The Electronic Kiosk: Interactive Multimedia Goes Enterprise-Wide Copyright CAUSE 1994. This paper was presented at the 1993 CAUSE Annual Conference held in San Diego, California, December 7-10, and is part of the conference proceedings published by CAUSE. Permission to copy or disseminate all or part of this material is granted provided that the copies are not made or distributed for commercial advantage, that the CAUSE copyright notice and the title and authors of the publication and its date appear, and that notice is given that copying is by permission of CAUSE, the association for managing and using information technology in higher education. To copy or disseminate otherwise, or to republish in any form, requires written permission from CAUSE. For further information: CAUSE, 4840 Pearl East Circle, Suite 302E, Boulder, CO 80301; 303449-4430; e-mail info@cause.colorado.edu The Electronic Kiosk: Interactive Multimedia Goes Enterprise-Wide I. INTRODUCTION The University of Texas at Austin is in the early stages of deploying a series of electronic Kiosks designed to support a variety of campus-wide information services. These Kiosks are similar in concept to automated teller machines (ATMs) in that they are publicly available computers linked to common servers via network connections and are housed in enclosures protecting the equipment from damage or theft. While these Kiosks will not dispense cash, they do deliver a wide range of information from different sources and employ interactive multimedia as the user interface. All major elements of the Kiosks, from the software to the physical enclosures, were developed in-house at The University of Texas. The focus of this presentation is to relate the strategies employed in the initiation and management of this project and describe how it is being integrated into the existing information technology infrastructure. Original Project Description The Kiosk project is a joint development effort between The University of Texas at Austin and Apple Computer. The project was initiated with The University's Data Processing Department in 1992. In addition to the development of the Kiosks themselves, the stated purposes of the project included demonstrating the effectiveness of interactive multimedia as user interface and producing portable software tools which could be used by other institutions to build similar Kiosks. The functions which have been developed for the first version are: * A Multimedia Introduction to The University of Texas at Austin * A Campus-wide Events Calendar * An Interactive Map of Campus * Maps of Shuttle Bus Routes * A Directory of Students, Faculty and Staff The multimedia introduction makes extensive use of digital video (QuickTime) to provide visitors and prospective students and faculty with background on the history and traditions of The University as well as an orientation to campus and essential facts for visitors. The Student, Faculty, Staff Directory and the Events Calendar provide real-time access to information from The University's corporate database on the Administrative Computer System. Both the Events Calendar and the Directory are linked to the interactive map so that campus locations of people and events can be displayed directly from those lists. The map may be searched by a variety of criteria ranging from building name to department, and photographs of buildings are displayed whenever they are selected. In addition to the digital video described above, the major elements of interactive multimedia employed in the Kiosk include color graphics, digitized photographs and other images, sound, and touch screen monitors. The touch screen technology enables users to indicate selections by simply pointing. The project's goal is to make the system simple enough to be used by any adult without training or assistance. Survey results from early prototypes and other similar projects around the nation indicate that this objective will be achieved. Multimedia also enables the delivery of types of information not easily conveyed through text (e.g. the maps.) The Kiosk client machines are Apple Macintosh Quadra 840AVs. The Kiosk software is primarily written in HyperTalk with numerous external functions in other languages to perform some of the communications and utility functions. Communications with the Administrative Computer System are handled via 3270 front-end using Software AG's (SAG) Natural Connection and its HyperCard API. The 3270 front-end is intended to be temporary as work is currently underway to convert to the SAG ENTIRE family of products enabling true program-to-program communication with the Administrative Computer. Likewise, although the QuickTime movies are currently stored on CD ROM with drives local to each machine, real time access to networked video servers is planned for eventual implementation. One of the primary goals of software development for the first version of the Kiosk has been the creation of a developers' tool kit. The developers' tool kit includes a package of standard routines to process common functions, technical standards, documentation, and classes all of which are intended to enable individual administrative offices to produce their own Kiosk services as efficiently as possible. As of this writing, the Office of Personnel Services is the first department outside the core Kiosk Development Group to use the tool kit. They are creating an available job listings function. This paper is, in effect, a case study on how The