The check-in process alone---scanning each returned item, desensitizing its electromagnetic security strip, and updating the patron record---consumed staff hours that could have been devoted to more meaningful work. The check-out process, equally line-of-sight-dependent, created queues at service desks that lengthened during peak periods. And inventory, the periodic task of verifying that every item in the collection was where it was supposed to be, was so laborious that many libraries performed it only rarely, accepting that their catalog records would gradually drift out of sync with the physical reality of their shelves . |
The electromagnetic security strip---often called a Tattle-Tape after the 3M product that dominated the market---introduced its own complications. These thin metallic strips, embedded in book spines or affixed to media cases, would trigger an alarm if an item passed through a security gate without being properly desensitized at checkout. But the strip had to be manually desensitized and re-sensitized during every transaction, adding yet another step to the circulation workflow. And the strip was purely passive: it could sound an alarm, but it could not identify *which* item was setting off the alarm. A patron who walked through the gates with a book that had failed to desensitize properly would trigger a generic beep, forcing staff to conduct a manual inspection of the patron's belongings . |
This was the technological status quo that the National Library Board of Singapore confronted in the mid-1990s. |
|
2. The Singapore Problem: Scaling Without Staffing |
The National Library Board (NLB) of Singapore occupies a distinctive place in the history of library technology. Singapore is a small nation with a large ambition: to build a world-class library system that would serve as an engine of national learning and competitiveness. In the 1990s, the NLB embarked on an ambitious program of business process re-engineering, driven by a recognition that the existing library model could not meet the demands of a rapidly developing economy . |
The numbers told a stark story. With 5.7 million visitors a year, the NLB's libraries faced queues that could stretch to an hour during peak times---one hour to return a book, another hour to borrow one. The NLB's target for 2004 was thirty million loans annually, a threefold increase over 1995 levels. To achieve that target with the existing barcode-based workflow would have required hiring a small army of additional staff. The NLB had neither the budget for such hiring nor the desire to create a workforce that would be idle during non-peak periods . |
The 'time-to-check-out, time-to-shelf' project became the focal point of the NLB's re-engineering efforts. The goal was simple to state and extraordinarily difficult to achieve: reduce the friction of every circulation transaction to the point where the library system could triple its throughput without proportionally increasing its labor costs. The existing barcode technology could not deliver this. The NLB needed something fundamentally different. |
The answer, when it came, was RFID. |
|
3. The Bukit Batok Experiment: A World First |
In November 1998, the NLB opened a new branch library at Bukit Batok, a residential town in the western part of Singapore. This unassuming community library would become the site of one of the most significant experiments in library technology history. Working with a Singapore-based technology partner, ST Logitrack, the NLB deployed a fully integrated RFID system that handled borrowing, returning, sorting, and stocktaking---the first such deployment anywhere in the world . |
The system, eventually branded as the Electronic Library Management System (ELiMS), represented a radical departure from the barcode paradigm. Each library item was fitted with an RFID tag---a small microchip with an antenna that could communicate wirelessly with a reader. Unlike a barcode, which stores only a reference number that must be looked up in a database, the RFID tag could store information directly and could be read without line-of-sight, through book covers, and in multiple-item batches . |
The self-checkout stations at Bukit Batok worked on a principle that would become familiar to library users worldwide over the following two decades. A patron would place their library card and a stack of books on a reading pad. The system would automatically identify the patron and each item, updating the loan records in a single operation. No scanning. No alignment. No waiting for the beep after each individual barcode. The entire transaction took seconds . |
For returns, the system was even more transformative. Patrons could deposit books into a bookdrop, where an RFID reader would automatically detect the items and update their status. The loan was cancelled immediately, with no staff intervention required. The 'zero wait time' that the NLB emphasized in its internal documentation was not merely a marketing phrase; it was a genuine operational reality. Patrons could borrow a book the moment it was returned, rather than waiting for staff to process the return . |
