Case Western Reserve University: What Presidents Need to Know About the Impact of Networking on Campus Background paper for HEIRAlliance Executive Strategies Report #3 "What Presidents Need to Know ... about the Impact of Networking on Campus" ----------------------------------------------------------------- prepared by representatives of CASE WESTERN RESERVE UNIVERSITY Agnar Pytte President Raymond K. Neff Vice President for Information Services D. Kaye Gapen Director of the University Library ----------------------------------------------------------------- Copyright 1993 by HEIRA The Executive Strategies reports are published by the Higher Education Information Resources Alliance (HEIRAlliance), a vehicle for cooperative projects between the Association of Research Libraries, CAUSE, and EDUCOM. For information about paper copies, contact CAUSE at 303-449-4430, orders@CAUSE.colorado.edu ============================================================= What Presidents Need to Know about the Impact of Networks on Campus: The View from Case Western Reserve University Why network the campus? At CWRU, the network connects people and provides access to a wide variety of information resources and services which support instruction, research, and service to our community, as well as to other communities. The University believes that the network will make a substantial contribution to fulfilling its core mission; our campus-wide network was designed to support a wide variety of academic and administrative applications in all of its departments. The network is the hub of our Electronic Learning Environment. It connects all faculty offices, classrooms, student residences, laboratories, and libraries. What network services to offer? Five families of network-based services are offered at CWRU: data, voice, video, telemetry, and control signaling. Examples of each family include: computer data services using standards-based Ethernet, Token Ring, and FDDI data transmission to over 8,800 desktops on campus, electronic mail, and a campus-wide information system; our voice services include CENTREX telephone services, long distance, and facsimile transmission; video services include cable TV, a video information bulletin board, video mail, and videoconferencing; telemetry handles parking lot surveillance and security sensors in campus buildings; control signaling provides for energy management and electronic door locks. The campus network is the only comprehensive wiring plant on the campus; we perform all communications tasks using this network. How do the communications services interrelate? With the convergence of information technologies to a common digital format, all five of the communications families are themselves becoming interrelated. Today, each service family resides in its own world and is managed by its own corps of specialists. In less than five years, common transmission protocols will replace the present ones and offer substantial opportunities for achieving more effective campus operations at significantly lower costs. Therefore, each administrative service unit offering a communications technology should be challenged to find ways of integrating its communications technologies with those of the other units. Often this integration will result in some need for organizational restructuring. Do you know what your campus is spending on telecommunications services and installations, on distribution of television signals, whether produced in-house or acquired from the airwaves, satellite, microwave, or cable? How much is being spent to install wiring for other information distribution systems including energy management control systems, building and area security and surveillance systems, and other telemetry systems? What fraction of your telecommunications bills are for facsimile services? Nearly all universities are continuing to use independent wiring for each family of communications services, often with little or no coordination. At CWRU, we have installed a universal wiring plant which is used for all types of communications and is adaptable to the rapidly changing technologies of both analog and digital communications. As a result, we are saving money by eliminating the redundancies of separate wiring systems, and the wiring we do install is coordinated centrally. As a result, our wiring is utilized more intensively, so that we achieve the greatest payback from these investments. By carefully documenting our wiring plant, we save money in building renovations by reusing the wiring rather than having to install new wiring each time. After many years, we expect that our strategic wiring plant which has been managed in a coordinated manner will achieve significant savings for the university. Significant trends in networking technologies and services I. All information formats are changing to digital representations. Consider the following four examples: (1) telephony, including voice, facsimile, answering machines, voice-mail are changing from analog signaling to digital signaling and digital encoding of information; (2) television is being reengineered for programming production and distribution in a digital format; (3) printing technologies from xerography to off-set are changing from analog (e.g., light-lens) to digital scanning and printing processes; and (4) photographic technologies are changing from film-based, analog image capture to filmless, chemical-less, digital image capture. II. Digital technology transmits information with greater accuracy and precision, at lower cost per information item, and, over time, accuracies and precisions will continue to increase; transmission costs per unit will continue to decrease. III. Digital technology stores information in a far more compact format than analog storage. A dozen or more electronic books may be stored on a single CD-ROM, including the images and charts. Other forms of optical and magnetic digital storage span a wide range of storage capacities and retrieval performance. Such technologies will play significant roles in the library of the future. IV. Distribution of information in electronic form is more efficient and more effective than distribution of information in paper format. Electronic mail and electronic file transfer offer speed and certainty of delivery that no manual postal system can provide. The import of these trends is that (a) academic scholarship and instruction are beginning to feel the first effects of these revolutions, (b) by capitalizing on these trends, it is possible to increase the effectiveness of the administration of the academy and to reduce the costs of administration. The computer becomes the all-purpose information input device, output device, and information reformatter. This means that all faculty, staff, and students will have to have ready access to a computer. For the employees of the institution, it will be cost beneficial to equip each "knowledge worker" with a computer. For faculty and others who work at home, they will need a computer there as well. Given the relative cost of powerful computers today (under $2,000), the investment in capital for two computers for each faculty member will pay handsome dividends. The campus-wide communications network will make it possible to share information across these many individual systems. In the 1980's, small numbers of personal computers were often linked together in local area networks (LANs). The principal purposes for these linkages were to share information and to share printers and file server resources. The computer makes it possible to reuse information without having to expend large amounts of effort each time. Reuse may simply involve reformatting or it may involve complex modeling; the user selects the methodology appropriate to the task at hand; institutional data repositories become network-based information resources and provide the institution's administrative staff with consistent, high quality information. How does a campus-wide network relate to the distribution of information? Consider first the distribution of information in the form of paper. On campuses today, we generate memoranda, reports, and analyses on a computer. Generally, the material to