University tapped available resources, built on the existing information infrastructure, and leveraged alliances to create its own unique multimedia campus- wide information system. II. DEFINING OBJECTIVES Creating a True Campus-wide Effort Soon after the project began, the Data Processing Department determined that the Kiosk should be managed and developed as a full campus-wide effort. This direction was both logical, since the Kiosk was intended to be a campus-wide information system, and necessary, because the expertise required to build and manage such a system did not reside in any single department. With separate departments for academic computing (Computation Center) and administrative computing (Data Processing Department) at The University, overall responsibility for information technology infrastructure is relatively decentralized. Although institution-wide technology projects have become more common, there was little precedent for providing for the ongoing development and management of an information system which crossed academic and administrative lines and potentially served all students, faculty, and staff. To fill the void, the Data Processing Department sponsored an effort to form an ad hoc Campus-wide Steering Committee. The Steering Committee has no formal standing or authority with in The University structure, but it does have the practical effect of providing guidance from the perspective of most major campus constituencies. The positions represented on the steering committee are: Assistant Dean Graduate School of Business Associate Director Recreational Sports Assistant VP Development Office Assistant Director Division of Housing and Food Associate Professor Faculty Computer Committee Associate VP Student Affairs Associate Director Computation Center President Students' Association Assistant Director Library Registrar Registrar's Office Auditor Internal Audit From the beginning, both the Data Processing Department and the CWIS Steering Committee viewed the Kiosk project as one element in an emerging direct service strategy for administrative information systems. This perspective refined the specific objectives of the project and shaped it to complement other CWIS systems. The Direct Service Strategy The concept of a direct service strategy simply referred to the trend in administrative information systems to provide services directly to the ultimate consumer without the need for intermediaries. During the 1980's, The University had automated almost all of its significant administrative procedures from accounting and purchasing to student records and financial aid. For services used by students and faculty, however, most of this automation was designed around a model that involved intermediaries. Specifically, administrative personnel were typically the direct users of the computer systems, and they in turn exchanged information over the counter or via telephone with students and faculty. By the 1990's, striking examples of a new "direct service" model were evident everywhere, both inside and outside The University. Automated Teller Machines (ATMs) are probably the most common and widely used examples of this direct service paradigm. On The University of Texas campus, two of the most successful information technology projects of the last few years have been UTCAT, the On-line library catalog system, and TEX, our telephone voice response registration system. Both of these projects clearly demonstrated that providing information services directly to students, faculty, and staff could work the apparent marvel of improving service while containing costs. Both the Data Processing Department and the CWIS Steering Committee sought to articulate a generalized direct service strategy that would enable any administrative office to provide electronic information services efficiently to students, faculty and staff. From a technical perspective, primary objectives would be 1) to develop a technical infrastructure that would enable departments to provide new information services quickly and inexpensively, and 2) to make these services as accessible as possible for the customers. From the perspective of administrative offices, the goal is to have a cost-effective method for providing routine information services to their clients and free staff from a substantial portion of the "counter service" workload. From the customers' perspective, the goal is to reduce University bureaucracy by providing convenient "one-stop" shopping for many common administrative functions. In short, we wanted to provide the ability for all university offices to achieve on a generalized basis what TEX, UTCAT, and ATMs had done for their specific functions. We wanted a unified campus-wide information system that would be a cost effective information delivery vehicle for all information providers and an easy-to-use single system image for all information clients whether they were students, faculty, staff or any other constituents of The University. Kiosk as Part of a Multi-prong Approach By the early Fall of 1992, the discussion on campus about direct service systems shifted from defining goals to implementation strategies. The ideal would obviously be to have a single campus-wide information system. Everyone involved in the discussion wanted to avoid a situation in which individual departments were developing numerous single-purpose information services. Such a balkanized information environment would produce a confusing array of choices for users