The system also automated the sorting process. As returns were processed, the ELiMS could direct items to the correct bin based on their category or destination branch. Staff who had previously spent hours sorting returns by hand could now focus on other tasks. The NLB later reported that training part-time staff to shelve books correctly became dramatically easier, because the system itself guided the sorting process . |
The success of the Bukit Batok trial led to a broader rollout. The system was piloted at Toa Payoh Community Library in May 1999 and at library@orchard in September 1999. In August 2000, the NLB signed a contract with ST Logitrack to install the ELiMS system across all NLB libraries. In December 2000, the NLB and ST Logitrack were jointly awarded an international patent for the system . |
The financial impact was substantial. The NLB later estimated that the RFID system had saved more than twenty-eight million US dollars in staffing costs. Library users, meanwhile, saved an estimated 5.6 million hours per year in queuing time---hours that could be devoted to reading, learning, or any of the other activities that libraries exist to support . |
|
4. Why Barcodes Could Not Do This |
The Singapore experiment succeeded because RFID addressed the fundamental constraints of barcode technology. Understanding those constraints---and the specific ways RFID overcame them---illuminates why the library became such an important proving ground for RFID. |
Line-of-sight. This is the most obvious and most consequential difference. A barcode reader is essentially a specialized camera. It must 'see' the barcode to decode it. An RFID reader, by contrast, communicates via radio waves, which can penetrate non-conductive materials like paper, plastic, and cloth. A stack of books on an RFID reading pad can be processed simultaneously, while a stack of barcoded books would require each volume to be individually scanned . |
Speed. The line-of-sight requirement translates directly into time. Scanning a barcode takes a few seconds per item, including the physical motion of locating and orienting the symbol. RFID reading is effectively instantaneous for multiple items. One academic study found that RFID-based circulation reduced per-item check-in time from forty-two seconds to eight seconds---a reduction of more than eighty percent . For a library processing thousands of items per day, these seconds accumulate into hours of staff time and patron waiting. |
Read/write capability. A barcode is a static identifier. Once printed, it cannot be changed. An RFID tag can be read and rewritten, allowing the tag to store not just an identifier but also security status, location codes, and other metadata. This read/write capability is what enabled the NLB's system to automatically update security status during checkout and return---eliminating the separate step of desensitizing an electromagnetic strip . |
Security integration. In a barcode-based library, security and identification are separate systems. The barcode identifies the item; the electromagnetic strip secures it. RFID can combine both functions in a single tag. When an item is checked out, the RFID tag's security bit is deactivated; when it is returned, the bit is reactivated. This integration reduces the number of steps in every transaction and eliminates the failure modes that arise when the two systems fall out of sync . |
Inventory and shelf management. Perhaps the most laborious task in a barcode library is inventory. Staff must walk the shelves, remove each book, scan its barcode, and replace it---a process that can take weeks for a large collection. RFID enables a fundamentally different approach. A handheld reader can scan an entire shelf at once, identifying not only which items are present but also which items are missing or misshelved. What once took weeks can be accomplished in a single day . |
These advantages were not merely incremental improvements. They represented a qualitative shift in what a library could do. The NLB's ability to triple its lending volume without tripling its staff was a direct consequence of this shift. |
|
5. The Global Diffusion of Library RFID |
The Singapore experiment was not immediately replicated elsewhere. In the late 1990s, RFID tags cost approximately one dollar each---prohibitively expensive for libraries with collections numbering in the hundreds of thousands or millions of volumes. The first library RFID installation in the United States occurred in 1999, the same year as the NLB's broader rollout, but adoption remained slow. By 2001, only about fifty library installations existed worldwide, compared to approximately five hundred thousand warehouse and retail installations. The technology was proven, but the economics were challenging . |
The turning point came with the dramatic reduction in tag prices over the following decade. Economies of scale in manufacturing, driven largely by RFID's adoption in supply chain and retail applications, drove tag costs down by as much as eighty percent. By the mid-2000s, tags could be purchased for a quarter or less, making RFID economically viable for a much broader range of libraries. By 2005, an estimated three hundred libraries worldwide had adopted RFID, with approximately 120 million tagged items. By 2010, a survey of UK libraries found that 116 organizations had deployed RFID to some extent---a 414 percent increase year over year . |