be distributed by the campus' mail service arises as output from a convenient office-based laser printer, which in turn is attached to a computer. If that computer is also attached to a campus-wide network, then the potential exists for distributing the output over the network. Thus, distribution of information over a network replaces the distribution of physical stacks of paper. Using the network, the time to distribute the information decreases to virtually nothing, the certifiability of actual delivery approaches 100%, and the marginal cost of the delivery nearly goes to zero. Networks are generally environmentally benign, and their use is positive for reducing the use of paper. Impact of the CWRU campus-wide network The impact of networks on teaching, learning, and the curriculum Case Western Reserve University is building an Electronic Learning Environment (ELE). Our ELE was conceived to provide students and faculty with tools for creating, organizing, storing, and transmitting knowledge as an integral part of the learning process. In the next century, universities must have the capability to draw widely on resources held in various forms by other institutions and in enormous electronic databases, as well as to shape and deliver these resources to students and faculty. The CWRU ELE would be a response to this challenge. The ELE would be the integration of seven major components of information technology: powerful software tools running on personal computers with multimedia capabilities; digital libraries; a campus-wide information system; specialized computational servers; curricular transformation through the use of instructional technology; re-engineered classrooms; and a universal campus network. Each of these components would require both development and investment because none was in existence in 1988 when the ELE was first conceived. The ELE vision was based on the synergy among these seven discrete elements. Indeed, it was the integration of these elements that would ultimately fulfill the promise of an electronic learning environment. The campus-wide network would be the glue that held together all of the pieces of the puzzle; it would be the essential infrastructure and key to the entire enterprise. With the ELE in place, we envision that the library would provide electronic documents which students could use from any campus location, including their residence-hall rooms. Classrooms would be re-engineered to permit faculty to use electronic information in their teaching, and students would be able to interact with these electronic materials at a later time and at their own pace from any campus location. A campus-wide information system was envisioned to serve both discrete information and to provide software tools for the manipulation of this information. CD- ROM publications would be joined to other digital formats as another network-based resource. The curriculum would evolve as faculty incorporated electronic information and software tools into their courses. Electronic mail would provide more frequent contact between students and their teachers and would also facilitate learning among students who would form ad hoc study groups across the campus. Students and faculty working off campus would be able to use campus-based information resources on demand, making accessibility to information the key to use rather than ownership. Linking the CWRU academic center to other scholarly centers would also be facilitated. Taken together, CWRU would be preparing its students for a future in which they would have the familiarity and competence to handle complex information and to do so without regard to its origination or its destination. As we started to develop the ELE concepts and as we looked at the specific ways students learn, we saw increased opportunities where information technology could be used in the educational process. We saw the potential of the electronic library expanding services to scholars and becoming more efficient in disseminating information, although no one believed that the electronic library of the future would be less costly than the library of today. We were, however, confident that the cost per information transaction could be significantly reduced. Now four years into this ten-year developmental effort, the ELE is partially implemented; the essential infrastructure is in place, and the faculty are being trained in its use. Additional computing staff members have been hired and integrated into the academic fabric. Because the ELE would eventually require network functionality which we did not have initially, we decided to architect the network in such a way that it could be enhanced as these needs materialized. Our designs would permit these enhancements to be made with the least possible effort and cost. The classroom of the future is wherever the student is. Through the use of advanced networking, the student and the teacher can be linked together as effectively as if they were in the same room. In fact, given the condition of many of our classrooms today, we believe that with the newest types of high resolution displays, students using these projection systems will likely see more and see it better than they can in a 200-seat lecture hall. This is especially relevant in considering computer-oriented multimedia, because in many respects, the workstations of tomorrow will combine high quality video display screens with the ability to process the multimedia information in the ways we want it, customized to the time and the purpose of the user. Another significant aspect of the new learning modalities is the built- in capability of replaying a learning segment. We do not all learn at the same pace, and as we learn, we relate new information to what we already know. Since we have different experience bases with which the new information is to be related, we will not be homogeneously incorporating it. Being able to replay and to learn a new subject at the depth appropriate to the purpose of the individual student will give the user a capability much beyond what the student is offered today in our "one-size fits all" formats for presenting information. The logical extension of this idea is to use information technology to deliver customized instruction. The validity of this is based on the differential "inputs" of prior experience each student brings to the learning situation and on the increasingly important learning goal of attaining differentiated outcome levels of achievement as "outputs." Because students learn in different ways and at different rates, there will also need to be a differentiated learning process in which students assimilate new information. Information technology, properly used, can produce the customized electronic learning environment we believe will be the hallmark of higher education in the twenty-first century. In graduate-level professional education, we see the importance of using extensions of CWRUnet to bring the campus and its information environment to the off-campus offices and laboratories where our students and faculty interact in their training. By using the same advanced communications technologies of ATM and SONET, which our telephone and television vendors will be using, CWRU will facilitate the extensions of the network to off-campus locations, including to health care settings, hospitals, clinics, physicians' and dentists' offices, law offices, social work agencies, and governmental agencies. By combining some of the distinct technologies mentioned above, it will be possible for the university of tomorrow to offer a form of time- shifted learning, the presenting of learning opportunities to students customized to their own schedules. Time-shifted learning breaks up the rigidity of scheduled classes which universities are finding increasingly limiting in meeting the needs of their non-traditional students. Time-shifted learning will clearly be useful in continuing education programs. The system will give our community of potential students more options for meeting their multiple needs. Use of readily available information and increasing the access to information is part of the information technology culture we are developing. Time-shifted