and an inefficient use of resources by The University. At the same time, it was clear that a single information system capable of meeting all requirements was impractical in the near term under existing budgets. So, a compromise plan was devised under which a limited number (primarily three) of campus-wide information umbrellas would be encouraged. Each of these systems focused on a different needs, built on existing technologies, and provided open architectures within their defined limits. The first CWIS, UT's implementation of the Internet Gopher, was already in existence and operating successfully. Gopher is supported by the Computation Center and is used primarily for academic purposes. It is not yet feasible to place sensitive, real-time administrative functions under the Gopher menu, so the Data Processing Department set about to create a campus- wide information system umbrella that would fill administrative needs and complement Gopher. In addition to the Kiosk, Data Processing is supporting the development of a public access application, called UTACCESS, to provide campus-wide information services from the Administrative Computer System (see "Building on Previous Projects" below.) Strengths and Objectives of the Kiosk Concept The Kiosk was quickly accepted as a potentially important element in a strategy of providing administrative services directly to students, faculty. It is, however, definitely seen as complementing network-based services which are accessed from users' own microcomputers in offices, homes, and dorm rooms; and not as supplanting such services. Serving as one component of an overall strategy, the Kiosk concept has many strengths. First, there is no requirement that the individual user own or have access to his or her own computer workstation. The entire information delivery infrastructure is provided by The University. A further technical advantage of institutional ownership of the workstation is that developers know the exact hardware and software configuration of the client machine and can optimize the system for those requirements. Another advantage of the Kiosk concept is that it is a true client/server system and can provide information services from a variety of services. The Kiosks are extremely convenient to use. The user is not required to go through any logon procedures (although PIN numbers will be required for personal information) or in any other way set up the machine before it can be used. Kiosks make it easy for anyone to access information quickly as they move around campus irrespective of whether they own or can use a networked personal computer. The Kiosk is ultimately intended to bridge the barrier between administrative and academic resources through a series of networked public workstations that draw information from a variety of sources. III. STRATEGIC ALLIANCES: GATHERING RESOURCES "Don't try to do more with less, and don't do less." Carole Barone, Associate Vice Chancellor for Information Technology at UC/Davis, speaking at CAUSE92. From the beginning, the Kiosk project has sought to follow Carole Barone's advice and increase available resources by pursuing unconventional approaches to funding and in-kind support. This strategy was not an exercise but a necessity. The University of Texas is no exception to the national trend toward tighter budgets. For the last several biennial sessions of the Texas Legislature, state-funded institutions of higher learning have witnessed ever strengthening movements to freeze and even reduce their allocations as the state battles to balance its own budget in difficult economic times. These budget constraints are coming at a time when overall administrative workload is expanding, and expectations for quality of service are increasing. In this environment, the Kiosk project would not have been possible, either financially or technically, without the extensive use of alliances and partnerships. Campus-wide information systems are by their very nature projects which span The University's organizational hierarchy. No single entity has the resources or authority to undertake the challenge alone. Internal Alliances: Building on a Base of Cooperation Several important informal processes began on The University of Texas campus in the early 1990's which greatly facilitated the joint efforts necessary to undertake campus-wide information systems. The General Libraries initiated formation of an informal "Information Services Discussion Group" which included management representatives from the Computation Center, Data Processing, Student Affairs, and General Libraries. Although the group has taken no action to date, a number of important cooperative efforts have arisen out of the monthly discussions of institution-wide information technology challenges and solutions. The Direct Service Strategy for administrative campus-wide information systems was first discussed and refined in ISDG. Prior to the start of the Kiosk project, the staff of the Data Processing Graphics Center had sponsored the formation of a campus-wide users group for graphics developers. The membership of this group included many of the most talented and accomplished graphic artists and multimedia developers on campus. When the Kiosk project started, a presentation was made to the graphics group introducing them to the project and requesting their input. Many of the members were immediately enthusiastic about the potential for the project and over time participated directly in the design