Design and Implementation Considerations for an

Interactive Multimedia Kiosk: Where to Start

An information kiosk is a public stand that supplies text, graphics, video, animation and sound information to the user. The application is an interactive multimedia program which can run on a local computer or via an intranet or internet connection within the enclosure.

The very nature of an information kiosk demands fast, accurate information delivered via a user-friendly interface. Design not only encompasses the normal software development cycle to produce an interactive multimedia program, but there are also additional issues unique to a kiosk situation.

This paper aims to provide advice to those new to design and development of information kiosks based on experiences gained creating an interactive presentation for the Department of Mathematics & Computing at Central Queensland University (CQU). It elaborates on some of the constraints experienced during this project (both hardware and software) in the hope that others may avoid the same problems.

1.0 Background

Kiosks are used worldwide at many university campuses for campus maps, libraries, staff and course information. The advantages of using an information kiosk at University Open Days and similar functions are:

• the public can query the system and determine information requirments specifically for their needs, and
• staff are released for other duties.

CQU is a multi-campus regional university with the main campus located at Rockhampton on the Tropic of Capricorn. Other campuses are located in Mackay, Emerald, Gladstone, Bundaberg, Sydney, and Melbourne. CQU has a strong

commitment to distance education, and has overseas operations in Singapore, Fiji, and Dubai.

In 1995 a pilot project developed an interactive multimedia kiosk to present information about CQU's Department of Mathematics and Computing [1]. The interactive multimedia program featured text, video, graphics, animation and sound files.

The project plan followed conventional lines: to research the area, followed by a design phase, development of resources, programming, user testing and implementing the resulting changes, followed by documentation (project report and technical information). While there is an abundance of information on multimedia programming, at the time of development it was difficult to find many references to design issues for information kiosks. By mid-1995 this situation had changed a little; by the end of 1995 Silk and Hodgson [2] reflected many of the issues experienced in our project. Information on kiosks is more prevalent on the World-Wide Web (WWW) today with commercial firms offering to develop information kiosks for clients.

In 1996 university funding was provided to develop a standard prototpye information kiosk for CQU based on the departmental pilot. This support affirmed both the suitability of the pilot design and CQU's recognition of the benefits to be gained by information dissemination in this manner across a multi-campus university.

This paper relates the experience of a kiosk within a university environment. Other developments may discover factors that are more or less important, depending on the specific application.

2.0 The 1995 Pilot Project Experience

The pilot project incorporated the following steps:

• determining work already reported on information kiosks,
• investigation of touch screen technology,
• design of touch screen user interface,
• review of development tools,
• program design,
• program development, and
• user testing.

Each of these is now addressed in detail.

2.1 Previous Work

Much research has already been undertaken in multimedia programming. It became obvious early in this project that putting a multimedia program together requires a wide range of skills [3] [4] [5] [6]. A balanced project team would usually consist of a content expert, a graphic designer, a programmer, a video producer, a sound technician, a project designer as well as someone with management skills: typically these skills are not all found in one person.

A search using the WWW revealed little information regarding information kiosks at the beginning of 1995. However within 6 months this had changed dramatically. By June 1995, commercial firms offering to create custom multimedia kiosks for clients were advertising on the Internet and including sample screens at their WWW sites [7]. In addition, "plugin" software was freely available to allow programs developed using commercial authoring systems to be viewed via WWW browsers (e.g. Shockwave for programs developed using Macromedia Director authoring language and MM-WWW-PC for programs developed using Asymetrix Multimedia ToolBook 3.0 authoring language (now Neuron for Toolbook II), etc.).

2.2 Touch Screen Technology

Worldwide, an increase of 25% per year in the use of touch screens is predicted [8]. Vendors identified five groups of touch screens uses: point-of-sale, information kiosks, industrial, gaming, and business applications. Of these, the business and education markets were slowly picking up the technology while point-of-sale and information kiosks were adopting touch screens at a very rapid pace. Industrial applications were also on the increase, aided by improved ruggedness of the touch screen and lower manufacturing costs. Importantly users with little or no computer literacy do not feel intimidated by the technology.