learning will likely involve the use of multimedia servers and advanced network-based delivery systems from the university campus over ATM networks to desktop learning stations, which will combine elements of personal computers, digital television, and electronic libraries. In summary, the network turns the university into a student-centered learning environment. With the network, we see faculty working more like coaches than like lecturers. Faculty and students will come together at any hour, whenever communication is necessary. In this way, the "classroom" is always open and the professor's office hours are 24-hours per day. The impact of networks on research It is easy to predict the impact of networks on research because it is from the research domain that sophisticated national networking began in the 1970's. It started in computer science and quickly spread to other disciplines where immediacy of information dissemination was of critical importance. In these disciplines, scholars could not and did not wait for journal publications to learn about the latest discoveries of their distant colleagues. Although formal peer-reviewed journals are still used in these fields, the circulation of electronic "preprints" has become a way of life. In some disciplines, specialized computer "servers" have been created and are maintained whereby manuscripts can be deposited and from which electronic articles can be selected for downloading and study. Faculty and their students are able to keep up with fast-changing fields of study by connecting to the "server" which covers their field or subdiscipline. In many intellectual domains today, great compendiums of electronic scholarship are being created; one subscribes to these resources automatically by sending electronic mail to the appropriate "listserve" system which maintains the service's subscription lists. In research, then, we see networks making accessibility to information the key to use rather than ownership. As more and more research problems involve necessary collaboration and teamwork, the network is providing the critical linkage of the CWRU academic center to other scholarly centers worldwide. It matters less and less where you are based as a scholar today because the network allows you to interact with others working in your field. Because electronic mail is so democratic in its style, one's electronic contributions to the network-based servers are judged more on their inherent worth than on the authority of the sender. The impact of networks on public service At CWRU, we are using the network to provide a pioneering information service to the entire Cleveland community. It is called the Cleveland Free-Net (CFN). Developed during the period from 1986-1989, today it offers over 300 distinct information services from law and health care to movie reviews and popular science; it provides several venues for people to dial into the system free of charge to exchange information. The CFN has been copied by at least forty other communities because CWRU licenses the CFN software. Today, the CFN has over 40,000 registered users. They range in age from 8 to over 88, with a sizable population being local teenagers. We anticipate that they will log over four million on-line sessions during 1993, and the cost to CWRU will be less than $200,000, making the cost per session average less than five cents (an average user session is about 30 minutes). We know of few ways that a research university like CWRU can have this amount of community impact with so little cost. At the same time, this Free-Net service projects the type of image we want to have in our community, that of a sophisticated, "high tech," contributing neighbor. The Cleveland Free-Net also serves the public and private (K-12) school systems in the Cleveland metropolitan area. Teachers at these and other schools nationwide exchange information about their successes and failures in using information technology in primary and secondary education. It is hard to quantify the benefits that this type of information exchange can have, but we know that the teachers appreciate this service; all free of charge. The impact of networks on libraries Case Western Reserve University is creating an environment composed of hardware, software, infrastructure, and highly skilled personnel that, together, will enable scholars to utilize electronic tools and resources to learn, teach, research, and conceptualize in new and powerful ways. This new Electronic Learning Environment will at the same time support the transformation of education while still reflecting the honored and classical values of scholarship and librarianship. Networking is transforming scholarly discourse; more people at all levels can participate in scholarly forums connected by a variety of networks at the campus level, the city level, the state level, the regional level, the national level, and internationally. Libraries have always had as part of their mission the support of scholarly discourse as the intellectual resources they have housed have been grist for the scholarly mill. That support extends to include what is held for scholars in the physical library as well as the information and knowledge which can be made available in digital form on the same networks which support both informal and formal scholarly communication. This is the first time in the history of humankind that all forms of scholarship and communication can be conveyed and shared in one communications format--digitalization. Text is the simplest form of communication which can now be shared in digitized format. However, through the potential afforded by digital electronics, other formats can be equally available, namely, sound recordings, art, photography, graphics, and moving images. The only sensations which remain as a challenge for electronically supported sharing are smelling and feeling. And we can even begin to approximate those in the form of electronic information sharing as electronics and computing provide a "virtual reality" which evokes sensations of smell and feel using computer- simulated psychological events at the receiving end of the networked microcomputers. It is the case today that we can think, teach, learn, and conceptualize at one time with the assistance of electronically formatted multimedia, taking advantage of color, sound, text, and moving image to enhance our conceptualization experience. The challenge in the electronic learning environment is for library staff to create new tools and interfaces for collection building, new intellectual organizing principles and mechanisms for information and knowledge organization, new partnerships in the creation of supporting infrastructures, new service models, new prototypes for experimentation and learning, and a new skill base which encompasses and exceeds the knowledge and skills which have comprised classical librarianship. It is the librarian of the future who will create the library of the future. The implications for the university library are many in a networked electronic learning environment. Conceptually and practically the library is being transformed in its ability to support unique patterns of faculty and student use, as well as to extend the sharing of its resources beyond the physical walls of the library building. Conceptually, the library of the future models the intellectual process as an active, on-site experience where technologically-driven tools allow the users to reach out for knowledge, wherever it may be, whatever form it may be in, and to gather that knowledge with tools that will record it and allow it to be manipulated, compared and contrasted to other related knowledge and thus, to facilitate both analysis and synthesis. The vital components of the library of the future will be: intellectual access to information and knowledge resources in all formats; a networking infrastructure to act as a transmission highway; workstations that feature hardware and software to facilitate on-line access and manipulation of the gathered knowledge; a library staff skilled in the tools of knowledge management, interface design, and user education; adequate and varied training facilities for both users and