Mitigating Motion Sickness in Automated Vehicles With Frequency-Shaping Approach to Motion Planning

Abstract

With automated driving, human drivers have been partially, and in the future will be fully, relieved from driving tasks, and it becomes more important to mitigate car-sickness in self-driving vehicles. Previous research on reducing motion sickness mostly focus on improving the cockpit environment through human factors engineering, while few efforts have been made through vehicle motion optimization, especially from the view of vehicle longitudinal and lateral accelerations. To this end, we propose a new conception to settle car-sickness issues in motion planning rather than in motion control, i.e., to first plan a desired set of vehicle motion less likely to cause car-sickness and then to track the desired motion. We develop a novel motion planning algorithm via frequency shaping approach, which incorporates the essential mechanism of car-sickness. Simulation and experiment results indicate that the proposed approach can reduce motion sickness dose value (MSDV) by 21% and 37%, respectively, comparing with the polynomial-based planning algorithm that optimizes acceleration and jerk magnitudes.