Comparative Study of Younger and Older Drivers’ Emergency Response and Gaze Behavior in Sweden

Conference poster, 2026

Age-related visual behavior changes may impair drivers’ hazard responses. While some research has explored drivers’ visual behavior, detailed studies on the drivers’ visual characteristics across ages under varying conditions remain insufficient. This study explores how age affects visual behavior in intersection scenarios, and whether these differences influence reactions to hazards.
Using a CARLA-based driving simulator, 36 younger and 24 older drivers completed two urban scenarios. The first scenario consisted of a 1081 m roadway with five intersections and two 3.25 m lanes. At the final intersection, a cyclist emerged from left to right across the vehicle’s path, resulting in a time-to-collision (TTC) of 2.0 s. The second scenario comprised a 3136 m roadway with 10 intersections and two 4.05 m lanes. A cyclist appeared from the left side at the 9th intersection, with a TTC of 2.5 s.
Drivers in both age groups focused on the right side in both scenarios, with no significant effect of age. Older drivers tended to gaze narrowly and closer to the ego vehicle, which was more pronounced on the narrow road. When approaching intersections, drivers scanned more often, however, their scan frequency decreased within 20 m of intersections as they navigated through it. On the narrower road, drivers scanned the right side more often within this 20 m zone. Younger drivers had higher scan speed than older drivers. Moreover, older drivers showed longer scan time on the right side, showing their awareness of the near side.
The findings demonstrate that understanding drivers’ gaze behavior, particularly increasing traffic fatalities within aging populations, can support the development of more effective driver monitoring systems and advanced driver assistance technologies. This study also has potential to inform driver behavior modeling by incorporating key influences such as age, cultural context, driving side, and variability in traffic environments.