staff; a variety of flexible user spaces that can evolve along with the transformation process; and a vigorous and viable working collection that supports teaching and research. The boundaries presented by these elements are formed only by our understanding of the concepts and the evolution of the transforming tools. A user can choose from physically accessible, on-site collections, but the user is not limited or bounded by physical location of material or its format. In fact, wherever the user can access the tools of on-line access and the expertise of knowledge management, the user can access the resources of the library of the future. The staff of the library of the future are highly visible, active collaborators in the process of knowledge management, and much of their work is constantly evident to the users in the form of interfaces designed to overlay digital information resources. These digitized collections of information and knowledge are organized by thought processes, and the tools the user employs to manipulate them will think with the user, modeling, simulating, and engaging all the user's senses as part of the process of synthesizing knowledge. The new knowledge that emerges will be in a format that can be managed and transported. Thus, as the classical library offered physical access, the library of the future complements physical access with intellectual access and the ability to facilitate knowledge synthesis. Instead of a hallmark of discreteness, the library of the future offers integration as its enduring theme--integration of formats, integration of knowledge, integration of user spaces, and integration of the experienced library staff. The library of the future can be conceptualized as a collection of virtual libraries. A virtual library is defined as the juncture of an on-site collection of current and heavily used materials in print, microtext formats, and electronic form, with an electronic network which provides access to, and delivery from, external library and commercial information and knowledge resources and services worldwide. In essence, the faculty member and student are provided the "effect" of a library which is a synergy created by bringing together technologically the resources of many libraries, information services, and knowledge stores. Networking and the Library of the Future: Pragmatic Considerations o The Network In the networked library of the future, all of the available networks must be designed and implemented to support the transmission of all forms of digital materials: text, sound recordings, graphics, art work, and so forth. Many of these digital files are quite large and require broad bandwidth and high speed. Fiber optics network, point-to-point, is today's most effective network capability, and it should be the goal for all universities as they move toward the library of the future. o Budgets The library is in transition from its historical classical form of collections shelved in a physical space to a combination of owned and networked digital collections. Historical library budget models have been in transition for the past ten years and will continue to be refined over the next ten years. It is essential for the university community to maintain flexibility in funding to support not only the classical library, but also the library of the future. A portion of this flexibility can be viewed from a practicality point-of-view as "venture capital" which is available to support new patterns of library collections and service as both prototypes and as operational systems. Both owned and accessed collections are increasingly reliant on equipment (both mechanical and electronic) for the use of the collections. This is a major shift for library budgeting which has been traditionally limited in the funding available for equipment, for example. New patterns are also emerging in the funding of digital and network-accessible collections in which the cost is initially ill- defined from the publishers' perspective. The campus community must understand that funding patterns for creating the library of the future will be irregular for some time to come. o Collections The library's collections are becoming a mixture of owned print and other hardcopy materials, such as sound recordings, microtext formats, and photographic slides. Adding to the mix are digital intellectual resources which can either be owned or accessed via a variety of networks. The digital collections can be network-accessible on the campus and may be purchased and made available as CD-ROM resources or as mainframe-supported data bases. Resources which are initially published as print may be digitized in the library or elsewhere on the campus, becoming part of the network accessible collections. Essential in the networked intellectual resources framework is the monitoring of collection use for copyright royalty payment. Copyright law in the electronic information and knowledge world is in flux and careful monitoring of the use of digital collections is essential to the tracking of new patterns of use and budget support of the library. o Physical Space In the networked environment, the library continues to exist as a physical space, as well as a suite of networked access. In the physical space, working collections continue to be housed. Implications for physical space with the networked environment is the need to add space and equipment in the library for faculty and students to created digital copies of print, sound, and graphical materials, accompanied by electronic multimedia workstations for the merging and synthesis of digital information. The network supports not only distributed digitalization and media conversion, but also distributed printing; high-speed printers (some with color) can also be located in the library for faculty and student use. Finally, the library staff are also readily available in the library for face-to-face problem-solving support for faculty and staff. While many of the dialogs which currently occur in the physical library among faculty, staff, and students can occur over the available networks, many information-seeking problems will continue to be solved in face-to-face working situations. The library's physical space will have to be adapted or constructed to be network friendly, i.e., connections to the network need to be readily available. This may require physical remodeling or, in the case of new library buildings, a design approach which incorporates sophisticated network and electrical connections from the ground up. o Document Delivery The library staff will continue to provide physical document delivery of hardcopy collections. In addition, the staff can also move in the direction of providing networked document delivery through the use of scanning of print materials for delivery as standard, low-resolution facsimile, electronic mail documents, or by transmission of high- resolution image files. o Standards The standards which support the classical library are also in transition. The library of the future encompasses digital collections which have their own set of electronic standards both for the networks and for the multimedia collections. It is important for the campus community to discuss and implement a suite of digital standards which will provide consistency in the use and networking of electronic collections. Examples include standards for graphics, markup languages, and open network architectures. o Multiple Computing Platforms A variety of micro and mainframe computing platforms are used in universities. The implications for the library include a similar support of multiple computing platforms for both accessing, scanning, and manipulating digital collections. The library staff need to be skilled in the use of a variety of computing platforms, and a variety of computing platforms needs to be available in the library. o Education in the Use of Library-Supported Intellectual Resources As the universe of available information and knowledge continues to expand exponentially, it is important that the campus community have a program for continual education in the use of library-supported intellectual resources, both electronic and paper-based. o Research and