The diffusion followed a predictable pattern. Academic libraries, with their larger budgets and more technologically sophisticated staff, were early adopters. The University of Nevada, Las Vegas, served as a beta tester for 3M's library RFID system, eventually tagging over one million items in its Lied Library. Public libraries, facing greater budget constraints, followed more slowly, though the pace accelerated as tag prices fell and the success stories multiplied . |
Standardization played a crucial role in accelerating adoption. Early RFID deployments used proprietary systems that locked libraries into a single vendor. Beginning in the mid-2000s, industry standards for tag manufacturing and data encoding began to emerge, enabling interoperability between different vendors' tags and equipment. The NLB itself migrated its collection to ISO 15693-compliant tags in 2004, a move that signaled the maturation of the technology . Standardization lowered another barrier to entry by giving libraries confidence that their investment would not become obsolete. |
By the 2010s, RFID had become the dominant technology for library circulation management in much of the developed world. A library that opened without RFID self-checkout would have seemed as anachronistic as one without an online catalog. |
|
6. The Anatomy of a Library RFID System |
The system that the NLB pioneered in the late 1990s has evolved considerably, but its core architecture remains recognizable. Understanding the components provides insight into how the silent network of RFID readers and tags maps the physical space of a library. |
The RFID tag. Each library item---book, DVD, CD, or other material---carries a small RFID tag, typically embedded in the spine or affixed to an inside cover. The tag contains a microchip and an antenna. The microchip stores a unique identifier for the item, along with a security bit that can be toggled to indicate whether the item is checked out. Some tags also store a location code, allowing the system to track which branch or section of the library the item belongs to. The tag is passive, meaning it has no battery; it draws power from the radio signal emitted by the reader . |
The self-checkout station. These patron-facing units typically consist of a touchscreen display and an RFID reading pad. The patron places their library card and items on the pad; the system reads all tags simultaneously, identifies the patron, updates the loan records, and prints or emails a receipt. The absence of a line-of-sight requirement means that items can be stacked or placed at any orientation. Some modern systems can process multiple items in a single transaction, further reducing the time required . |
The automated return and sorting system. Bookdrops equipped with RFID readers allow patrons to return items at any time, with immediate account updates. In larger libraries, returned items travel along conveyor systems where RFID readers identify each item and route it to the appropriate sorting bin---by category, by destination branch, or by whether the item has a hold request. These automated materials handling systems can process thousands of items per hour, dramatically reducing the labor required for returns processing . |
The security gate. RFID security gates at library entrances detect tags that have not been properly checked out. Unlike electromagnetic strips, which trigger a generic alarm, RFID gates can identify the specific item that triggered the alert, allowing staff to resolve the issue more efficiently. Modern gates are designed to accommodate wide aisles and wheelchair access while maintaining detection accuracy . |
The handheld inventory reader. This portable device allows staff to walk through the stacks at normal speed while scanning shelves for inventory, misshelved items, and items flagged for weeding. A single pass can capture the status of thousands of items, transforming a task that once took weeks into one that can be completed in a day . |
|
7. Beyond Singapore: Applications Across Library Types |
The Singapore experiment demonstrated what was possible. The decades that followed saw libraries of every type adapt RFID to their specific needs. |
Public libraries. The most visible application of library RFID remains self-checkout. For public libraries, which often face unpredictable peaks in demand---after school, on weekends, during summer reading programs---the ability to process multiple items simultaneously and without staff assistance has been transformative. A parent with a stack of picture books can check out in seconds rather than minutes. A teenager returning a bag of graphic novels can deposit them in a bookdrop without waiting in line. These small conveniences accumulate into a fundamentally different patron experience. The NLB's own evolution reflects this: after pioneering RFID self-checkout, the NLB added mobile app-based borrowing in 2014, allowing patrons to use their smartphones to scan barcodes and check out items directly, building on the infrastructure and patron expectations that RFID had established . |
