Motion sickness incidence estimation incorporating back-measured torso biodynamic response: Modeling and real-vehicle evaluation

Accurate estimation of motion sickness incidence (MSI) is important for vehicle comfort optimization and motion-sickness mitigation. Conventional vibration-based functional models generally use vehicle acceleration directly and therefore do not account for mechanical transmission between the vehicle and the occupant. This study develops axis-decoupled four-degree-of-freedom (four-DOF) models of the effective vehicle-floor-to-back vibration transmission of seated occupants. The model-reconstructed back/upper-torso acceleration is used as a configuration-specific exposure descriptor in a functional MSI estimation model that accounts for frequency weighting and temporal accumulation. Real-vehicle experiments showed that the coefficients of determination R 2 between the model-predicted and experimentally measured magnitudes of effective floor-to-back acceleration transmissibility were 0.9648, 0.9665, and 0.9524 in the z-, y-, and x-directions, respectively. These results indicate that the four-DOF models closely reproduced the experimentally observed magnitude characteristics of vibration transmission, supporting their subsequent use in MSI estimation. Under straight and slalom driving, the back-response-based estimation model reduced the group-level root mean square error (RMSE) by 22.10% and 17.54%, respectively, compared with the direct model. Additional analyses examined estimation performance across sex and body-mass groups, naturalistic smartphone viewing, and input perturbations.