Teaching Merging Electronically Our experience at CWRU indicates that the electronic learning environment increases the potential for merging research and teaching. We have found that an increasing amount of research can be supported by the microcomputer through, for example, the downloading of electronically available and/or scanned collections. As a result, it has become much easier for a faculty member to incorporate digital research materials into digital courseware. Both library staff and faculty members should work together to ensure that the standard formats for digital materials are useful not only in the research process, but also in the teaching process. o Knowledge Management At CWRU, we are defining a new context for the scholarly process, namely knowledge management. Knowledge management is a model for scholarly communication in which faculty and research librarians share the responsibility for the collection, structuring, representation, dissemination, and use of information and knowledge, using computing and communication technologies. In the networked environment, it is possible for knowledge diffusion to expand beyond classical publishing to include the placement of newly created digital research results in the campus database system. Thus, new knowledge can be made available by the faculty member via a variety of networks. Knowledge management can be employed at any university because it is based on a construct of hardware, software, expertise, and policy development. o The Virtual Library At CWRU, we have determined through an extensive survey of faculty members and students that their use of information, knowledge, and information technologies varies by discipline. As we create the library of the future, one technique we are building upon is the creation of a series of virtual libraries. The CWRU virtual libraries are not only the "library without walls." We are thinking of collections and the library- staff designed programs which add value to the collections, but also in the information and knowledge resources to which we facilitate access. Our CWRU virtual library is a system intellectually conceived with and delivered to faculty members in each academic department, tailored to their unique user information environments, drawing upon the capabilities of a variety of computer and intellectual technologies. o The Universal Finder The world of information and knowledge is increasingly diverse and, in some instances, chaotic in terms of finding tools which are effective over a variety of databases of information and knowledge. One of the major challenges for higher education is the exploration and implementation of software which provides powerful retrieval capabilities. The Universal Finder is created by bringing together software and hardware in support of the storage and retrieval of information and knowledge from the CWRU information system and from connections to national and international information and knowledge databases. The Universal Finder incorporates the present on-line library system (including its public catalog), digitized bibliographic tools, the full-text of monographs and journals, and digitized images of non- textual resources (e.g., slides, photographs, sound-recordings, art objects, moving images, and so forth). University support for the development of sophisticated searching software needs to be a high priority. o Staff Development and Organizational Structure As we have defined the library of the future in the electronic learning environment, the knowledge management construct, and the creation of the virtual libraries, we have begun to define new roles, responsibilities, and organizational structures for library staff. At the same time, we are identifying the skills and supporting staff development programs which are integral to moving into the future. The patterns we have identified include an organizational structure which moves from library departments organized by function to cross-functional teams of library staff organized around the major disciplinary patterns of library use which frame the virtual libraries. In addition to a new organizational structure, we are developing a series of classes and other educational experiences for introducing library staff to the skills and interpersonal relationship building which are characteristic of self- directed teams. New skills are integral to the creation of virtual libraries, including a higher level of expertise in the use of electronic multimedia microcomputers, database management and authoring software, the use of scanning and other digitizing equipment, the creation of controlled vocabularies for the disciplines represented by the virtual libraries. Library staff responsibilities are becoming more sophisticated and varied. Budgeting for and supporting staff development is a high priority for creating the library of the future. The likely result of creating the librarian of the future who will, in turn, create the library of the future is an increase in library staff salaries. The impact of networks on administration and campus management The campus-wide network at CWRU was designed to support administrative and business service applications. The University is developing on-line, transaction-oriented applications that acquire and use information to serve students, faculty, and staff. These applications use servers and microcomputer-based clients connected by the campus-wide network. This configuration will replace batch-oriented, mainframe applications which were designed in the 1960's. Although we envision that these administrative applications will not in and of themselves be demanding of the network, the fact that the network is ubiquitous and standardized in its interfaces will facilitate the development and implementation of these newly designed applications. A new service emerging from the network is a real-time, charge account system. Initially, this system takes all telephone toll and long- distance charges and keeps both the detailed transactions and a running total which the students can see using their user-id's on our campus- wide information system. Later, we expect to collect charges for the use of photocopy machines in the libraries and soft-drink machines on the campus, for course-related custom-published books as created in our innovative electronic library, and for other cash-associated transactions on the campus. Monthly, this system will transfer an authorized debit to the student's chosen banking account using an electronic funds protocol, with CWRU receiving the credit to its account. The time-value of the cash money will be used to provide a discount incentive to use this real-time system. Again, it is the ubiquity of CWRUnet which makes these related applications fit together in a cost-effective solution. CWRUnet conducts telephone traffic to the local central office of our telephone company, Ameritech-Ohio Bell. At present, we purchase CENTREX services for our office, laboratory, and residence-hall telephones. CWRUnet's premise wiring, as well as its fiber backbone, are being used to carry telephone signaling efficiently and compactly to all areas of the campus. CWRUnet carries a variety of other types of signaling: parking lots are monitored by surveillance television cameras which transmit the information over CWRUnet; residence halls and many other buildings have electronic door locks which open only to authorized persons during predesignated periods of time, customized to each individual's needs as recorded in a central data base accessible via CWRUnet; and energy system controls in campus buildings are managed centrally by transmitting signals over CWRUnet. Since e-mail connects students to each other and students to both faculty and staff, electronic mail capabilities are important network services for the CWRU community. The CWRU electronic mail system must serve the entire campus community and have linkages to networks beyond the campus. Further, there should exist the capability to send and receive facsimile (fax) messages through the electronic mail system in an integrated manner. Thus, users should be able to create documents on their computers and fax them to receivers