Determining new approaches in road safety based on driver-road dynamics interaction

The primary objective of this study is to develop a safety-oriented and holistic analytical framework that extends beyond conventional physics- and geometry-based evaluation approaches widely used in highway geometric design by incorporating psychophysiological indicators reflecting driver behavior and responses. In this context, the adequacy of existing highway design criteria in terms of driver safety has been critically examined, and an alternative evaluation framework centered on the human factor has been proposed. Within the scope of the research, traffic accident reports from Hatay Province, Türkiye, were examined in detail, and traffic accident spots (TAS) were identified. The geometric characteristics of the identified critical road segments were analyzed using physical and geometric formulations commonly employed in highway design. Subsequently, these segments were transferred into a driving simulation environment, where controlled experimental driving tests were conducted with real drivers. During the simulation process, drivers' heart rate data were recorded in real time using smartwatches, while facial thermal color distributions were simultaneously captured using thermal camera systems, resulting in a multi-layered and integrated dataset representing drivers' psychophysiological responses. The effects of road geometric features, route conditions, and drivers' experiential characteristics on speed behavior, facial thermography, and heart rate variations were analyzed using Linear Mixed Models (LMMs). The findings indicate that pronounced physiological stress responses may occur even on road segments that are considered geometrically adequate according to conventional design criteria. Moreover, among the examined dependent variables, the heart rate model exhibited the highest model fit, with the lowest AIC (19,051.66) and BIC (19,683.71) values. These results demonstrate that identifying traffic accident locations alone is insufficient for comprehensive traffic safety assessment and highlight the critical necessity of additionally defining "psychophysiological accident spots" that reflect drivers' physiological and cognitive responses.