College and Research Libraries KAREN HORNY Automation of Technical Services: Northwestern's Experience Library automation requires significant adjustments in attitudes and procedures to work effectively. When Northwestern University Library made a commitment to develop computer support of the entire tech- nical processing system, a chain of basic decisions was necessary, each with an individual impact on staff outlook and system productivity. Key decisions involved reassessing manual operations prior to automa- tion; defining resources-human and financial; determining on-line and batch processing requirements, hardware usage, and necessary modifications; considering essential staff reorganization and retrain- ing; and assuring communication. These decisions, the resulting on- line interactive system, and its impact are described. wHAT DOES A UNIVERSITY LIBRARY EN- COUNTER when it makes a commitment - to develop computer support of the en- tire technical processing system? North- western's experience began with the need for a chain of basic decisions in- volving determination of the potential for basic operational improvements, identification of available hardware capabilities and possibilities for modifi- cation, and examination of financial implications within the university as a whole and directly related to library pri- orities. Total library systems analysis, re- assessing existing manual operations, and identifying possibilities for staff reor- ganization became essential. A primary aim of any automated sys- tem is to reduce per unit costs for acqui- sitions and cataloging while allowing the library to handle more material without a proportional increase in staff size in years to come. Hardware costs Karen Horny is assistant university li- brarian for technical services, N orthwestem University, Evanston, Illinois. 364 I are decreasing as personnel expenses soar. Northwestern University's opera- tional system owes its existence to uni- versity and library administrative recog- nition of this essential fact. BACKGROUND As libraries have become rather pain- fully aware, automation does not arrive instantaneously. Reaching the imple- mentation phase requires substantial ad- vance planning. At Northwestern, uni- versity administrative interest was given added impetus by a librarian-faculty- student Library Planning Committee which served as an advisory group dur- ing the design and construction phases of a new library building. In the early 1960s this group projected the develop- ment of an automated library support system and indicated the strength of its confidence by providing a special room for a "dedicated" library computer as the blueprints developed. This special room, which was not removed from the plans by the eventual decision to share a computer already available elsewhere Automation of Technical Services I 365 on campus, has proved most useful as a typewriter terminal center for the li- brary project. As planning progressed, rec.ognition that a "dedicated" library computer was financially impractical resulted in a deci- sion to utilize the hardware at the Ad- ministrative Data Processing Center, which already maintained payroll, stu- dent registration, and alumni records. This machine has text handling capabil- ities which are superior to those of the larger computer at the scientific research computing center, and the library also was assured priority in usage. Unlike many library systems which share a com- puter with other campus applications, Northwestern University Library has al- ways had sufficient computer time for both operation and development. Al- though the equipment limitations of the very small capacity 96K IBM 370/ 135 computer cause occasional irrita- tion, the use of such a machine keeps computer costs down to the bedrock lev- el the library can afford. An important early decision was to develop the system in an on-line mode. Unlike batch processing, the on-line in- teractive mode provides for immediate access to and modification of informa- tion in the computer file and prevents the automated system from becoming as unwieldy as the previous manual sys- tem. One unified computer record re- places a multiplicity of order and pro- cessing control slips formerly filed in separated sequences. Printed purchase orders, claim forms, worksheets for catalogers, catalog cards, pocket labels, circulation charge cards, and various in- dexes are batch produced; but inquiry and update operations for some 100,000 in-process and recently cataloged items are all on-line. Order, receipt, payment, location, call number, and internal pro- cessing data are immediately entered in the computer record via direct terminal input. THE TECHNICAL SERVICES SYSTEM The technical services system, imple- mented October 4, 1971, after nearly two years' successful operation of the computerized circulation system, 1 cur- rently utilizes nine IBM 27 40 typewriter communication terminals linked to the computer via two telephone lines. This is a total system in that it is designed to cover every aspect of acquisition and processing. Library computer records in- clude all monographs on order, in pro- cess, and recently cataloged; biblio- graphic and holdings information for some 35,000 current and noncurrent se- rial titles held by the main and branch science libraries; and the circulation file. These records were originally stored on data cell, long the most inexpensive form of computer storage, but after ap- proximately one year of library system operation disc packs were installed to in- crease general operating efficiency. At the project's inception the computer had no tape drives, a situation which neces- sitated transfer of MARC records from tape to disc at another institution. This arrangement was most inconvenient and often resulted in considerable delays. Both the disc pack and tape drive hard- ware alterations were instigated because of library computer usage. The small size of the computer has challenged the impressive talents and in- genuity of the programmer, while the constraint of limited core and disc stor- age available for the teleprocessing pro- gram has kept on-line data validation and index development to a minimum. Limited file capacity has also resulted in the necessity to ''dtifilp" data onto historical file tapes for monographs cat- aloged at least nine months and without . added copy orders. Operationally, the most significant current equipment limi- tation is the display rate (only 14 char- acters per second) of the typewriter ter- minals, which is far from ideal when long bibliographic records are involved. 