Academic libraries. University libraries face distinctive challenges: large collections, seasonal surges in demand, and a user population that expects seamless digital experiences. RFID has enabled academic libraries to handle semester-start rushes with greater ease, to offer extended-hours access through remote locker systems, and to conduct frequent inventory sweeps that keep catalog records accurate. At Hong Kong Metropolitan University, the library migrated its entire collection to RFID specifically to enable mobile self-checkout, allowing students to borrow materials using their phones via NFC or camera scanning. The library also implemented fast inventory sweeps using handheld wands, a capability that would have been impractical with barcodes . At Augsburg University's Lindell Library, the installation of remote locker systems for hold pickup helped drive a five percent annual increase in circulation, with self-service accounting for ninety-eight percent of all transactions . |
School libraries. Even smaller school libraries have adopted RFID, often integrating library borrowing with student ID cards and cafeteria payment systems. In the Philippines, some schools have deployed RFID-enabled ID cards that serve multiple functions: borrowing books, purchasing items at the school shop, and tracking attendance. The library becomes one node in a broader network of identification and transaction . |
Special and corporate libraries. Organizations with specialized collections---law firms, medical libraries, research institutes---have adopted RFID for the same reasons as public and academic libraries: efficiency, accuracy, and the ability to track materials that may circulate among a distributed workforce. The smaller scale of these libraries means that RFID adoption is often simpler, but the benefits are proportionally similar. |
|
8. The Barcode's Enduring Role |
The rise of RFID in libraries did not eliminate barcodes. In most libraries, the two technologies coexist in a hybrid arrangement that leverages the strengths of each. |
Barcodes remain the universal identifier. Every library item, even those with RFID tags, typically bears a barcode as well. The barcode serves as the human-readable and machine-readable link to the library's catalog system. If an RFID tag fails---and tags do occasionally fail---the barcode provides a fallback. Many library management systems still use barcodes as the primary identifier, with RFID tags storing the barcode number as their core data element. This arrangement means that a library can adopt RFID incrementally, tagging items as they circulate rather than requiring a complete collection conversion overnight . |
Barcodes also remain important for interoperability. A library that lends items to other libraries, or that participates in consortia, cannot assume that all partners use the same RFID standards. The barcode is a lingua franca that transcends vendor and technology boundaries. |
The hybrid model extends to security as well. While RFID tags can carry security bits, some libraries continue to use electromagnetic strips alongside RFID, either as a backup or because their existing security gates are not RFID-compatible. The transition from electromagnetic to RFID security has been gradual, driven by the replacement cycle of security gates and the desire for a unified identification and security system . |
This coexistence is not a failure of RFID adoption but a recognition that different technologies serve different purposes. The barcode is a low-cost, universally readable identifier. RFID is a higher-cost, higher-capability system for automation and tracking. A well-designed library uses both. |
|
9. The Measurable Impact |
The case for library RFID rests not on technological novelty but on measurable outcomes. The numbers from libraries that have adopted the technology tell a consistent story. |
The NLB's experience is the most extensively documented. The system saved more than twenty-eight million US dollars in staffing costs. Patrons saved an estimated 5.6 million hours per year in queuing time. The library was able to pursue its ambitious lending targets without proportional increases in staff . |
A study of a university library's RFID deployment found that per-item check-in time dropped from forty-two seconds to eight seconds---an eighty-one percent reduction. For a library processing hundreds of items per day, this translates into hours of staff time reclaimed for other purposes . |
Inventory, as noted, is transformed. What once took weeks of tedious scanning can be accomplished in a single day with a handheld RFID reader. The accuracy of the catalog improves, the visibility of misshelved items increases, and staff can redirect their time to patron-facing services . |