at other locations. Electronic data interchange, or EDI, is a standard (X.12) which is currently gaining favor as a means of performing various administrative and business functions, including exchange of data on student applications for admission, issuance of purchase orders to a wide variety of vendors, and reporting on employees for pension-plan participation. On our campus network, we have established three sets of servers, one each for e-mail, facsimile, and EDI. Rather than installing EDI software on each system or mainframe which requires an interface for EDI, we have established an EDI server (currently a 486 microcomputer) with out-bound ports for establishing EDI connections. This server also resides on the campus network and acts as the clearinghouse for all EDI transmissions, just as our electronic post office server handles our e-mail traffic. Our fax server provides both in-bound and out-bound delivery of facsimile messages; users do not have to purchase or rent a dedicated fax machine to avail themselves of this capability. CWRUnet is the vehicle for providing applications software to our campus community. Standard commercial software is made available without charge to students, faculty, and staff from the network-based software library that includes a variety of word processors, spreadsheets, symbolic and numerical mathematics tools, graphics and statistics packages, data base management systems, and locally-created courseware. All owners of microcomputers know that the costs of software now dominate the total cost of ownership. To help control costs, we are using the network's software libraries to provide application software to campus-based users at no cost. Thus, users do not have to purchase their own copies of common software packages but can share them through the network-based software library. This saves many hundreds of thousands of dollars each year. More importantly, centralized maintenance of the software library facilitates the enforcement of procedures for proper software licensing and the protection of intellectual property. There is very little incentive for software piracy in an environment where software is so readily available through the network. Wiring the campus Communications technologies exist in two basic domains: the atmosphere and the cable. Today, voice, data, telemetry, and control signaling mainly use wires, and video uses the airwaves. In one of those interesting reversals of technologies, we are seeing the future of video as being based on the cable, not the airwaves, and of telephones, which for so long have been based on cabling, moving principally to the airwaves. The other families of communications technologies will remain largely wire-based, but a new domain for some types of computers is now emerging in which a subset of networked computers becomes totally free of the cable, a so-called "nomadic computer." Such devices are expected to use the infrared spectrum of radiation to communicate with other computers and with conventional wire-based networks. What about the future of wiring? Ethernet and Token Ring are good short- term technologies for data transmission, but for multimedia and other image-intensive applications, these technologies will ultimately prove to be inadequate. We believe that full multimedia transmission will require a new physical layer and data link layer. Thus, we have adopted technologies with higher performance ratings using optical fiber to replace copper at the physical layer. Since optical fiber would not be found wanting in terms of bandwidth potential, an investment in a fiber- based network would have strategic value for this university. We also found that existing data communications technologies could be worked in a cost effective way in terms of hundreds of megabits per second transmission rates using multimode fiber strands, but it was single mode fiber strands which offered the potential of transmission rates of tens of gigabits per second. The problem with single mode fiber, however, was that the optoelectronics would be prohibitively expensive at this point in their technological life cycles; these devices would very likely decrease in price over the next five to ten years. Which fiber to use? We knew that we would need to start using the network immediately; the cost of the single mode fiber was not within the limits of the budget. Even the multimode fiber was expensive relative to the use of data-grade unshielded twisted pairs (category 5), shielded twisted pairs, or thin or thickwire coaxial cable. In less than five years, we believe that copper-based wiring will become the limiting factor in the transmission of multimedia information. We made the decision to use multimode fiber for data and to pull single mode fiber to every desktop. We will be ready to use single mode fiber when it becomes a more mature technology. Developing a networking strategy for the campus A high-ranking member of the campus administration should be charged with the responsibility for developing a campus-wide networking strategy. For the campus, there will need to be developed a vision for the network. Issues in developing a specific strategic plan will have to be addressed; each university will have to ask some basic questions and then write down the answers that make sense in its particular context. There seems to be very little that is right or wrong in any answers that might be given; it is important that they are valid and realistic. o Determining the aspiration level for network services. How state- of-the-art is it necessary for the network to be? Will academic programs and administrative functions be compromised if the network does not have the latest functionalities and the highest performance levels? o Determining the phasing of technological developments. With rapid changes taking place in networking technologies, how will the institution handle the changing technological base over the course of the network implementation and thereafter? It should be understood that installing a network is generally a process that takes several years during which technologies will continue to evolve. o Determining the role of individual, microcomputer-based computing systems. Some universities have come to the realization that the microcomputer is an important tool in the management of scholarly information whether done by the student, the professor, the staff member, or the administrative officer. In short, these schools have said that the microcomputer is a fundamental tool and that they are building their institution's strategies with the microcomputer in mind. Unfortunately, the microcomputer is far less useful as a standalone device than as a networked device. Thus, an institution's aggressive stance toward microcomputing may lead it to take a similar stance toward networking. o Determining who will have access to the network and the level of services provided? Does the network provide services to students, faculty, and staff equally or are some services available to some groups but not to all groups? Is there a separate administrative network for reasons of security and confidentiality? Are there restricted sub- networks? o Charging or not charging for network use. Charging puts the "have not" groups at a disadvantage. If funds are distributed to departments to pay for networking, why not budget the operations costs centrally and then operate the network from a central administrative unit. There are many possibilities here: to charge for network traffic, say, so much per packet of information, to charge for network access, say a fixed amount per month, or not to charge at all and consider the network like the university's library. The library model is interesting because it says that basic information resources are a staple of the university and access to and usage of information resources should be independent of the ability to pay. This model recognizes the importance of universal access, and the network achieves its real value when everyone has access to it. If one uses a charging scheme, less than 100% of the campus population will choose to have network access. There is also the electrical system model. On nearly every campus, electrical