and creation of the graphics and other media used in the Kiosk. At least four members of the graphics users group became active participants in the Kiosk Development Group. Forming a Campus-wide Development Group The core Kiosk Development Group was staffed by the Data Processing Department, but DP had neither the available personnel nor all the expertise required to complete every phase of the Kiosk. The core Kiosk team structured the projectas a campus-wide development effort and actively sought assistance from other University offices. The response to these appeals was overwhelmingly positive (see below). In all cases, this support was voluntary. At least three general factors motivated the enthusiasm and spirit of cooperation that characterized the Kiosk Development Group. First, the project was positioned both as a showcase for University talent and as a rare opportunity to create something in which the entire institution could take pride. Secondly, many members viewed the project as an opportunity to work with new technologies and learn new skills. The College of Pharmacy Staff indicated that their considerable commitment of time in the Kiosk was repaid by the skills they acquired in the production of QuickTime videos. That expertise is already being applied in many of their academic projects. Finally, the Kiosk became an excellent vehicle for University service. The Architecture professor who designed the physical enclosure stated that he appreciated the opportunity to make a professional contribution to the University environment. Listed below are some of the contributions made to the Kiosk project by University departments. * Graphic artists from several departments, such as UT Publications, volunteered their services to assist in the creation of color art work and backgrounds used in the Kiosk. * The staff of the College of Pharmacy Learning Resources Center assumed primary responsibility for the design and production of the video, animation, and graphics used in the multimedia "Welcome" section of the Kiosk. * The Manager of the Computation Center's Micro Technologies Group taught classes in HyperTalk to Kiosk developers. * The School of Architecture donated the services of one of their faculty members to design the Kiosk physical enclosure. * A programmer on the Computation Center Staff wrote several of the external utilities not producible in HyperTalk. * Programmers from General Libraries, the Computation Center, Apple Computer, and Data Processing participated in writing HyperTalk code. IV. INTEGRATING INFORMATION TECHNOLOGIES The electronic Kiosk is not an isolated, independent project, but rather an integral part of The University's overall information technology environment. The Kiosk is built on the foundation of existing applications, databases, networks, and communications. It is designed to complement and extend both current and future information technology projects. Tapping into the Existing Infrastructure The focus in administrative computing at The University during the 1980's was automation of official administrative procedures. Electronic documents have now replaced paper forms to hire and reclassify employees, purchase anything from office supplies to laboratory equipment, transfer funds, pay vendors, and process internal billing. These electronic documents automate virtually every aspect of the manual procedures they replace, including auditing of routine data, routing for review and approval, elimination of duplicate data entry, updating of accounting records, and storage of official information. Almost all of The University's official records are maintained in electronic form on a single database management system that is easily accessible by authorized personnel. In 1980, the Data Processing Department adopted Software AG's DBMS ADABAS as its official DBMS. Since that time, ADABAS has evolved into the central repository for virtually all of The University's official electronic records. The University of Texas now has a single, unified corporate database which supports the relatively easy development of real-time, integrated applications. During the same period The University has also constructed a comprehensive campus-wide network system. In fact, two essentially parallel wide area networks exist on campus to serve the different requirements of administrative and academic computing. The administrative network, UTAN, is a fiber optic SNA token ring network that provides 16 Mbps service to all administrative and most academic building on the UT main campus. UTAN is a controlled access network providing a high level of security and capacity for confidential administrative traffic. UTAN is primarily used for 3270 traffic between the Administrative Computer System and departmental LANs. UTAN is supported by the Data Processing Department. The University's academic network, UTnet, is a fiber optic TCP/IP network providing open access to a full range of Internet services primarily for academic purposes. UTnet serves all major academic buildings on campus and is accessible via departmental LANs, microcomputer labs, and a high speed modem system called TELESYS. UTnet is supported by the Computation Center. The current Kiosk prototypes are connected directly to the administrative SNA network for secured access to administrative services via the UTACCESS application (see below.) A joint project between the Computation Center and Data Processing is now under way to develop the encryption software necessary