366 I College & Research Libraries • September 1974 The use of cathode ray tube (CRT) terminals, now in a testing phase, will greatly speed inquiry displays and be particularly advantageous for the indi- vidual issue check-in operation for se- rials. STAFF RELATIONSIDPS In the development of a viable com- puter support design, the importance of a strong, direct relationship between li- brarians and systems staff is impossible to overemphasize. More than five years prior to automation implementation a systems analyst was officially appointed to the library staff, a decision which al- ,lowed the development of thorough familiarity with both the theories and specific applications of library opera- tions. The mutual confidence which evolved was a significant factor in facil- itating the massive retraining project necessitated by systems implementation. The library also gained a unique and inestimable benefit by interesting a member of the university's Computer Sciences and Electrical Engineering Fac- ulty in the automation project. On an official half-time appointment to the li- brary staff, he singlehandedly designed and programmed the teleprocessing sys- tem and wrote most of the assembly lan- guage batch programs which allow the best possible utilization of the limited amount of computer memory available to the library. The official library staff appointments of both systems analyst and faculty computer expert have been vitally important in insuring essential communication, developing designs tai- lored specifically to meet the library's sometimes rather peculiar needs, and al- lowing quick action when problems are detected. One of the most important aspects of the automated system implemented by Northwestern is that it was put into full operation without the addition of any extra line staff. This proved to be a con- siderable accomplishment since techni- cal services had been demonstrably un- derstaffed in the most recent years of manual operations (based on the Uni- versity of Washington formula in Model Budget Analysis System for Li- braries, 1970, as well as in-house calcu- lations). The decision to add no new staff to aid in the transition was partial- ly a result of the tight financial restric- tions under which the project must op- erate. Northwestern's automation pro- gram has developed as a "shoestring" operation with a need to justify its ex- istence and methods to the university administration, which provides its sole support. No grant funds have aided the development. This was mainly due to a general feeling that there would be less trouble suiting Northwestern's indi- vidual needs if the library faced the realities of university funding from the outset. Otherwise, it might plunge into a potentially overambitious project using special funds which would even- tually run out and leave operations in a position where a retrenchment might be necessary. Since new staff were not added, it was necessary to redefine or reallocate cer- tain positions within the technical ser- vices division. Interdepartmental proce- dural coordination has been ably man- aged by a professional also serving as head of the Search Department. A va- cancy from the Catalog Department provided the nonprofessional supervisor for a newly established Data Center where the six 27 40 terminals for mono- graph . processing are located (the room in the new building originally intended for a library computer). Terminal op- erators from the general staff have been assigned in short shifts, returning to their regular ·departments with a broad- er understanding of the total system. This arrangement provides variety in job assignments and helps sustain the peak efficiency which is difficult to main- tain in any long-term typing assignment. It also has been important to avoid ere- ' ' ,. [ Automation of Technical Services I 361 ating an "elite" group of Data Center staff. Another advantage of using gen- eral library personnel as terminal opera- tors lies in their grasp of the format and importance of bibliographic data. Since these staff members work with the same data as it applies to their regular assignments in the Search, Order, Cata- log, and Bindery & Marking depart- ments, they are less prone to make mis- takes at the terminal than new typists without such additional experience. Un- like administrative keypunch operators who are trained to operate their ma- chines as a primary objective and under- stand the material only secondarily, the library terminal operators have been trained to deal with the bibliographic data first and then given part-time ter- minal typing assignments. PREPARATION A nwnber of procedural and organi- zational changes in manual operations resulted from anticipatory planning during the years immediately preceding implementation of the ·automated sys- tem. Preorder and precatalog searching were consolidated in a special Search Department. Cataloging with LC copy was organized into a Fast Cataloging Unit staffed by nonprofessionals under the supervision of a professional cata- loger. The Acquisitions Department was split into monographic Order and Se- rials departments. These staff organiza- tional patterns prevented a number of potential bottlenecks during system im- plementation and lent themselves ef- fectively to group retraining. After the inauguration of the automated system, the need to face several other long- standing problem areas became clear, and two additional specialized units were organized. The first deals with problems in correctly identifying and handling added copies. In the second, serial bindery preparation procedures have been reorganized. Automation sometimes makes a library face up to problems which previously could be ig- nored but were significant nonetheless. Perhaps the most difficult problems encountered involved transitional pro- cedures rather than equipment and pro- gram limitations. Since the library im- plemented a total system, all aspects of the technical services operations were affected simultaneously. The change- over was not done on a "piecemeal" ba- sis, and everyone became immediately involved. Misconceptions about various aspects of the system were rampant, in spite of an extensive informational campaign which began during the sum- mer preceding implementation and con- tinued intensively throughout the first year. The strain of simultaneously deal- ing with materials still tied into the manual system while inaugurating new methods of handling items ordered via the