Personalizing motion sickness models: estimation and statistical modeling of individual-specific parameters

Kotian, Varun; Pool, Daan M.; Happee, Riender · 2025 · Crossref

DOI: 10.3389/fnsys.2025.1531795

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of predicting motion sickness in automated vehicles and driving simulators, where sensory mismatches between visual and vestibular inputs cause significant discomfort. Because individuals vary widely in their susceptibility to motion sickness, group-averaged models are insufficient for designing personalized countermeasures. The authors aim to develop a personalized modeling framework that combines a group-averaged "conflict generation" model with an individualized "conflict accumulation" model (AM) to accurately predict individual sickness responses across diverse conditions. The researchers utilized four existing datasets involving passive motion scenarios, including vehicle experiments with varying vision conditions, matched vehicle and simulator trials, and non-driving-related tasks. The modeling framework integrates the Subjective Vertical Conflict (SVC) model to generate a scalar sensory conflict signal from six degrees-of-freedom motion and visual inputs. This signal serves as input to a nonlinear Accumulation Model, which predicts sickness levels using the MIsery Scale (MISC). The study explicitly optimized the number of parameters required for personalization, comparing a full five-parameter AM against reduced versions to identify a "minimum effective" implementation that ensures generalizability and prevents overfitting. The results demonstrate that a reduced two-parameter model (AM2), utilizing only gain ($K_1$) and time constant ($T_1$), effectively fits individual motion sickness responses across all tested conditions. This AM2 model achieved an average improvement factor of 1.7 in fitting accuracy compared to a group-averaged model (AM0). Furthermore, the authors developed a Gaussian mixture model to describe the statistical distribution of these individual parameters across the population. This statistical model successfully predicted motion sickness in an unseen validation dataset with an average Root Mean Square Error (RMSE) of 0.47. The significance of this work lies in its ability to personalize motion sickness prediction without requiring extensive individual calibration data. By reducing the necessary parameters to two, the framework enhances model robustness and generalizability across different motion and vision conditions. The derived statistical model allows for the prediction of individual sickness susceptibility based on population distributions, thereby reducing the need for large-scale population experiments. This approach accelerates the development of optimized motion control strategies for automated vehicles and improved motion cueing algorithms for driving simulators, ultimately enhancing user comfort and safety.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-09; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• passenger motion sickness comfort
• simulator sickness

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Theoretical Contribution: computational model