Chernich [9] provided details of touch screen technology and compared the different methods implemented. Three technologies have been widely used for touch screens: infrared, capacitative, and piezo-electric. The disadvantages of using infra-red technology are many including: low reliability; susceptibility to foreign body interference; and a high degree of parallax error. The main disadvantage of a capacitive touch screen (as listed by Chernich) is that condensation can form between the screen and the overlay. However the Australian supplier for the capacitive touch screen used in this work guaranteed their product and, in fact, included the 'non-condensing' feature in their literature. Piezo-Electric touch screens are more reliable than a capacitive system however the cost was much higher. There were two main reasons for the decision for a capacitive touch screen: price, and location of an Australian supplier.

2.3 Touch Screen User Interface

Design of a 'walk-up-and-use' information kiosk requires a simple interface that takes into account the computer skills and the educational level of possible users [10]. These skills range from the complete novice to those familiar with using a computer. The program is considered successful if the user is able to accomplish the goal of obtaining information required in a minimum amount of time without experiencing trauma with the technology.

Basic screen design for the current work was greatly influenced by Russell and Redhead [11]. Good balance was achieved while incorporating a variety of layouts to avoid boredom and yet still maintaining a familiar screen environment throughout the program.

Merrill [12] states that the user should always know how to exit the program. This is true with a conventional program but for information kiosks the user does not have to 'exit the program', they merely walk away when they have finished.

Colour is an important element in the design considerations, and Perry [13] provided the idea of a different colour scheme for each branch of the program. Russell and Redhead [11] gave guidelines for colour combinations and reported that users favour blue, green, grey and black combinations. While other colours were used, a basic blue/green theme was implemented in the current work.

Cheung [14] discouraged the use of many levels within a hierarchy to allow the user to jump around at the same level. His advice of a regular hierarchical structure has been followed to allow easy navigation in the current work.

Silk and Hodgson [2] report their campus information kiosk was programmed to return to the opening menu after a period of 5 minutes of non-use. They further included screen-saver software to activate after a period of 20 minutes of the main menu being displayed continuously. The presentation developed at CQU follows this model and will return to a "loop" of screens after a period of 5 minutes of non-use. Presently the loop consists of just 5 screens but could be extended to many more, delivered in random order to attract attention and maintain interest. The loop will play continuously until a touch of the screen returns the program to the 'Main Menu'.

2.4 Development Tools

The original intention was to use Asymetrix Multimedia ToolBook 1.53 (MTB1.53) as the authoring tool. However a free upgrade to MTB3.0 was available at that time. A decision was made to wait for the upgrade as reports of the later version were most favourable [15]. This time delay proved costly to the project. In the meantime development began using Asymetrix's full-featured graphics presentation program, Asymetrix Compel 1.0. The prototype presentation was developed using Compel 1.0.

2.5 Program Design

An aim of the project was to create an application that would be generic for any location of the Department.

The 'film type' storyboard was developed taking into account feedback from the meetings with the client (a departmental working party for the recruitment of undergraduate and postgraduate students). This approach seemed a natural way to create a flow chart for the program on paper. The major advantage in this type of plan is that all the screens can be seen at one glance to determine the logical sequence of the presentation.

All regular navigation buttons were placed at the bottom of the screen to ensure that the user knows how to proceed. A 'Help' button was provided along with a prompt bar showing actions available to the user. As an example, the Main Menu screen of the presentation is shown in Figure 1. This screen is indicative of design features used throughout, based on the research results. A design decision to remove navigation buttons not applicable to a particular screen ensures that all options presented are active. This avoids confusion by not presenting 'greyed out', non-active buttons to the user.

As well as the navigation buttons, the presentation includes menu options, hotwords, and active areas on the screen (usually a photo) which access more information. This is explained in detail to the user in the 'Help' branch of the presentation.

Figure 1: Representative Screen Design used throughout the Presentation

2.6 Program Development

Delays were experienced both with the purchase of a touch screen for the Department as well as delivery of software upgrades. However, this time was spent in the organisation of resources to be used in the information kiosk presentation. As well, familiarity with several software packages required to collect and edit those resources was acquired.

2.6.1 Collection of Resources

Great care was taken with copyright considerations throughout this project. It is believed that there is no infringement of copyright in any of the resources presented in the information kiosk presentation.

Video sequences included in the information kiosk need to be a coherent part of the multimedia program. Whalley [16] advised that each sequence of film needs to be understood in isolation. It is generally accepted that the ideal length of video clips used in a multimedia presentation is between 10-20 seconds. A studio was used for most clips, but several amateur recordings using hand-held cameras were satisfactory.

