Journal of safety research · 2026
Environmental context is associated with differences in driver behavior around partial automation alerts.
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.
Mueller Alexandra S, Gershon Pnina, Haus Samantha H, Cicchino Jessica B, Mehler Bruce, Reimer Bryan
Аннотация
INTRODUCTION: The study's objective was to understand how driver behavior around Tesla Autopilot alerts varies with environmental context.
METHODS: Using on-road data from 13 drivers who drove a 2020 Model 3 as their personal vehicle for 1 month, epochs were captured around alerts that were thought to be associated with environmental features rather than driver behavior. The presence of roadway features in the 10 s leading up to the beginning of these alerts were used to cluster the alerts by the road scenario. Hand, eye glance, and secondary task behavior surrounding alerts were compared across clusters.
RESULTS: Alerts were evenly distributed among three clusters, which we described as leftmost lane, rightmost lane, and nonmotorway road scenarios. Drivers spent a larger percentage of time looking away from the road and doing driving-unrelated activities during alert events associated with the leftmost lane compared with the other scenarios. Furthermore, these drivers exhibited a clear preference for having one hand on the wheel across the three scenarios, which is consistent with Autopilot's steering-torque monitoring strategy.
CONCLUSIONS: These findings indicate that the driving context ought to be considered when it comes to understanding how drivers behave around partial automation safeguards. Differences in behavior between the clusters indicate a possible influence of driving demand. Mechanisms for factoring in environmental context into adaptive driver support to keep drivers engaged in driving are discussed in the paper.
METHODS: Using on-road data from 13 drivers who drove a 2020 Model 3 as their personal vehicle for 1 month, epochs were captured around alerts that were thought to be associated with environmental features rather than driver behavior. The presence of roadway features in the 10 s leading up to the beginning of these alerts were used to cluster the alerts by the road scenario. Hand, eye glance, and secondary task behavior surrounding alerts were compared across clusters.
RESULTS: Alerts were evenly distributed among three clusters, which we described as leftmost lane, rightmost lane, and nonmotorway road scenarios. Drivers spent a larger percentage of time looking away from the road and doing driving-unrelated activities during alert events associated with the leftmost lane compared with the other scenarios. Furthermore, these drivers exhibited a clear preference for having one hand on the wheel across the three scenarios, which is consistent with Autopilot's steering-torque monitoring strategy.
CONCLUSIONS: These findings indicate that the driving context ought to be considered when it comes to understanding how drivers behave around partial automation safeguards. Differences in behavior between the clusters indicate a possible influence of driving demand. Mechanisms for factoring in environmental context into adaptive driver support to keep drivers engaged in driving are discussed in the paper.