I see libraries in a time where it's important to question all of the assumptions that we have held regarding technology and the automation products in which we invest. Many of the current automation scenarios were informed by inherent limitations of computing of times past and models of functionality and processing workflows that may no longer apply. Some of the key issues in play today deal with the scale in which technology can be implemented to support libraries or groups of libraries and how different areas of functionality are organized within automation products.
We're in a time when the scale of computing far exceeds the bounds of times past. In this era of cloud computing, platforms can be assembled at a global scale. In the commercial realm, we see all sorts of services that support millions - if not billions - of users with limitless capacity for handling images, video, text, or practically any other type of data imaginable. These high-capacity services rely on thousands of computers clustered together in data centers to handle enormous transaction loads with seemingly unlimited data storage capacity and with data centers distributed around the globe to support customers in each geographic region. Facebook, Google, Amazon, Flickr, and the like could not exist apart from the limitless scale of computing power that can be amassed today.
Libraries, on the other hand, mostly continue to rely on comparatively smallscale computing. The vast majority of automation implementations involve software installed on a single computing platform serving a single library. These implementations are based on an instance of the software configured for that library and an isolated database representing that library's collection. At only a slightly larger scale, many libraries join together in consortia to share an automation system. Library consortia, regional library systems, or other organizational constructs allow libraries to share hardware and software platforms and offer larger collective collections to their patrons while reducing technology costs. A consortium may involve dozens, or even hundreds, of libraries but still operates at a small scale relative to the global commercial services.
Rethinking the Possibilities
The automation scenarios that we see in place today were shaped largely by the constraints of computing platforms of previous eras of computing. There was a time when the automation of a single large library could stretch the limits of available hardware and the sophistication of software. While smaller libraries could elect to band together to share a system, the number of members was limited by constraints imposed by current hardware and software capacities.
Roll the clock forward to today. Given what happens on the commercial web, we know that computing platforms can scale infinite, but library software often continues to impose some constraints. Products designed to support individual libraries or consortia of finite numbers of libraries may not necessarily be able to handle the complexities of very largescale implementations. But as library software evolves - or is newly created - in Une with the models of multitenant software-as-a-service products routinely deployed in the commercial sphere, these constraints begin to fall away. As the realm of library software enters the infinitely scalable realm of modern cloud computing, the opportunity opens to reassess some of the fundamental patterns in which libraries individually or collectively automate.
Cloud Computing for Library Automation
Part of the mix seen in this new phase of library technologies involves some systems built anew following the basic premises of cloud computing. Some of the automation products in this category would include the OCLC WorldShare Platform, Ex Libris Alma, and Intota in development by Serials Solutions, as well as several resource discovery services and other general-purpose and nbrary-specific applications.
OCLC takes an expansive view of collaboration through its WorldShare Platform. Applications such as WorldShare Management Services allow libraries to participate in a global automation environment modeled after the commercial web destinations. This environment provides automation services to libraries, allocating computing resources needed, but more importantly through sharing global database structures. The basic approach involves tagging holdings and items onto WorldCat bibliographic records rather than the traditional model of having each library maintain its own isolated database of bibliographic, holding, and item records. WorldShare embodies the modern multitenant cloudcomputing approach.
Alma from Ex Libris likewise follows a similar cloud-based approach, though with some additional nuances. Rather than entirely relying on a shared data model, Alma follows a hybrid approach, allowing libraries to share data globally when appropriate through what is called the Global Zone, as well as to maintain institution-specific data held in a Local Zone. All the data resides on a cloud platform, including the local zone.
Intota, currently under development by Serials Solutions, likewise follows the multitenant software-as-aservice model. Based on a new knowledgebase that Serials Solutions is creating, Intota will provide a cloud-based platform that allows libraries to manage their print, electronic, and digital resources based on a data model that allows each library to maintain its own bibliographic database, but within the hosted environment and sharing one global instance of the software.
The WorldShare Platform, Ex Libris Alma, and Serials Solutions Intota each allow libraries to take advantage of a global cloud-based platform, either as individual libraries or organized in consortial groupings. Each of these products takes a somewhat different approach in how automation functionality is organized, how bibliographic and other data resources are shared, and how it handles individual or collective groupings of libraries. But each has been designed both in terms of hardware resources and software architecture to scale to support very large numbers of libraries. The number of libraries currently implementing these systems is relatively modest, so this scalability is more potential than actual.
Libraries Consolidate for Shared Automation
Apart from the realm of systems built anew under the models of cloud technologies, the realm of traditional integrated library systems (ILSs) continues to see ever-larger implementations. While not necessarily engineered for global deployments, today's hardware allows even traditional software to perform faster and with higher capacity. Speed of processing and storage limitations have increased to the point that an instance of an ILS can support incredibly large groups of libraries.
Last year, for example, the Illinois Heartland Library System, an amalgamation of four antecedent regional libraries that each previously operated its own separate ILS, set out on a strategy to automate 450 of its library members with a single automation system. This organization, as far as I have been able to determine, constitutes the largest consortium in the U.S. and will implement the Polaris ILS. I find this project very interesting in that it reinforces a trend toward larger-scale library automation implementations.
The Orbis Cascade Alliance provides a similar example in the academic library arena. This organization of 37 academic libraries was formed through the merger of two previously separate consortia in 2003, and additional organizations have joined subsequently. Each institution currently operates its own ILS, and different technologies have been used over time to enable the libraries to share their collections. The most recent strategy, driven by an interest to press the limits of cooperation, involves shifting from individual ILS implementations to all sharing the same automation system. Following an extensive evaluation and selection process, the Orbis Cascade Alliance announced last fall that it would collectively implement Ex Libris Alma.
Shared State or National Infrastructure
Beyond individual libraries organizing together in consortia to use a common automation environment, we are also seeing more examples of state or national governments opting to implement a shared automation infrastructure for all the libraries within their jurisdiction. Some of the places that have had projects in place for a while include Iceland, where all of its libraries, including public, school, special, and academic, share a single instance of Ex Libris Aleph; all the public libraries in Chile are in the process of implementing an automation environment that consists of Aleph and VuFind; and the state of South Australia is executing a strategy of a statewide library management system and has selected SirsiDynix Symphony. All of the public libraries of Northern Ireland in the U.K. are automated using the OpenGalaxy library management system from Axiell Ltd. and have recently amalgamated from four regional systems into one.
The country of Denmark is working toward national infrastructure to support its public libraries. There are currently about 90 separate automation implementations spanning more than 400 libraries. A procurement process is now under way to select a system that will provide a nationally shared technology infrastructure for these libraries.
I'm not necessarily suggesting that bigger is always better, that libraries should necessarily strive to form the largest possible shared automation implementations. Rather, those wanting to explore larger-scale collaborative projects should be aware that the technical constraints that may have shaped their strategies in the past have changed radically and a new range of possibilities is open for consideration. Organizational concerns for some libraries may continue to demand individualized automation, but the possibilities for broader collaborative relationships are almost limitless.