Video clips, sound tracks, and graphics were all manipulated to produce suitable resource files (.AVI, .WAV, and .BMP files respectively). Many software programs were used to produce the final results (including AdobeCap Video Capture Utility, Corel DRAW and Corel PHOTO-PAINT, Paint Shop Pro, PhotoStyler and Creative WaveStudio).

Graphics files were produced from a variety of sources, namely

• digital camera images,
• capture of a single frame from a video tape,
• photos scanned using a flatbed scanner,
• slides scanned using a slide scanner,
• creations using graphics packages, and
• sample graphics provided within Compel 1.0.

Graphics were used in a variety of ways throughout the presentation: in animation sequences, for visual representation of concepts, as ways of linking the user to further information, as backgrounds for screens, and included in built .AVI files. Screen resolution, image size, and purpose are of paramount importance when preparing graphics for use in a multimedia presentation [17].

2.6.2 The right tool for the right job!

During 1996 the prototype was ported to a more robust program using Asymetrix Multimedia Toolbook 3.0 authoring language (MTB3.0). A number of deficiencies (or difficulties) in the Compel version were rectified proving that the choice of software is critical - the right tool for the right job.

2.7 User Testing

User testing was performed at a standalone station organised for Open Day in August '96 at CQU. The program did not falter once despite the fact that it was thoroughly used by passers-by. One simple lesson learnt on the day was that the monitor needs to be firmly positioned at the correct viewing angle for the public.

3.0 CQU Campus Kiosk Project

Funding was sought in early 1996 from the University to develop a prototype campus information kiosk for CQU. The aim was to prove the concept prior to implementing the physical kiosks on a grander (wider) scale across this multi-campus university. A very small project team is involved who are presently working with the University's Interactive Multimedia Unit to develop the prototype program. Concurrently, the team is organising the purchase of an enclosure and system unit. The project is planned for completion early in 1997.

4.0 Design Issues

4.1 The User

An information kiosk should pique the curiosity of passers by, enticing them to spend a few minutes browsing the display. Once engaged, those passers by must be able to find information easily - no matter what their level of computer expertise. [18]

The question, "What information is the user seeking or expecting to find at the station?" is fundamental. After full investigation of the type of information to provide (including surveys of the target audience's need, and consultation with experts on the knowledge base required), a thorough analysis of the data is needed to decide the best way to structure information for retrieval by the user. One way of achieving this is to conduct a card-sorting exercise. The resulting structure must be thoroughly tested by intending users prior to implementation and modifications incorporated into the final program. Table 1 summarises the design issues involved in relation

to providing the correct information for the user.

A successful result of this phase is that the user will retrieve the information required efficiently and without frustration. This is central to the success of the whole system.

4.2 The Enclosure

A basic decision to purchase the skin of the kiosk or to build it locally is often related to financial considerations.

Table 2 summarises the issues involved when considering the provision of a physical kiosk.

4.3 Hardware

Internet access, or at the very least Intranet access, must be considered when designing a single physical unit. It is appropriate to take advantage of a client/server system to allow the programs running on the kiosk to be the front-end client with access to a database on a central server. Presuming that the central database is updated regularly, it follows that the kiosk will always be accurate with up-to-date information. If connection time to a server is too slow for comfortable user interaction, the database file could be stored locally and updated automatically at regular intervals (the number of times per day depends on the changing nature of the data involved - early hours of the morning is recommended for this type of housekeeping). Table 3 summarises some of the hardware issues for the kiosk.

4.4 Software

Information kiosks need an interactive multimedia program and software to provide an automatic return to loop or screen saver when the

kiosk has not been touched for a certain length of time. Careful thought needs to be given to monitoring the status of the system, log file generation to track users, and statistics generation for management and potential advertisers. Additionally, a built-in survey tool to collect feedback from users on their perceptions of the kiosk is useful.

The choice of programming language is very important and will depend on cost and availability of skilled programmers. The final program needs to be robust. As well, a major consideration is whether the program can run on the WWW, that is can be viewed using a Web browser.

Table 4 summarises some of the issues relating to software considerations when planning an information kiosk.

One final issue is whether the interactive multimedia program is provided as a runtime version or will the authoring program be necessary on the system unit of the kiosk? If the latter, provision will be required for a licensed copy of the program.

