REMEMBER Lawnmower Man, the tacky Hollywood movie that tried to transport its audience into some psycho-filled virtual future? Or the hype a couple of years ago when Sega and Nintendo claimed to be on the verge of bringing virtual reality to the home? Not much has happened since. Go down to one of the large arcades, and you can don a headset and whiz around a simulated motorcycle racing track, but that’s about it. We are still a long way from the promised world of a VR set in every home.
Things might be different if quality VR was available at an affordable price. Any system costing less than six figures lacks the picture quality and speed needed to make it anywhere near realistic. But perhaps this slow progress is all to the good because there is another problem, too, and that won’t be fixed simply by waiting for the cost of computing power to fall. What is beginning to worry the VR industry is the effect on living, kicking human beings when they are immersed for any length of time in an artificial electronically generated environment.
That existing VR systems can have ill effects on their users is not in doubt. But how serious are they? Claimed problems range from temporary nausea to permanent damage to a VR user’s vision. Also in question is what, if anything, can be done to avoid these hazards.
Advertisement
Dangerous nausea
More is at stake than just the future of electronic games. There are great hopes for VR as a training medium for people such as firefighters and others who need to learn how to tackle dangerous tasks without putting themselves at risk. And any problems with VR are likely also to appear in related applications, such as the use of head-mounted displays by surgeons to view the site of an operation, or by nuclear technicians to get a robot’s eye view of the inside of a reactor. Few patients would be willing to undergo an operation using this technology if the person manipulating the probes in their stomach is going to have to rush off and be sick halfway through the procedure. Nor would a sudden bout of nausea help someone moving fuel rods in an emergency at a nuclear power plant.
Probably the best-documented problem is that linking VR with motion sickness. Since 1993, work carried out at the Army Personnel Research Establishment in Farnborough, Hampshire – now part of Britain’s Defence Research Agency – has confirmed this form of sickness as a common nuisance, especially for those new to VR. Clare Regan of the DRA observed 150 people who each spent 20 minutes wearing a head-mounted VR system through which they explored a virtual environment consisting of a series of rooms. She found that at first 61 per cent of the 150 subjects reported some side effects either during their time in the virtual environment or in the 10 minutes immediately afterwards; of these, 28 per cent reported symptoms of nausea. Eight people had to opt out of the study because their nausea or dizziness was so severe.
The most popular explanation for the link between motion sickness and VR is that it is caused by a conflict between the messages received from different parts of our sensory systems. According to this theory, VR confuses the vestibular-ocular reflex (VOR), the neural link between our eyes and our inner balancing system.
Many VR setups are controlled by a set of buttons on a hand-held joystick that the participant uses to move around the virtual environment. The display shows the computer-generated scene changing as it would if your body was in motion. Meanwhile your vestibular or balancing system is telling you that you are standing still. The brain might interpret this conflict as a dysfunction caused by poisoning – hence the urge to vomit. A variant of this theory suggests that it is time-lags between making a movement and seeing the result that cause a difficulty for VR users because here too there is a discrepancy between the movement of the head and the perceived result.
But Regan is confident that the problems with nausea can be overcome. Generally, she says, people adapt to VR, just as sailors get their “sea legs”. Regan has found that the incidence of nausea in a second VR session is about half that for people’s first session, and after four sessions the figure is less than a quarter. “I think this problem is surmountable to a greater or lesser extent,” she says. “The data show quite dramatic effects after quite a short time. About 5 per cent of people will not adapt, but they might be cured by some other method.” For these more sensitive souls, motion sickness pills might be one solution.
Evasive action
Some researchers suggest that ill effects can be kept at bay by allowing people to move in a way that is consistent with what is happening in the virtual enviroment – riding an exercise bicycle while cycling in the virtual world, for example. Other ideas include the use of lightweight headsets or systems capable of generating sharper images.
There has also been some research into how reducing the time lag between a change of scene and the movement that caused it affects the user. This has the benefit of strengthening the VR user’s feeling of presence in the artificial environment. A heightened presence is particularly important if VR is to be used in training situations, says Mel Slater, formerly at Queen Mary and Westfield College, who now leads the VR team at University College London. He says it is crucial that people being trained should respond as they would in real life. He and his team knew they had got their firefighting simulation right, he says, when the firefighters’ legs began shaking as they walked along a virtual precipice.
But as it turns out, improving the response rates to gain this kind of presence seems to increase rather than reduce the number of users reporting nausea. One explanation may be that when there is an obvious conflict between what the eyes are seeing and what is happening to the body, the brain’s VOR learns to compensate – which is why symptoms become less likely the more someone uses VR. But when the time lag is very small, this is only perceived at a subconscious level and the brain cannot compensate, although it remains aware that there is a conflict and so the user may still feel sick.
Nausea isn’t the only threat to the VR experience. Two years ago researchers from the University of Edinburgh’s department of psychology upset the VR world by suggesting that some setups could cause visual problems. Mark Mon-Williams, John Wann and Simon Rushton found that after spending as little as 10 minutes in a typical stereoscopic virtual environment, people experienced a range of problems such as blurred vision, eye strain, headaches and even double vision. All the symptoms were temporary and, except for the headaches, lasted minutes rather than hours, but the results, which were the first evidence of VR causing measurable changes to the visual system, shook many researchers and VR manufacturers.
