Full Paper

Lawrie M Schafe and Rodney J Clarke

lawrie_schafe@uow.edu.au, rodney_clarke@uow.edu.au

Department of Business Systems

University of Wollongong

One of the main problems in teaching introductory information systems subjects is conveying the context of use of information systems in organisations. While the best textbooks describe the functions of general types of information systems, (eg. transaction processing systems and management information systems), they generally over emphasise the hardware and system software aspects of these systems. These textbooks tend to pay little more than lip service to the issue of the context of use of information systems: some provide case studies, typically described in 100-500 words! Thus student comprehension of information technology and information systems use is relatively shallow, biased toward the technology, and disconnected from its context of use.

This is a systemic problem, particularly at undergraduate levels, given the generally widespread lack of student experience in organisations. This problem could be addressed by the provision of extended text-based case studies, but many of our students come from non-english speaking backgrounds and this would disadvantage them in comparison to local students.

Thus, in a CAUT project application in 1995 (Schafe and Clarke, 1995) we suggested that the problem might be usefully addressed by the use of a multimedia approach, despite the fact that at that time we found very little published empirical support for the widely held belief that a multimedia approach was beneficial to learning. This strong belief in the value of multimedia to student learning gains its support mainly from the constructivist theory of learning (Tergan, 1997) which suggests that multimedia, via its ability to provide multiple representations of differing conceptual views, formats and contexts of the material to be learnt, can and will actively engage the learner in constructing their own understanding of the material to be learnt. There seems to be more evidence in support of this belief now (eg, Najjar, 1996; Relan and Smith, 1996; Yaverbaum and Nadarajan, 1996) than there was when we wrote the CAUT project proposal. However, the accumulating evidence is neither unequivocal in its support nor without need of qualification in its applicability (Najjar, 1996; Tergan, 1997). We look forward to the prospect of evaluating our product next year and adding some empirical evidence of our own to this debate.

In order to obtain an industrial/business site of perceived relevance and importance in which to set our multimedia CAUT pilot project, we approached the BHP IT group and following lengthy discussions and negotiations, determined that we and they would be happy for us to investigate some aspects of the BHP Supply System. This gave rise to circumstances in which a set of possibly appropriate educational uses for Apple's QuickTime VR technology suggested themselves within the context of bringing this industrial case study of the use of 'Information Systems in Context' at the BHP Port Kembla steelworks in Wollongong to our students. In support of this possibility, we note that Ainge (1997) argues:

The physical and logistic difficulty of organising computer centre site visits for about 650 undergraduate and 150 postgraduate information systems students in introductory classes suggests that we might try to bring the computer centre (and other relevant organisational settings having limited capacity to handle such large numbers of visitors safely) to the students via the use of VR and other media. It must be noted however, that our purpose was not solely to bring the computer centre to students, but rather to bring an integrated and linked set of material settings to the students in order to illustrate to students how organisations use information technology and information systems to assist them carry out their day to day business.

Thus, in our CAUT pilot project (Schafe and Clarke, in press) we used a number of Apple QuickTime VR movies to provide students with a partially immersive virtual reality environment in which they can explore and interact with via hot-spots in the VR movie. For example, the 30 node VR movie of the computer room allows the student to 'walk through' a large computer centre, opening some of the computer cabinets to see what they contain and being given an explanation, via graphics and/or text, of what they do and how they do it.

From a design and data collection point-of-view, the processes involved in producing a QuickTime VR movie of a scene involving multiple nodes are as follows (Apple Computer, 1995a, 1995b).

If you are not familiar with the site and the scene(s), as was the case with our project, arrange a visit and tour of the site. Fortunately, our site guide was in charge of the computer room and was able to answer most of our questions and to provide us with a floor plan for the site we were interested in. In particular we noted that the false or floating floor in the computer room was composed of square removable floor 'tiles.' Having obtained a clear visual and conceptual image of the site and the scenes, we then addressed the following planning issues.