The patron experience improvements are harder to quantify but no less real. Self-checkout reduces queues. Automated returns eliminate waits. The ability to borrow a stack of books with a single action, rather than scanning each one individually, is a small convenience that, repeated across millions of transactions, becomes a meaningful improvement in the library's usability. |
These outcomes explain why, despite the higher initial cost of RFID compared to barcodes, libraries have continued to adopt the technology. The return on investment comes not from the tags themselves but from the labor savings, the accuracy improvements, and the enhanced patron experience that RFID enables. |
|
10. Conclusion: The Library as Proving Ground |
The library RFID deployments of the 1990s, and particularly the National Library Board of Singapore's pioneering work, occupy a distinctive place in the history of the silent network. They were not the largest RFID installations by any measure---retail and logistics operations dwarfed libraries in scale. They were not the most technologically sophisticated---the tags and readers were, by later standards, rudimentary. But they were among the first civilian applications to demonstrate that RFID could transform not just efficiency but the fundamental experience of a public institution. |
The NLB's decision to deploy RFID was driven by a specific, urgent problem: how to triple library usage without tripling staff. The solution it developed, in collaboration with ST Logitrack, became a template that libraries around the world would follow. The self-checkout station, the automated bookdrop, the sorting system, the handheld inventory reader---these components, first assembled at Bukit Batok in 1998, became the standard architecture of the modern RFID library . |
The library also provided an early answer to a question that would become increasingly important as RFID spread: how to balance efficiency gains against privacy concerns. Library RFID tags store only item identifiers, not patron information. The link between an item and a patron exists only in the library's database, and it is broken when the item is returned. This design principle, articulated by the American Library Association and adopted by libraries worldwide, established a model for responsible RFID deployment that other sectors would later emulate . |
The 1990s were a decade of experimentation for RFID in libraries. The technology was expensive, the standards were immature, and the outcomes were uncertain. But the Singapore experiment demonstrated that the potential was real. The decades that followed saw that potential realized across libraries of every type and size, as tag prices fell, standards matured, and the benefits of RFID became too compelling to ignore. The quiet library, it turned out, was an ideal place to prove that a silent network of radio waves could map and manage the physical world. |
|
Detailed Summary |
This chapter has examined the deployment of RFID technology in library systems during the 1990s, focusing on the pioneering work of the National Library Board of Singapore at the Bukit Batok Community Library in 1998. The chapter traced the limitations of barcode technology---line-of-sight requirements, per-item scanning, separation of identification and security functions---that created operational bottlenecks as library usage grew. It then detailed how the NLB, facing a mandate to triple lending volume without proportional staff increases, partnered with ST Logitrack to develop the Electronic Library Management System, the world's first fully integrated RFID system for library borrowing, returning, sorting, and stocktaking. |
The chapter examined the technical advantages of RFID over barcodes in the library context: simultaneous reading of multiple items without line-of-sight, read/write capability that enabled integrated security and identification, and the ability to conduct rapid inventory sweeps with handheld readers. It described the measurable outcomes of the Singapore deployment---over twenty-eight million dollars in staffing savings and 5.6 million patron-hours saved annually---and traced the subsequent diffusion of library RFID globally as tag prices fell and standards matured. |
The chapter also explored applications across library types: public libraries using RFID for self-checkout and automated returns, academic libraries deploying mobile self-checkout and remote locker systems, school libraries integrating RFID with student ID cards, and special libraries adopting RFID for specialized collection management. It noted the continuing coexistence of barcodes and RFID in hybrid systems, with barcodes serving as universal identifiers and fallback options. |
Finally, the chapter considered the broader significance of the library as a proving ground for RFID. The library demonstrated that RFID could transform not just operational efficiency but the fundamental patron experience of a public institution. It also established early models for privacy protection that would influence RFID deployment in other sectors. The quiet library, the chapter concludes, proved that a silent network of radio waves could map and manage the physical world. |
cs@easiersoft.com