to provide secure access to administrative data over the Internet. Once this project is complete, the Kiosks will be moved to UTnet. Building on Previous Projects Much of the underlying technology used in the Kiosk was pioneered in earlier projects at The University. In particular, it would not have been possible to produce the Kiosk economically without the benefit of software and techniques developed as part of the Executive Workstation and the UTCAT online library catalog. The Executive Workstation is a Macintosh-based executive information system developed within the Data Processing Department to give management level personnel point-and- click access to official administrative functions. In addition to providing a GUI style interface to mainframe functions, the Executive Workstation gathers and assimilates information from several different applications to deliver summary displays which are not available in any other system. The Executive Workstation is a HyperCard application which communicates with the Administrative Computer System via an API to a 3270 emulation session running on the Macintosh. The Kiosk, also developed in HyperCard, employs 3270 emulation communication with the Administrative Host which is based on the code developed for the Executive Workstation. The UTCAT online library catalog system is a mainframe- based 3270 application which was developed in-house. In recent years, UTCAT has been used as the foundation for an expanded menu system, called UTCAT Plus. UTCAT Plus is essentially a campus-wide information system incorporating not only Library information, but also administrative policies, The University telephone directory, and available job openings. Access to UTCAT Plus is available to anyone via terminals in the UT libraries, by connection to the Internet, and by dial-up to the UT TELESYS system. The Library has contributed the UTCAT Plus core software for menu management, navigation, and access to be used as the basis for the UTACCESS project. From that starting point, UTACCESS has become an independent CWIS in its own right which provides easy access to information services on the Administrative Computer System. UTACCESS is also the 3270 system that provides official administrative information to the Kiosk. As with the contribution made by the Executive Workstation, UTCAT Plus paved the way for the Kiosk. Coordination with Complementing Projects The Kiosk has also achieved a level of cost savings and efficiency by coordinating closely with separate but related projects on campus. For example, a task force is working on producing an RFP for a campus-wide image server. The goal of the Image Server Task Force is to acquire imaging technology that is compatible with existing information technology infrastructure and can be implemented on an institution-wide basis. The task force is working to ensure that the Imaging products can be integrated smoothly with the administrative corporate database (ADABAS) and our administrative programminglanguage (NATURAL 2) so that images can be readily included in existing and future administrative applications. Another project with direct implications for the Kiosk is the Campus-wide ID Card project. There, yet another task force is generating a recommendation for a single, multi-purpose ID card that can serve all needs on campus for student, faculty, staff identification. The project is based on the principle that a credit card-like ID with a standard, machine-readable magnetic stripe can be implemented to address the full range of identification requirements of all University offices. In addition to the traditional manual identification using the photograph and signature, the new card can be "machine read" by electronic card readers. Through network connections, the machine card readers can be fully integrated in with existing databases and electronically validate requests against existing information systems. The Kiosk Development Group communicates regularly with both the Image Server Task Force and the ID Card group to ensure that the technology in all three is fully compatible. The benefits to the Kiosk are clear. The availability of institution-wide, network-based image servers mean that the ability to display graphics such as images of buildings, floor plans, and maps will soon become a practical option. The ID Card project will enable the Kiosk to serve as a true ATM for University services that require positive identification and, in some cases, charge for services. For example, in the future, students may be able to buy concert tickets or pay library fines and have those fees charged against the debit card stripe on their campus-wide ID cards. V. MANAGING A CAMPUS-WIDE EFFORT Role of the Steering Committee Since its formation, the Campus-wide steering committee has met at irregular intervals and communicated frequently through e-mail to oversee overall project management, establish policy, and set priorities. The Kiosk project has benefited enormously from the experience and diverse backgrounds of committee members. They have been able to provide crucial advice and guidance on how to tailor systems for students. This was an issue of particular concern within the Data Processing Department since, until recently, our focus has been on services for administrative staff. We had not acquired the expertise of many of the offices represented on the Steering Committee in serving the large student population. Below are examples of some of the issues handled by