service is provided to all campus constituencies without charge, even though the university has to pay for the electricity. A network outlet is placed wherever information is needed, analogous to placing electrical outlets wherever electricity is needed. Charging policies may seem like a necessity, but considerations of the library model and the electric utility model show other ways to meet the operating and capital costs. Of course, it is well known that libraries have huge capital costs for buildings and scholarly materials, as well as large operating costs for people-based services; electrical systems have huge capital costs for putting the electrical plant on-line and for the wiring system, and they need large operating budgets to cover the costs of fuels and maintenance of their equipment. The network enjoys these similarities also. o Determining the relative proportion of institutional funds, operating and capital, to be used for networking. Many campuses have a budget for telephony, with virtually all of it coming from the operating budget. Networking infrastructure is extremely capital intensive with a long useful life. Reconciling the needs for large amounts of capital and amortization within operating budget guidelines is the task that needs to be considered. There is also the question of how much can be budgeted for networking on an annualized basis. If one takes into account all forms of communications (voice, video, data, telemetry, and control), then a realistic annual budget figure for networking and communications is in the range of 1 to 2% of a university's budget. This figure includes amortized costs. o Determining the sources of new funds for networking. In general, it will take an infusion of new funding to initiate a networking project. This is simply because the current communications capability will be needed until the new system is operational. The institution will have to pay for both systems at the same time. Also, the network will require additional funding because new and expanded services are being provided. It is completely unrealistic to try to operate an expanded networking service on the old budgeted amount; new sources of funding are a requirement. Some universities are initiating a student computing fee to bring in sufficient revenue to cover some of the additional networking costs. o Determining the organizational structure for networking and the degree of network decentralization. Historically, computer centers have been the organizational home of data communications services, but with the advent of library-based computing for the catalog, circulation, and acquisitions systems, another major campus organization needs to be recognized as an information resource. When one factors in voice, video, telemetry, control, and the other decentralized computing facilities, one sees that the computer center should not have a monopoly over the network. Fundamentally, the networking organization should be in neutral territory with respect to the various computing facilities and the library, and this calls for an independent organization. If the campus has a chief information officer, then the networking management should report to this person. There is also the question of decentralized networking in which schools or even large departments build their own networks. In general, this is a poor strategy because it leads to operational complexity and greater operating costs. o Determining the role of the administration and its organizational structure. Is there a networking czar? To whom does s/he report? Are all of the communications technologies merged into one organization or are they dispersed across the administrative units? At CWRU, we favor centralized networking because we believe that the advantages of common systems, managed by full-time specialists, are greater than the disadvantages. This centralization is based on the electrical utility model of the network. o Determining the requirements for high-level technical support and modest skill-level support services. CWRU has deployed over three thousand workstations in the past four years, all with networking support. Our experience is that the end user is ill-equipped to handle (and does not want to be bothered with) the details of installing networking hardware and software. The end user wants the workstation provider and/or the network provider to set up the network and make it work. Further, our experience is that networks are operationally fragile. It is still too easy for the end user to create network dysfunction or to leave the workstation in a nonfunctional state with respect to the network. In such cases, how does the user get restarted? There has to be a simple way to get user assistance; there has to be a staff of persons ready to give technical assistance, to make a "house calls." o Determining the role of the students in filling networking staffing positions. Students can be excellent members of the network technical staff; typically, they are of high energy, work with enthusiasm, perform network evangelism, do not expect high salaries, and are empathetic with the end user. How network management reconciles the exigencies of student life with the needs of maintaining the network will dictate what roles, if any, students play. Critical network staff positions should not be filled by students, but many auxiliary positions are entirely suitable, and the experience for the student is generally quite meaningful. o Determining where networking technology standards will be set. No network can accommodate all possible devices and interfaces. Standards for the campus network will have to be set and promulgated. On some campuses, newer, emerging standards will also be adopted for the latest in functionality or performance. What organ of the university has the responsibility and authority to set these standards? Where should this responsibility be delegated for day-to-day functioning? What happens when a faculty member finds out that his "pet" project cannot be accommodated by the adopted standards? Can one appeal? And what does an appeal actually mean? Having hard and fast standards will be necessary. The network will need to be managed with a palpable discipline. If not, then the effort will fail. Imagine how the electrical system would function if faculty members could appeal the decision to have outlets with only standard 110 volt AC service? Of course, it is possible to use transformers to meet other electrical requirements, and one can derive DC circuit from AC. The network analogs are gateways, digital computers specially designed to provide for arbitrary translations of one networking technology or protocol into another. Even with network gateways, it is not generally possible to have all of the functionalities translated across arbitrary network boundaries. Users, especially faculty, should be realistic and follow the campus's networking standards. Whatever standards are adopted, they should aim to be sufficient for the services to be provided. Networking services, however well intentioned, cannot possibly be all things to all people. Some limitations are necessary. o Determining the design and implementation of campus-wide, local area, regional, and national (and international) networking. The campus network will not be an island. It will be connected to networks in the local community, if only to the telephone system. Regional networks have developed within and across many states. Both private and public agencies are offering national and international networking, and there are offerings from a variety of national telecommunications companies. There is also the wildly successful Internet and BITNET. Therefore, it should not be a question of if the campus network is connected to other networks, but how it is to be done. o Determining the degree of integration of technologies for the five different families of communications services that are relevant to strategic planning for networking: voice, based on the principles of telephony; video or television, including CATV; data, as in digital computer originated, coded information; telemetry, as in information related to status of the physical environment; and