Motion Sickness Prediction in Self-Driving Cars Using the 6DOF-SVC Model

Drivers who assign the driving task to a self-driving car switch to a passive role to work or enjoy leisure time like a traditional passenger. Consequently, the risk of developing motion sickness (MS) symptoms increases significantly. Adapting one's own driving behavior, e.g. by choosing an alternative route or decreasing the velocity, offers future intelligent vehicles a way to independently prevent MS. Accurate predictions help to improve journey's planning of the vehicle and make correct decisions so as to minimize disruption to the traffic flow. In the present study our contribution is as follow: We conduct two studies by focusing on real-world driving under self-driving conditions and induced MS symptoms in passengers. We simulated driving parameters of the conducted studies to extract simulated driving dynamics and contrasted them with recorded driving dynamics. A well-known model of MS, namely the six-degrees-of-freedom subjective vertical conflict model (6DOF-SVC model) was utilized to predict motion sickness incidence (MSI) for both studies. In order to do so, we implemented a customized Human-Vehicle-Model to map the car's dynamics to the head, which is crucial to apply the 6DOF-SVC model. We evaluated different Human-Vehicle-Model conditions and optimized the parameters of the 6DOF-SVC model to increase prediction accuracy in the case of our experiments. Note that our modeling approach enabled capture effects of missing visual anticipation and cognitive distraction that we present in our experiment. It is concluded that the 6DOF-SVC model is applicable in realistic driving scenarios as the ones used in our study.