Frontiers in psychology · 2026
Impact of SA-related visual characteristics on driver takeover reaction time in L3 automated vehicles: an eye-movement-based evaluation method.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Wan Jian, Zeng Yue, Liu Huayu, Ma Yulin
Аннотация
INTRODUCTION: In Level 3 (L3) conditionally automated driving, drivers must rapidly regain situational awareness (SA) following a takeover request. This study examined pre-takeover eye-movement characteristics and their associations with SA-related states and takeover reaction time under different takeover scenarios and non-driving-related tasks (NDRTs).
METHODS: A driving simulator experiment was conducted under four conditions. Of 168 trials from 42 participants, 108 valid trials from 27 participants were analyzed. Pupil area, saccade duration, and fixation duration during the 5 s before takeover were integrated into an SA-related score using the CRITIC method. Machine-learning models were compared to predict takeover reaction time.
RESULTS: NDRT type significantly affected the three eye-movement indicators, whereas takeover scenario did not significantly affect these indicators or reaction time. The SA-related score was significantly associated with takeover reaction time. LightGBM achieved the best predictive performance, with an accuracy of 0.818 and an F1-score of 0.837, and the SA-related score showed the highest predictive contribution (29.40%).
DISCUSSION: Pre-takeover eye-movement characteristics may provide an objective physiological proxy for SA-related visual-processing states and support the prediction of initial takeover response speed. These findings may inform adaptive driver monitoring and takeover assistance, although the observed relationships are predictive rather than causal.
METHODS: A driving simulator experiment was conducted under four conditions. Of 168 trials from 42 participants, 108 valid trials from 27 participants were analyzed. Pupil area, saccade duration, and fixation duration during the 5 s before takeover were integrated into an SA-related score using the CRITIC method. Machine-learning models were compared to predict takeover reaction time.
RESULTS: NDRT type significantly affected the three eye-movement indicators, whereas takeover scenario did not significantly affect these indicators or reaction time. The SA-related score was significantly associated with takeover reaction time. LightGBM achieved the best predictive performance, with an accuracy of 0.818 and an F1-score of 0.837, and the SA-related score showed the highest predictive contribution (29.40%).
DISCUSSION: Pre-takeover eye-movement characteristics may provide an objective physiological proxy for SA-related visual-processing states and support the prediction of initial takeover response speed. These findings may inform adaptive driver monitoring and takeover assistance, although the observed relationships are predictive rather than causal.