control, as in activation of remote equipment, the measuration of phenomena at a distance, and the feedback to keep a process within acceptable limits. The campus network can be composed of five separate networks, each in its own track, or two or more of these families can be integrated to accomplish a set of objectives. A number of issues are present which bear on the question of what degree of technological integration is desired. First, all five of these communications services are going to have underlying digital technologies. This is a substantial change, of course, for those communications services that have been using information in analog form. Second, considerations of the space occupied by communications cabling, plus the need for higher performance levels, indicate that fiber optic cable will become the universal medium for all families of communications services. Fiber cables have better characteristics for avoiding interference from electrical force fields and are less susceptible to invasions of privacy. The initially higher cost of fiber is decreasing rapidly. Third, one does not want to switch all of the communications traffic. As yet, telephone-type switches are not designed for the high data rates of computer-type traffic. Consideration of the performance, i.e., speed, of the data flow (2400 bits per second for voice versus ten megabits per second for computer data) implies that a different type of technology is needed to address telephones versus workstations. Also a consideration is the fact that the typical telephone conversation is only a few minutes in length, whereas the typical computer session lasts over thirty minutes. This difference in "off- hook" times means that the switch designed for optimal telephone traffic is not the same as one designed for optimal computer traffic. Newer "gigabit" switches are being designed at this time based on Asynchronous Transfer Mode (ATM) technology, so that there is likely to be a reconciliation of voice, video, and high-speed data in our future. Considerations of load on the network must be taken into account, so that one type of service does not ruin some or all of the other types of services. A fully integrated voice-video-data network could degrade or even fail if some small number of high-performance users were to initiate bulk data transfers. The design of the network should provide for the segregation of large users of the data service. Since monitoring of the load on the network may not be adequate to prevent loading problems, we favor the installation of extra cabling to handle large users. The user community should inform the network services department as individual usage profiles change. On most campuses, there is an independence of management of the various communications services. Bringing them all together probably involves a major administrative reorganization. If computer services and video are on the "academic" side of the organization and voice, telemetry, and control are on the "administrative" side, then it usually means that several vice-presidential level persons will be involved in the integration. Such reorganizations maybe difficult to implement, yet communications vendors are increasingly expecting universities to act in a singular, coordinated manner. One suggestion is that the university create the chief information officer position to integrate the various communications services. Lastly, we have been witnessing the separate evolution of the various communications technologies. As they all become digital, they may very well evolve more synchronously, and at a pace dictated by improvements in integrated circuitry. If voice services becomes peripheral to the digital computer base, then we may see a high degree of integration very soon. Clearly, more progress will need to be made on voice input and output units, and the costs must decrease before the personal computer and the telephone will truly merge. Digital television exists, albeit it is very expensive. We believe that digital television is the wave of the future, and that the mass production manufacturing scenarios from the 1950's and 60's that made analog televisions cheaper will be replayed. Television signals in commercial service have been carried over fiber optic cables since 1984. How to fund the network? In our experience, the campus-wide network needs two types of funding strategies, one for capital and the other for operating. The capital funds are one-time funds for constructing the network. To network our 85 buildings, we will have to place some 11,000 network outlets, mostly in older buildings; it will take five years to implement and cost about $18 million. We have raised the funds specifically for the retrofitting of existing buildings as part of a general university-wide capital campaign and as an integral part of the construction budget for each new building. All buildings follow the newly established standards for campus-wide network wiring and conduiting. For CWRUnet, the average cost as expressed in terms of the cost per wall outlet is $1,650 per faceplate. This is an all-inclusive cost (cf. reference (1) for more details). Buildings are renovated from time to time, and the cost of reestablishing CWRUnet services in the renovated space is another construction cost analogous to wiring for electrical service. Because the CWRUnet cabling was designed for a long period of usage, we have found that we can reuse it in the renovated spaces. For all types of network-related capital expenditures, one should have a set of policies to cover the depreciation of this asset. At CWRU, we built the networking capital expense into the building's capital value and depreciate the total value with each type of depreciation following its own schedule. The accumulation of depreciation funds creates a funding stream for the eventual replacement of electronic and non-electronic network components. The electronic components will often need upgrading over a five-to-eight year cycle; the use of fiber optics cabling allows for extended depreciation cycles of 20 and more years. Each institution should explore the use of building depreciation finance policies to generate on a continual basis funds sufficient to renew the critical networking infrastructure. CWRUnet also requires an operating budget. With all of its services, the cost per faceplate is budgeted at $117 for the 1993-94 academic year. Each building has a designated number of installed faceplates, and this amount per faceplate is factored into the operating costs for the communications utilities of each building in the same manner as we apply costs for electricity, steam, and water. We do not charge for the amount of information actually transmitted over CWRUnet; there is no charge analogous to telephone "message units." Therefore, CWRUnet does not follow the telephone utility model; we use a library-like model that apportions costs according to access, not usage. We have watched how people actually use networks, and we see that usage is often quite unpredictable. A charging scheme based on the message unit concept penalizes the user who works in a "discovery" mode, as contrasted with users who preplan each network activity. Therefore, we have adopted the library budgeting model because network usage involves substantial unplanned activity, much as library usage in browsing the stacks involves books and journals you may not have planned to use. At CWRU, we see the network as another major utility on the campus, being like the electrical service in its ubiquity and like the library in its impact on academic and social life. References (1) Neff, R.K., and Haigh, P.J.: "CWRUnet--Case History of a Campus- wide Fiber-to-the-Desktop Network" CAUSE/EFFECT, 15 (2), 26-29, 33-38, Summer 1992. (2) Watkins, B.T.: "Information Technology: University Hopes Campuswide Network Will Help Give It a Competitive Edge" The Chronicle of Higher Education, April 29, 1992. (3) Haigh, P.J.: "LANS and Fiber Optics: Building a New Infrastructure in Higher Education" Higher Education Product Companion, 2 (2), 8-14, January/February 1993. (4) Elmer-Dewitt, P. "Take a Trip into the Future on the Electronic Superhighway" Time, April 12, 1993.