• The type of experience to be made available to the user - we decided that we wanted the user to be able to walk through the computer room, go and look about in the tape room (adjacent to the computer room) and the printer room (across a corridor from the computer room), in a way not dissimilar to what used to happen in the past when we took students on a tour of a large computer centre. However, the computer room was quite large, containing many different systems which were not related to the supply system. Thus we decided to plot a path through the computer room that incorporated most of the computing equipment being used by the supply system. Including a node outside the doors to the computing room (for the purpose of continuity and connection to other single node VR movies of the tape and printer rooms), we had a 30 node VR movie of the computer room.
• The importance of the scene to the product - quite important since it would provide the context in which to show and explain many of the fundamental components of a large scale computing system in use in a 'real-world' environment.
• The size of the scene in terms of disk space - at the time we were planning the VR, we were not in a position to make any reasonable estimates of this since details of other data we needed to describe and explain the supply system were still under investigation.
• Aspects of the scene that might prove difficult to shoot - since the computer room was lit by banks of closely spaced baffled fluorescent lamps only, there was likely to be a slight greenish/blue colour imbalance in our negatives. A Minolta 1A filter was attached to the lens to limit this effect. Reflections and direct light 'flares' were not a problem. The light level would be fairly low and we would need to use a fairly sensitive light meter to accurately measure the light level. Relatively long exposures (0.25 to 1 sec) would be needed in order to shoot at f11 or f16 to increase depth of field so that most of the objects in the shot would be in focus. Another difficulty was that some of the node points we planned to shoot were in relatively tight 'corridors' between banks of computer equipment cabinets that had few or no distinguishing features on them to help the QTVR 'stitcher' program to accurately merge adjacent shots.
• The amount of time and money to be budgeted for the scene - since it was primarily our time and the resources (Kodak film, processing and transfer of images to Kodak Photo-CDs) were relatively cheap, we did not worry about this.
• Post-processing time to be used on the scene - we hoped that we would do a good enough job on the shooting of the scene that we would not need to spend much time on post processing of the photographic resources.

The location of the nodes was determined by where the key pieces of computing equipment relevant to our supply system were, and then we determined the 'in-between nodes' so that we would have a fairly smooth transition from one node to the next. Having the computer room floor divided into 20 sq. in. tiles made it much easier to locate the nodes on the floor. We decided that the in-between nodes would be 60 in. (3 tiles) apart where possible, otherwise 40 in. (2 tiles) apart. Apple (Apple Computer, 1995a) recommends they be between 3 and 7 ft. apart. The further they are apart the quicker and 'jumpier' are transitions between nodes. Since all the nodes were to be photographed at the same height and at the same horizontal pan angle, there was no need to distinguish between key nodes, entry nodes, special nodes and in-between nodes as suggested by Apple. We just marked the nodes on the floor plan of the computer room and numbered them sequentially.

According to Apple Computer (1995a) you need the following type of equipment to photograph a QuickTime VR node panorama:

• A 35mm SLR camera with manual focus lens and motor driven film advance. Recommended is a Nikon N90; we used a Minolta SRT101 - while not as good as the Nikon, still a good quality camera capable of producing very good results. It does not have a motorised film advance mechanism.
• A high-speed colour print film. We used Kodak Ektapress Multispeed Professional color negative film, rated at 400 ASA (it can be rated anywhere between 100 and 1000 ASA). This gave us the flexibility needed with exposure timing (the camera permitted automatically timed exposures of 1 sec, 0.5 sec, 0.25 sec, etc) under normal indoor commercial lighting conditions to permit an f-stop in the range f8 to f16 so that we achieved good depth of field. The 400 ASA rating resulted in exposures with no obvious graininess. It is also important to note that Kodak claim that reciprocity effects do not need to be factored into exposure calculations for this film until exposure times equal or exceed 10 sec.
• A wide angle distortion-free lens. They recommend a Nikkor 15mm lens because it provides a large vertical field of view and requires only 12 shots per node to capture a 360 degree panorama. We used a much cheaper Minolta Rokkor 28mm (with 1:3.5 aperture) lens. The longer focal length meant that our vertical field of view was not as large as we would have liked, but it was sufficient to capture all the detail we needed.
• A sturdy tripod. Recommended is a Bogen 3426. We used a much less sturdy and considerably cheaper aluminium Slik 35S tripod. This lack of sturdiness meant that the time taken to shoot each node panorama was longer than it might have been since we had to be very careful not to move the tripod/camera as we wound the film-on, or rotated the camera on the pan head, or removed and refitted the camera in order to change the film. What we had estimated to be about an 8-10 hr shoot in fact took us over 15 hrs!
• Two sturdy pan/tilt heads mounted one on top of the other, both with axial bubble levels for easier camera leveling. We used only one pan/tilt head- the one built into the tripod to which we attached a KAIDAN Quickpan Model QPA-1 pre-calibrated pan head (designed specifically for the Apple QuickTake camera) with one bubble level built-in to enable the camera to be levelled. One of the KAIDAN's pre-calibrated settings was for 20 degrees which was exactly what we needed for our 28mm lens, however, we found we could not mount the camera as required so we had to do a little metal working on it to permit the camera to be secured properly to the attached L bracket!
• A large horizontal mounting plate such as the Slik Twin-Camera Plate 400, mounted on the top pan/tilt head, to allow horizontal movement of the L bracket, and an L bracket for mounting the camera on the top of all this in portrait orientation. The KAIDAN QuickPan pan head has an attached L bracket for mounting the camera in portrait orientation, but it has no capacity for horizontal movement. Because of this and the geometry of the KAIDAN head, we were unable to align the camera film plane with the axis of rotation (it was a few inches out) as required!
• A sliding quick-release camera mount for mounting the camera to the L bracket so that mounting and unmounting the camera (to change films) is quick and easy. Recommended is a Bogen 3273 Pro QR Adapter. We just used the normal camera base-plate screw hole and the bolt attached to the KAIDAN pan head L bracket. A small adjustable spanner was used to ensure that the mounting of the camera to the L plate was very firm so that the camera could not move in the vertical plane.
• A glare shield was unnecessary for our site since there were no sources of glare, but we did attach the lens glare hood that came with the lens.
• A remote shutter release. This was essential given that we were having to use relatively long exposure times and a relatively lightweight tripod.
• They also optionally recommend the use of a dolly for the tripod rig so that moving the camera around in a multi-node scene is easy, as is setting the rig at each node ready for shooting the next panorama!
• A light meter - we used a Gossen Lunasix 3 Light Meter since the level of the light was outside the range of the camera's built-in light meter to measure with accuracy.
• Finally, a plumb line is recommended to locate the camera accurately above a node marked on the floor. We attached ours (a piece of lead sheet crimped around the end of some twine) to the central height adjustment post of the tripod and used the floor tile intersections as the node points.

Apple (Apple Computer, 1995a, 1995b) then recommend a very detailed procedure for setting-up the camera rig (as specified above) at a node. It emphasises the need to get the camera exactly level, precisely above the axis of rotation, as well as accurately above the node marker. While the caution and precision they advise is understandable and would undoubtedly yield excellent results, we have found from experience that it is not necessary to be so precise and yet still achieve quite reasonable results. Also clearly evident from the above details is that it is not necessary to use the highest quality equipment to achieve acceptable results. The equipment we used was that owned by a keen amateur photographer. Hopefully, camera equipment of this quality or better should be available for use in most Universities.

In doing the shoot of our scenes and getting the resulting films developed and transferred to Kodak Photo-CD, we found the following issues needed attention:

• Unless very, very careful, it is easy to lose track of whether or not you have taken a particular shot at each of the required pan angles at a node. This difficulty is most likely to arise when you reach the end of a film and need to unmount the camera, remove the exposed film, load a new film and put the camera back on the rig. If you have any doubt, we suggest you re-shot the whole node again.
• Watch people who come in and out of the scene very carefully so that if they move, remove or add anything to it you can restore the scene to its prior state if necessary. Clearly, it's preferable not to allow people to enter/leave the vicinity of the scene, but this may not be practicable in the real world.
• Make sure you keep a careful record of which film numbers and negatives contain the shots for each node. If your film processor and Kodak Photo-CD producer was like ours, you will need that information to identify which photos are which, particularly if some nodes have similar panoramas to others.
• When sending off your films for processing, tell your film processor NOT to cut the resulting negatives up into the usual strips of four or six negatives. It is much easier to have the images transferred in the proper order to Kodak Photo-CD if the film remains intact as a roll.
• Tell your Kodak Photo-CD producer to look at one or two rolls of film and ascertain whether the exposure is reasonable. If they are reasonable, get them to transfer ALL negative images (no matter what the quality of each image, nor whether they think one is a duplicate of another, etc.) from each film to the CD in film number and negative number sequence! Tell them to turn off 'automatic exposure compensation' when scanning so that artificial variations in exposure are not introduced.

(c) Lawrie M Schafe and Rodney J Clarke

The author(s) 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 author(s) 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 97 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.