by Benjamin P. Hughes

Motion sickness is a phenomenon that has been investigated for as long as humans have traveled by sea. In the 2000 years since Hippocrates wrote about the experience of seasickness, physicians and researchers have developed theories of why our bodies often react so negatively to certain types of motion. The most widely-accepted explanation is referred to as sensory conflict theory (Oman, 1990). The idea is that when your eyes are telling you one thing (like “I’m staying still”) but your inner ear is telling you something else (like “I’m moving forward”), your brain triggers symptoms like nausea, stomach awareness, and dizziness in response. Some evolutionary psychologists have suggested that this response is a sort of false-positive attempt to purge any potentially poisonous or toxic substances one might have ingested (Treisman, 1977).

Sensory Conflict Theory

The sensory conflict theory can also explain how motion sickness can be triggered by visual stimuli alone, in the absence of any actual physical motion. Although this sort of situation is fairly rare in the natural world, it often presents itself for users of virtual reality (VR) applications that simulate self-motion. In VR, it’s much more common for the eyes to tell you one thing (“I’m moving”) and the inner ear to tell you another (“I’m sitting still”). When nausea occurs in this circumstance, it is sometimes referred to as “cybersickness” and it represents a significant obstacle to the broader adoption of VR technology (Davis et al., 2014).

My colleagues and I were interested in how various visual and behavioral factors might affect the intensity of cybersickness. To that end, we developed two VR experiments involving 88 college students from UC Santa Cruz. The results were published in last month’s issue of Scientific Reports. In each experiment, we presented participants with several laps around part of a virtual race track (a rendering of Nürburgring Nordschleife) that were pre-recorded using a popular driving simulator called iRacing. The laps were presented using an HTC Vive head-mounted display (see figure below). Half of the laps were presented at full speed (an average of 100-120mph) and half at half-speed (50-60mph). Participants viewed a total of 12-16 laps that lasted 60 seconds each. After each lap, they rated (1) their sense of vection (how much they felt like they were moving), (2) how sick they felt, and (3) how present they felt in the virtual environment, on 20-point scales. We also tracked participants’ head movements during the virtual laps along three axes: pitch, yaw, and roll.

We wanted to find out whether the virtual driving speed (fast vs. slow) and the direction of the virtual movement (forward, backward, or sideways) affected the intensity of participants’ cybersickness. Furthermore, we wanted to know whether there were particular patterns of head movement that were associated with a reduction in cybersickness.

The Effects of Speed and Direction on Cybersickness

Since cybersickness arises from a conflict between vision and vestibulation, we examined whether degree of sensory conflict would prompt a higher intensity of sickness. At higher velocities, the visual “I’m moving” signal should be stronger, while the vestibular “I’m sitting still” signal should remain the same. As we expected, fast laps consistently induced stronger ratings of cybersickness than slow laps. Fast laps also induced a greater sense of self-motion and a stronger sense of presence in the virtual environment, despite neither of these factors being correlated with cybersickness. (See figure below.)

In Experiment 2, we tested whether the direction of the car’s motion relative to the participant affected cybersickness. We included three conditions: “forward,” in which the car is driving forward; “side-facing,” in which the participant is looking out the side of the car; and “reversed,” in which the forward recording is played in reverse. As expected, the forward condition elicited higher ratings of both self-motion and cybersickness compared to the side-facing and reversed conditions. (See figure below.) Given that it’s pretty rare to move sideways or backwards at high speeds in real life, it makes sense that both of these variables were higher in the condition that aligns more closely with real-world expectations.

The Effect of Head Movement on Cybersickness

We mapped out participants’ head movements during the forward-facing laps along the pitch, yaw, and roll axes. While no two participants moved their heads in the exact same way, a surprisingly consistent pattern emerged in the average head movements along the yaw axis (i.e. left and right head movements; see figure below). Participants tended to turn their heads in anticipation of the upcoming turns in the virtual road, initiating head turns about three-quarters of a second prior to the turn in the road itself, which is consistent with how people tend to move their heads in real-life vehicles. We also found that participants whose head movements were closer to the group’s average pattern of movements reported a greater sense of self-motion and a stronger sense of presence in the virtual environment. During the slow laps, this type of “normative head motion” was also positively correlated with reports of cybersickness.

Future Directions

While many aspects of cybersickness are still not well understood, this study provides several useful avenues of exploration to better understand the phenomenon and its underlying mechanisms. In particular, the consistency of participants’ head movements during a passive virtual driving simulation underscores the potential of using head movements as a behavioral measure for future studies on cybersickness and motion perception in general.