Chen, Junyi (Tongji University), Wang, Peiyi (Tongji University), Wu, Xinzheng (Tongji University), Xiao, Wenbo (Tongji University), Li, Duo (SAIC Motor R&D Innovation Headquarters), MENG, Haolan (TONGJI UNIVERSITY)

A Study of Cognitive Discomfort in Urban Intersection Scenarios Based on Vehicle Occupants’ Eye Movement Characteristics

Scheduled for presentation during the Poster Session "Human Drivers in Intelligent Transportation Systems" (FrAT15), Friday, September 27, 2024, 10:30−12:30, Foyer

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 3, 2024

Abstract

Ensuring riding comfort is crucial as it affects the acceptance of Automated Vehicles (AVs). Cognitive discomfort of vehicle occupants, based on the cognition of the driving situation, is generally induced by visual stimuli. Therefore, it is necessary to study occupants’ eye movement behavior, which helps to reveal the mechanism of cognitive discomfort. This research conducts a Naturalistic Driving Study (NDS) to analyze the relationship between eye movement and occupants’ cognitive discomfort. First, an experiment on urban roads is conducted to collect data including participants’ eye movement, real-time subjective discomfort value, and driving situation. Intersections are chosen as the study scenario and 44 samples are finally selected. Then, descriptive statistics and comparative analysis are employed to reveal the occupants’ eye movement characteristics. Subsequently, with two eye movement metrics (mean fixation duration and saccade count per second) as input, a multiple linear regression model is established based on naturalistic driving data to predict cognitive discomfort levels. Furthermore, eye movement characteristics during driving are also influenced by various factors, such as driving experience level, ego vehicle speed level, and road type. In order to explore the influence of these factors, ANalysis Of VAriance (ANOVA) is employed. This study obtained a multiple linear regression model with a coefficient of determination (R2) of 0.788, indicating a strong predictive capability for the model. The above results demonstrate the potential of predicting discomfort through occupants’ eye movement data. And the results on the influencing factors of eye movement characteristics can further contribute to improving riding comfort in automated vehicles.