Predicting Driver Emergency Response Onset using Time-to-Impact 2022-01-0806

Conventional forensic analyses of collision avoidance behaviours involve assigning a perception and response time (PRT) interval to a driver which precedes the onset of evasive action. This approach relies on the investigator identifying a ‘time zero’ for when to begin the PRT interval, which nominally aligns with the first moment of hazard detection. However, depending on the incident circumstances, identifying the initial moment of hazard detection poses challenges. Recent research has shown the potential of an alternative method of forensic analysis that is based on the relationship between response onset and the projected time-to-impact (TTI). Accordingly, the aim of this research is to further investigate this relationship.

Twenty subjects viewed driver-perspective recordings of a simulated vehicle travelling down a major roadway past several two-way stop-controlled minor roads. At various intersections, intruding vehicles entered from the left or right. These intruders were either initially stationary at the stop bar and accelerated into the major roadway, or were travelling at a constant speed from a distance. Other vehicles turned left in front of the subject’s path from the opposing direction (i.e., left turn across path, or LTAP). All intruding vehicles entered the lane ahead or were first visible if in motion either 3 or 6 seconds prior to impact. Subjects were instructed to brake their vehicle when they perceived that evasive action was required to avoid a collision.

Results revealed that subjects responded within a narrow TTI range, on average, to lateral intruders regardless of the intruders’ kinematics. However, responses to LTAP intruders did not conform to the prevailing trends, suggesting that the hazard perception processes may differ depending on the type of collision threat. These results advance our understanding of the perceptual cues used by drivers when confronted with collision hazards.