the Steering Committee. * Selection of the on-campus sites where the Kiosk will be located * Establishing policy on the types of services considered appropriate for delivery via the Kiosk * Review of ADA compliance measures * Review of testing and quality control measures * Ensuring all effected campus constituencies are consulted regarding significant implementation issues * Review and approval of implementation procedures and schedule * Communications with campus constituencies and Executive Officers Getting Input from 70,000 People The Data Processing Department has generally employed users groups as the chief mechanism for promoting customer participation in the process of designing and refining administrative information systems. Traditional users groups were not practical, however, for a potential customer base of all 70,000 students, faculty, and staff at The University. Other techniques had to be found to evaluate how well the Kiosk would perform and be received in real-world situations. The techniques chosen were those now commonly used in the testing and evaluation of all types of commercial products which are intended for mass markets: focus groups, customer surveys, and observation. Two successive prototypes of the Kiosk software have now been produced. Each prototype has been subjected to focus group analysis. The feed back from these groups has served as the basis for refinements incorporated into the next version. The focus groups for the first prototype were structured, but generally informal and not conducted according to strict scientific method. Specifically, the Students' Association was invited to send up to eight students each to twenty different sessions scheduled shortly after completion of the first prototype. No effort was made by the Kiosk Development Group to ensure that statistically random groups were selected. Each group met for 90 minutes in which they were given the opportunity to use the Kiosk, participate in a panel discussion, and fill out a survey questionnaire. The Development Group then generated a package of changes and enhancements for the second prototype based on their own observations of focus group participants using the first prototype, the suggestions made in the panel discussions, and the results from survey forms. The evaluation of the second Kiosk prototype was significantly more structured and thorough. Once the second prototype was ready, it was set up for public access on an information desk in the student union. This unenclosed unit was used as the base for several different types of evaluation. At the same time a class in the Marketing department had requested permission to use the Kiosk as the basis for a class project on product testing. The project was approved, and the Marketing students have since conducted formal surveys of students who have used the public prototype. The class has also conducted telephone surveys of students and administrative offices to determine what types of services the campus would most want to see on a Kiosk. Members of the Development Group have periodically spent one to two hours a day unobtrusively watching students use the prototype. These observation sessions have been used successfully to identify elements of the Kiosk user interface that are confusing and difficult to use. The observations have served as the basis for a second round of refinements, and the Marketing class surveys will be reported to the Steering Committee for its consideration. VI. CONCLUSION It is too early to predict what overall impact the electronic Kiosk will have on University operations. At the time of this writing, two unenclosed prototypes are available for public access. These units are being used in part to assess the effectiveness of the overall Kiosk concept and to measure the costs associated with their operation. Bids are being taken for the construction of the first five Kiosk enclosures. These units will be installed in the early part of the 1994 Spring semester. The remaining 12 enclosed Kiosks and possibly more will be deployed as soon as the performance of the first five has been thoroughly evaluated. Many important questions remain unanswered. Will we be able to fund and support enough units to have a significant impact on a campus community of 50,000 students and over 17,000 faculty and staff? Will administrative offices view the Kiosks as a cost- effective mechanism for delivering routine services? What will the operational and maintenance costs associated with the Kiosks be? Does the Kiosk concept represent a viable long term solution for delivering administrative information services in a large University environment? What is the Kiosk's logical relationship to information services delivered directly to user-owned client computers over the network or telephone lines? It is reasonable, however, to draw at least four significant conclusion at this stage. * Interactive multimedia technology has advanced to the point that it can be used to produce practical information systems which are both easy to use and well received. * Multimedia Kiosks make it realistic to provide public information services that almost anyone can use with no prior training. * Kiosks alone do not constitute a comprehensive campus-wide information system, but they can play a vital role as one element in an overall CWIS strategy. * The concept of a multimedia CWIS at The University has attracted sufficient enthusiasm, support, and resources to produce a competent first version.