The spotlight on computer games in connection with virtual reality (VR) means that most people associate VR with entertainment. They brush VR off as trivial and think it is primarily aimed at children of all ages. It is true that the entertainment industry has been an important user of the technology and a primus motor in the development of hardware for 3D visualisation. On the other hand, work is being done on VR for industrial purposes. In the long term it is likely that such applications will have a much bigger impact than pure entertainment – though probably not in financial terms.
If VR is to be credible, two ingredients are crucial: the user must feel himself present in the data set (”immersive environment”), and interaction with the data must be natural. This is achieved best by stereoscopic representation in which the visual impression is governed by the viewer’s own movements. Such a concept is called SSVE (Surround Screen Virtual Environment).
One of the Norwegian communities aiming to become of the market leaders in VR is to be found at Christian Michelsen Research (CMR) in Bergen. CMR’s
so far biggest VR project is development of software for a CAVE™ for Norsk Hydro. This is a SSVE in a room measuring 3 x 3 x 3 metres, in which pictures are projected onto three walls and the floor. Reservoir data are shown in the room in such a way that users with stereo glasses are given the feeling of being inside an oil or gas reservoir.
The programme is organised in such a way that the primary target data, the reservoir description, is replaced with other kinds of data suitable for VR representation. In this way the system has been used to visual medical data and gas explosions.
Several medical instruments generate 3D data sets. The traditional procedure for analysing this kind of data is to study 2D cross-sections, but it often requires long experience to be able to see the whole picture and draw conclusions from these. Use of VR provides a better holistic understanding of the anatomy represented. Medicine is one of the disciplines in which VR finds many promising applications, for example the training of medical personnel, surgical simulations and telemedicine.
VR visualisation of gas explosions can give the users the feeling that they are waling about in the structure where the explosion happens, and they can then study the flame front from various vantage-points in the structure or from outside, at the speed they choose. This provides a greater understanding of the process than traditional representations on computer screens.
All representation of 3D graphics requires computing-power, and even though price has fallen relative to performance, there will always be applications that demand more than current technology can offer. In such applications it is acceptable for the picture to show the familiar hour-glass while it updates; at other times such a wait will ruin the experience, as for example in games simulating flying or car racing. In such cases it is usual for the programme to test the platform it is running on and adjust the level of detail so as to maintain the minimum requirements for image refreshment. In representation of seismic data in a SSVE, however, neither solution is acceptable. The user must be able to move, rotate and scale the volumes with flexible lighting and shadows in real time, and stereoscopically. If the user’s visual experience is out of synch with his actions, or if delays are flagged with any kind of hour-glass, the impression of operating in a virtual reality will be ruined.
The primary goal in the short term is to slash the costs of VR technology and thereby make it available to a wider market. For many applications, the data sets will be much smaller than typical reservoir data, so that the hardware requirements can be reduced on this basis alone.
The graphics accelerator card for the home PC has seen an explosive development in capacity and performance over the last few years, but without a price rise. Up to now the properties of the card have been largely managed by and optimised for the games market, with the consequences that properties essential to industrial VR are entirely lacking. This is now changing, in the current generation of graphics cards. The desired functionality will be incorporated into such cards relatively soon.
Want new tools
This year CMR has initiated a Strategic Institute Programme supported by the Research Council of Norway with the objective of analysing and developing scaled-down VR systems. It will survey the kinds of problems suitable for running on computers of all shapes and sizes. In addition, the VR communities want a paradigm shift: current methods of working based on workstations, screens, keyboards and pointing tools are poorly-suited, perhaps even worthless, as a VR user interface. Alternatives must be introduced, tested and accepted. But what this actually means is an only partly-explored territory. There is nothing particularly original about predicting that in a few years programs that are as demanding as the biggest of today’s VR applications will be run on small or medium computers. This will allow this technology to be used in disciplines we can hardly imagine today.