5.0 DISCUSSION

Currently, many universities in Australia and across the world have initiated or installed some type of across campus information kiosk system. These vary widely in both the style and the manner in which they have been developed. For instance, there is a consortium of universities in the USA who have contracted externally to provide a system of kiosks around campus and across multi-campus sites. Within Australia there have been prototype kiosks developed (ANU, CQU) as well as across campus systems linking physical kiosk with virtual systems on the WWW (e.g. Curtin Enews, Southern Cross).

A prototype information kiosk was developed for the Department of Mathematics & Computing at CQU in 1995. It is recognised that the time spent in the design phase of this project paid huge dividends during the development phase. The final product reflected only minor changes to the film-type storyboarding used. University funding was granted to extend this prototype to a cross-campus information kiosk in 1996. This

kiosk (one complete physical unit) is under development and is expected to be finalised early in 1997. A decision to implement the system more widely at the multi-campus CQU will be made in 1997, subject to financial and political considerations.

ACKNOWLEDGEMENTS

The authors acknowledge the support of Professor John D. Smith, Professor of Computing and Head of Department of Mathematics and Computing at CQU, who passed away in August 1995. He keenly supported this project in the early days and this is gratefully acknowledged.

REFERENCES

1. CRANSTON, M. (1995). The Development of a Prototype Presentation for an Information Kiosk, Undergraduate Project, Department of Mathematics and Computing, Central Queensland University.

2. SILK, D.N. and HODGSON, R.A. (1995). Of Kiosks and Campuses, T.H.E. Journal 23(1) : 69-72.

3. BOIKO, B. (1994). The back-end of a multimedia title, Multicomm '94 Proceedings, Vancouver, pp43-57.

4. BUNZEL, M.J. and MORRIS, S.K. (1992). Multimedia applications development: Using DVI technology, McGraw-Hill, Inc., USA.

5. PHILLIPS, R. (1993). Managing computer-based learning projects at Curtin University, Proceedings of the 10th Annual Conference of the Australian Society for Computers in Learning in Tertiary Education, Lismore, N.S.W., pp491-501.

6. DURLAK, J. (1994). Designing and producing collaborative CD-ROMs in university settings, Computing Application Learning and Training B Resource Material, Department of Mathematics and Computing, Faculty of Applied Science, Central Queensland University.

7. PACIFIC RIM NETWORK SYSTEMS (1995). Custom Multimedia Kiosks, URL: http://www.alaska.net/~pacrim/kiosks.html.

8. ENGLISH, E. (1995). Touching on touch screen technology, Australian Personal Computer, 16(10) : 82-83.

9. CHERNICH, R. (1994). Touch Screen Hardware, unpublished report, Department of Mathematics and Computing, Central Queensland University.

10. CHERRY, J.M. and TURNER, J.M. (1992). Designing for diversity: The user interface for a hypermedia information system on a university campus, Behaviour and Information Technology 11(1) : 1-12.

11. RUSSELL, A. and REDHEAD, K. (1991). Style Manual: A guide to screen design for computer based training, Queensland University of Technology.

12. MERRILL, M.D. (1988). Don't bother me with instructional design - I'm busy programming! In Computers in education (4th ed.), Guilford, Conn.: Dushkin, pp37-45.

13. PERRY, M. (1995). Multimedia Kiosk Sparks Interest, Multimedia Asia Pacific, 1 (2) : 15-16.

14. CHEUNG, W. (1993). Common mistakes in developing computer-based learning packages, Proceedings of the 10th Annual Conference of the Australian Society for Computers in Learning in Tertiary Education, Lismore, N.S.W., pp123-127.

15. ALESANDRINI, K. (1995). Be an Instant Windows Author with Toolbook, Data Based Advisor, January 1995, pp108-113.

16. WHALLEY, P. (1993). Topic accessed video, Proceedings of the 10th Annual Conference of the Australian Society for Computers in Learning in Tertiary Education, Lismore, N.S.W., pp686-690.

17. LYNCH, T.A. (1994). The Art of Good Scanning, unpublished report, Department of Mathematics and Computing, Central Queensland University.

18. ASYMETRIX CORPORATION (1991). The Concise Guide to Multimedia, Asymetrix Press, USA.

COPYRIGHT

M. Cranston, D.J. Clayton & P.J. Farrands (c) 1996. The authors assign to ASCILITE and educational and non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The authors also grant a non-exclusive licence to ASCILITE to publish this document in full on the World Wide Web and on CD-ROM and in printed form with the ASCILITE 96 conference papers, and for the documents to be published on mirrors on the World Wide Web. Any other usage is prohibited without the express permission of the authors.