Sickness in Motion

Self-driving cars are no longer science fiction: fully automated taxis have been operational since 2020. Automated driving may offer societal, environmental, and economic benefits. Despite such advantages, a negative consequence is an expected increase in motion sickness. Therefore, the overall aim of my dissertation is to contribute to research on the mitigation of motion sickness, particularly in automated driving. I started by investigating how motion sickness can best be measured when using a self-report rating scale. To measure motion sickness reliably, a scale should capture its progression unambiguously. The results of Chapter 2 indicate that a scale focusing on the symptomatology of motion sickness does so better than a scale focusing on general feelings of unpleasantness. This motivated my decision to use the Motion Illness Symptoms Classification (MISC) in my other studies. In Chapter 3, I explored if cognitive cues could influence the perception of self-motion. The results indicated: profoundly. Cognitive cues that manipulated a priori motion expectations elicited a percept of oscillatory self-motion in the absence of corresponding sensory stimulation. This finding supports the assumption that our brain uses a predictive mechanism in self-motion perception, such as internal models. Passengers presumably suffer more from motion sickness than drivers because drivers can better anticipate the car's accelerations. Anticipatory cues that alert passengers of accelerations via vision or sound have been demonstrated to mitigate motion sickness. In automated vehicles, providing anticipatory cues via the tactile modality may be more desirable. In Chapter 4, I investigated whether anticipatory vibrotactile cues that announced the onset of a forward displacement mitigated motion sickness, and if the timing of the cue was of influence. To determine their effectiveness, I developed a new pre registered measure: R. With R, it becomes possible to quantify the reduction in motion sickness symptomatology between a session with anticipatory cues and a control session in a single value. Using this measure, in Chapter 5 I compared the effectiveness of anticipatory auditory and vibrotactile cues for a more unpredictable motion stimulus. In both studies, the anticipatory cues generated some reduction in motion sickness, but large variability between participants resulted in a lack of statistical power. To increase power, I performed an internal meta analysis from which I concluded that anticipatory cues are overall effective in mitigating motion sickness. In Chapter 6, I performed a test track study in which I compared the effectiveness of anticipatory auditory and vibrotactile cues during a real car ride. The same analysis using the measure R indicated that the vibrotactile cue mitigated motion sickness. The auditory cue was significantly less effective and generated no significant mitigation overall. Remarkably, the mitigating effect of the vibrotactile cue was larger than the overall effect I found in the meta analysis on laboratory studies. Overall, my findings indicate that anticipatory vibrotactile cues can effectively mitigate motion sickness in automated driving. In conclusion, I hope that my work contributes to a better understanding of the measurement and mitigation of motion sickness, so that we can drive comfortably in the future.