To try to pin down the cause of the problem, the Edinburgh researchers then carried out similar tests on people using “biocular” systems instead of stereoscopic ones. The key difference is that with a biocular system the two eyes see identical images, so the scene appears flat to the person viewing it. Stereoscopic systems, by contrast, aim to trick the brain into seeing a 3D effect by projecting a slightly different image to each eye. With biocular systems, users reported fewer problems than with 3D ones, even when immersed in the virtual environment for half an hour.
These results cannot, however, be used to pin the blame unequivocally on the 3D effect because there were a number of important differences between the two systems. Unlike the 3D system, the biocular system allowed users to alter the focus for each eye idenpendently. Users were also able to alter the separation between the two eyepieces so that they did not have to squint.
Split focus
But the Edinburgh researchers still suspect that there will be problems when using flat images to trick the brain into seeing a scene in 3D. When someone is wearing a VR headset their eyes naturally focus on the flat screens inside. But this can become confusing when the images shown to each eye differ in a way that gives the impression of depth, says Wann, who is now at the University of Reading. The displayed images then appear to extend both in from of and behind the plane of the screen, so the eyes will try to alter their focus accordingly. The worry is that this may have harmful effects, especially in people who have poor binocular vision to begin with.
Even if the effects turn out to be no more than temporary, the Edinburgh team warns that they could still be dangerous. Someone who needed to drive a car or operate dangerous equipment immediately after being immersed in the 3D environment would be a danger to themselves or others if their ability to see in 3D was impaired.
If stereoscopic images are the problem, the obvious solution is not to use them. This would, however, rule VR out for applications such as surgical training, which is seen as a key future market for the technology. The consensus among researchers is that a lot more investigation is needed on how stereo VR affects VR users’ vision, and how long any harmful effects are likely to last.
There is also the question of what effect repeated immersion in any kind of VR system might have on vision – an important consideration if VR ever makes its way into home computer games. Tom Furness of the Human Interface Tecnology Laboratory (HITL) at the University of Washington caused a stir last June when he was quoted as saying that cheap, poor engineering products could leave users with long-term visual disturbance. He seemed to suggest that frequent use of VR could impose long-term stress on the VOR that would leave it permenently damaged. The implication is that the brain might form new neural pathways to compensate for disruption to the VOR, leading to “flashbacks” as the brain later switched unpredictably between the old and new pathways.
This highly speculative argument has rung alarm bells in the VR industry, especially among people who have experienced confusion or flashbacks themselves. Several researchers say that after spending some time in VR they have ocassionally become disorientated or experienced images inspired by the virtual environment. One researcher even admitted needing to lie down for a couple of hours after being immersed in a high-quality virtual environment.
Erik Viirre, a colleague of Furness at the HITL, who leads the health and safety research programme there, likens VR to getting a new pair of glasses. “When we change glasses to a new magnification or switch between glasses and contact lenses, the scene is not what the eye is expecting. My hypothesis is that in a way VR is like wearing new glasses – you have a new context and you have to get used to it.”
While not disputing the seriousness of the symptoms described by Furness, Viirre plays down the idea that VR can damage the brain, suggesting instead that it causes the brain to add another context or neural pathway for use in the virtual environment. He says the brain appears to use various context clues to find out when it should switch its visual response. The act of putting on a pair of glasses, for example, can lead it to change its expectations, so the same could presumably apply to the act of putting on a VR helmet. “Where there would be concern is if the VR system was changing constantly as some people might not tolerate that very well,” Viirre says. “As long as the system is working consistently it’s easier to get used to.”
VR appears to cause a number of distinctly different types of visual disturbance that are all labelled “flashback”. Some people report a type of visual hallucination that involves something from the virtual environment appearing in the real world. Others have experienced short term flashbacks similar to the after-image you get after looking at a bright light. And yet others, such as Furness, have described a sort of motion flashback, where they experienced movement sensations similar to those in the VR environment, some time after using the system.
Knowledge gap
Different as they are, these experiences may have a common cause: the brain picking up inappropriate visual cues and momentarily switching into an inappropriate visual context. Researchers are warning that we do not yet know enough about VR technology, or our visual systems, to be confident that VR is harmless.
Many researchers are keen for health and safety guidelines to be established, so that VR systems which conform to them can be certified accordingly. However, John Wilson of the University of Nottingham virtual reality applications research team, feels there is still some way to go before any useful guidelines can be drawn up. “No one is certain how and why VR affects us and what its effects are, so nobody is certain what tests to do.” Wilson and his team hope to find some answers through developing tests that manufacturers could use as a way of guaranteeing standards.
Worrying as the potential hazards of VR may be, many researchers feel that it is also important to keep them in perspective. Viirre, for one, acknowledges that research into the effects of VR and possible ways of testing the safety of equipment is important, but he also points out: “We let teenagers drive cars, and that’s far more dangerous.”