The Price Gilbert Memorial Library at Georgia Tech recently contracted to purchase an online circulation system.
The need for such a system came about for several reasons. Reorganization of public services areas made dispersal of circulation functions and remote access to circulation information desirable. Budget cutbacks created a need for detailed collection use information. Finally, a rapidly growing user population, due to borrowing agreements with other institutions, demanded closer control of borrower information. Although many of the library's operations were automated, the circulation department had been limited to a manual system. This system had sufficed because collection usage was primarily journal-oriented, and circulating materials were less frequently used. When circulation functions were expanded, however, the library decided that staff time, knowledge and expertise could be put to more creative use with a computer handling the additional routine administrative work.
The decision was made to automate, and the library began evaluating available systems. None met all of Tech's requirements, and indeed, some had obvious disadvantages. Most had costs outside the range of the endowment funds which the library intended to use for the project. Because the Georgia Tech library independently developed most of its other automated systems, internal development of a circulation system was briefly considered. However, with the library's Systems staff already overburdened by the re-design of the COM catalog as wewll as the maintenance of other systems, it was decided that another approach was needed. At this point an opportunity arose to cooperate with a software firm in solving Tech's circulation problems.
Sirsi Corporation, a software vendor based in Huntsville, Alabama, is headed by Jim Young, a former systems analyst at the Tech library. When the associate director of the library learned that Sirsi had begun planning a circulation system based on use of an OCR wand and independent workstations, it was seen as an opportunity for the library to participate in the development of an innovative system. Aware of Young's knowledge of library operations and systems, the Prince Gilbert Library agreed to collaborate with Sirsi to evolve a unique new circulation system.
Sirsi was contracted to design and implement the circulation system and to provide detailed documentation, most of the programming, and maintenance for a specified period. For its part, Georgia Tech selected and purchased the mainframe computer and provided detailed specifications of circulation and inventory control functions. This provided the library with a means of going beyond the limitations observed in available turnkey systems. The Tech library further opted to have its own Systems staff write software to transfer already available machine-readable bibliographic and patron records to the new circulation system. The staff was also designated to write programs to produce OCR-readable call number labels for the collection, to handle daily accounting, to provide statistical data, and to create overdue and other notices to users. This involvement was deemed advantageous, because it permitted the Systems staff to gain knowledge useful for system maintenance after the initial service agreement expired. It also reduced costs to the library.
The heads of the Circulation and Systems departments developed specifications for functions considered necessary in a circulation system. In doing so, they considered not only the Tech library's distinctive needs but the general requirements of all libraries, since Sirsi wished to make the finished system highly adaptable and marketable. All professionals in the library, as well as Circulation and Systems staff, were given the opportunity to review the specifications and make suggestions. Sirsi then complemented the specs with additional useful features and submitted a preliminary design to the library. During this period, Sirsi and the library were in close contact, and Tech was continually consulted as major design decisions were made.
The final design was delivered in June 1981 with detailed descriptions of how operators wowuld interact with the system to perform circulation functions.
It was approved by the library, and intensive programming began in Huntsville and Atlanta. After six months the system prototype was installed; the complete system, Unicorn, followed three months later.
Several capabilities distinguish Unicorn from other circulation systems. All standard functions are handled, including charges and discharges, renewals, holds, recalls, fine and fee accounting, notices to patrons, management and collection-use reports, temporary record creation, tracking items to successive locations, etc. However, Unicorn differs from most other available systems in five features: the use of OCR-readable call numbers rather than barcodes; an inventory control ability; an effective back-up circulation function; a highly evolved reserve book subsystem; and the flexible Poligen policy-setting mechanism.
Librarians know that nearly every item in a library has its own unique number: the call number. Tech and Sirsi decided to use these numbers instead of the ubiquitous barcode numbers since current technology allows for optical character recognition of letters, numerals, and punctuation. A program was written to extract call numbers from Tech's bibliographic database, reformat them, and print them in OCR type on individual spine labels. Relabeling the collection was facilitated by the fact that the labels were generated in shelving order. There was no need for a link between an arbitrary barcode number and the call number, as the call number was already an integral part of the bibliographic record stored in the circ system.
The inventory control aspect of the system seems particularly attractive since inventorying any extensive collection has become a costly and complex project. The OCR-readable spine labels will allow library staff to walk through the stacks with wands reading each label. The computer will evaluate the input, match “missing” numbers with those charged out, on reserve, etc., and finally will yield a detailed report of the number of volumes and titles in the library, will identify missing titles, etc. The inventory control system is ultimately expected to replace the shelf lists plans are underway to store holdings information for serials, government documents microforms, maps, as well as monographs. Currently, monographs alone compose the circulation data base. Unicorn's highly evolved reserve book subsystem is likewise an important advantage for Georgia Tech since it provides access not only by usual means—author, title, call number, ISSN, ISBN, etc.—but also by course number and name of the professor who placed the material on reserve. Hourly loan periods may be specified, and items are tracked even if they are temporarily removed from the reserve shelf to go to the bindery, for example. Sirsi has provided security against computer failure by means of intelligent workstations using floppy disks. Should the central computer go down, the diskettes are inserted into slots on the front of the terminals and will record each transaction. Several days' circulation transactions may be stored on the diskettes, and can later be loaded into the computer.
Unicorn allows flexibility in setting up and manipulating circulation policies by means of its Poligen feature, designed for use by Circulation management personnel. When the system was first brought on-line, Tech's head of circulation set policies such as loan periods and fine structures, and created user and item profiles. Workstations are assigned specific functions through Poligen. This capability is particularly important to Tech since some terminals are for public use and should only allow access to certain types of general information. Access to more detailed information is regulated by operator level with highly privileged information available only by password.
Georgia Tech purchased a TI 990 mini-computer on which to run the system. This decision was made partly because TI's are widely used in the Georgia University System libraries. While Unicorn does allow some flexibility in the choice of hardware, it does require specific workstations which are purchased through Sirsi. Although they look like ordinary CXRT terminals, they are actually microcomputers and are capable of independent functioning, allowing reliable backup in the event of computer downtime.
Sirsi also supplies OCR wands which are used to read call number labels and user ID cards. While this is the most efficient means of charging materials, user ID and call numbers can also be manually input, whether for circulation transactions or for querying the data base, and items and users can be searched in a variety of ways.
Although the concept of cooperation in the library community has come to mean agreements between libraries, the development of Unicorn is evidence that the cooperation with the private sector is also a viable means of stretching library resources.