automated system, shifting respon- sibilities among individuals and depart- ments, giving up treasured files, and get- ting used to looking at the same infor- mation in a new format (e.g., log and work sheets instead of LC card copy) proved a major challenge. A decision to emphasize communica- tion encouraged cooperation, promoted special efforts, and reduced anxieties. General orientation sessions, progress re- ports, and demonstrations, as well as specialized training, occupied a consid- erable amount of staff time but amply repaid the investment in boosted morale both within the technical services divi- sion and in other areas of the library. A general meeting for all library staff was held shortly before the time of im- plementation, followed up by a staff an- nouncement briefly describing the ex- tent of the project and requesting pa- tience and cooperation during the tran- sitional period. During the first three months of operation an almost continu- ous string of information-discussion- problem-solving meetings was held. Demonstration sessions were arranged to allow divisional staff not normally 368 I College & Research Libraries • September 1974 scheduled at the computer terminals to receive basic instruction and have a chance to perform some actual opera- tions at the keyboard. These sessions preceded a slide-illustrated description of the automated system presented for the whole staff under the aegis of the Library Staff Association. In turn this was followed by another set of terminal demonstrations for nontechnical ser- vices personnel. Although some of the detailed slide narrative presentation pre- sumed a fairly clear understanding of technical processing operations under any system (manual or automated) and therefore was not completely compre- hensible to every staff member, the gen- eral reaction was very positive, indicat- ing intense interest. The slightly threat- ening mystery disappeared. ADJUSTMENT During the early adjustment period, lasting approximately three months, oc- casional negative remarks came from nontechnically oriented staff impatient with the slow-up in ordering and pro- cessing evident while training was in progress. Much material still arrived which had been ordered under the old manual system, and Thanksgiving and Christmas holidays intervened. By Jan- uary 1972, when out-going firm orders had reached a normal level (over 2,500 per month), objections diminished. One phenomenon demanding alertness dur- ing the implementation of the new 'sys- tem, particularly because it involved a computer, was the tendency to blame the system for unrelated or marginally associated problems arising during the same time period. Although · automation can be held responsible for certain amounts of stress and strain and result- ing conflicts, it is little understood by many librarians. Thus it often serves as a scapegoat for other kinds of difficul- ties (e.g., · slow-downs resulting from staff tum-over, problems related to absorbing processing for additional branch libraries). Faced with possibili- ties of this type, the only way to imple- ment any system successfully is to have the full cooperation and dedication of the library staff who must make it work. A superbly designed system, excellently programmed, will fail if · personnel lack understanding or are unwilling to con- tribute the extra effort needed to accom- plish a major change in procedures. Al- though an excellent system is a necessity, Northwestern's experience indicates that attitudes are at least equally important among the many requirements for suc- cessful implementation. RESULTS While certain key people bear the brunt of responsibility for systems plan- ning and explanation, general staff accept- ance, based on a perception of the proj- ect's actual worth, determines ultimate success. The immediate workload was certainly not lightened during the early months of operation. Some of the filing drudgery was, however, quickly eliminat- ed, and the increased accuracy in record- keeping became obvious as a long-term benefit. From the standpoint of im- proved operations, the greatest immedi- ate impact was in catalog card produc- tion. Cards are produced once a week at th.e rate of over 6,000 in two hours, thereby equalling the production of the original typing pool for approximately seven-and-a-half work days. Cards are printed in correct filing order, sorted to the eighth character, and grouped for each of the various catalogs. Another system benefit, automatic punching of charge cards, has reduced the workload of the Circulation Department, which formerly had a staff member operating a ·keypunch. The increased accuracy of these cards is another plus. Among · the possibilities offered · for the future by library automation is the charice to share files of cataloging data beyond those provided by the MARC project. Such ·sharing will reduce the I J Automation of Technical Services I 369 need to enter full non-MARC data via the terminal keyboard and spare the time currently spent tagging informa- tion for input. The computerized serial holdings records have potential for a re- gional serials data bank and promise for an expanded interlibrary loan op- eration. Naturally a cost savings can be anticipated if other libraries choose to join Northwestern's system in coopera- tive acquisitions and processing arrange- ments. Regardless of the cooperative poten- tial, Northwestern has decided that au- tomation is the only way to deal with the problems to be faced by technical services operations in a large research library now and in the foreseeable fu- ture. The ever-increasing volume of ma- terial to be acquired and processed without additional staff and the demand for the rapid preparation of items and accompanying cards for use permit no acceptable alternative. Northwestern's experience has confirmed the value of a direct systems and library staff rela- tionship, close attention to fiscal reali- ties, development of on-line capabili- ties, complete utilization of available hardware and personnel, and a primary emphasis on c.ommunication. With full attention to library needs and informed enthusiastic staff cooperation, automa- tion can and does succeed. REFERENCE 1. James S. Aagaard, "An Interactive Comput- er-Based Circulation System: Design and Development," Journal of Library Automa- tion 5:3-11 (March 1972); Velma Venezia- no, "An Interactive Computer-Based Circu- lation System for Northwestern University: The Library Puts It to Work," Journal of Li- brary Automat·ion 5:101-17 (June 1972).