Urgency-gating at intersections: impact of short-term cognitive priming on left- turn decision-making

Unprotected left turns pose a significant accident risk, especially in complex urban environments. While traditional road safety measures focused mainly on improving physical infrastructure, this study investigated the effect of short-term cognitive priming on driver's left-turn decision making and the time to collision. Using a customized web-based 2D simulation that we have developed (leftturnsafe.org; source code, https://github.com/eugenzio/Traffic-Priming), 100 adult drivers were randomly assigned to three perceptual decisions: Urgency, Safety, and Control, and participants performed 21 intersection entry tasks with various the time to collision applied. Studies have shown that cognitive status influences immediate decision-making within a simulated driving environment. Urgency-gating phenomenon in which the urgency group terminates visual information search early due to subjective time pressure, significantly reducing reaction time (median value of 1160 ms), and risk error surged to 30.0%. On the other hand, the safety group showed a very defensive attitude, with a delay in response time (1,650 ms) and a surge in conservative errors due to excessive waiting to 26.5%. The control group maintained balanced decision making. While traditional safety approaches emphasize infrastructure-based interventions such as signal timing and road design, less attention has been given to how drivers’ internal perceptual and cognitive states influence gap acceptance decisions. Our findings indicate that under the imposed simulation conditions, the time to collision evaluation relies on both the physical calculation of distance and speed, and the psychological weight of the time constraints. In conclusion, a human-centered approach beyond infrastructure improvement is required to prevent fatal intersection collisions, and psychological intervention considering the driver's internal cognitive status should be actively utilized in advanced driving assistance system design.