The cold driver: Driving performance under thermal stress

Exposure to cold environments can impact complex task performance and increase cognitive and physiological error in response to thermal stress. Critically, the task of driving a vehicle requires the use of calibrated mental and physical actions to be conducted safely without error. Few studies have examined the effects of cold stress on driving performance and none have explored the potential for advanced driver safety systems to detect error. Active vehicle safety systems which monitor dangerous driving behavior due to drowsiness have been researched and developed, though technology associated with thermal stressed driving error is unexplored. The current study examined the effects of cold stress by way of skin cooling on driving simulator performance and evaluated vehicle behavior metrics for possible dangerous driving detection systems by analyzing behaviors. The presence of lateral lane position variability and velocity variability are of particular interest when speculating detection technology in literature. Forty-four healthy young adults (20 males and 24 females, age 19.97 ± 2.98 years, 4.06 years of driving experience) participated in either a thermal neutral or cooled condition. Cold stress was indexed using skin temperature, core temperature, and subjective cold perception. Participants drove through a simulated track which incorporated vehicle following, stop signs, and curved sections of road. Additional tasks incorporating dexterity and psychomotor vigilance were also used to account for physical and cognitive decrements. The results of the current study found that while the skin cooling methodology was effective in eliciting a behavioral response to the cold, physiological decrements were not observed in the cold condition. Skin temperatures from multiple locations and subjective comfort were significantly lower in the cold condition while core temperature, arm dexterity, and psychomotor vigilance were unaffected. Results from the driving simulator found no difference in lateral lane deviation or vehicle velocity across the track, suggesting traditional drowsy driving detection technology based on these metrics may not be feasible. However, the simulator task did show that participants who had higher subjective ratings of cold followed lead vehicles closer and started to brake later. Participants in the cold condition followed the lead car 22% (0.82 seconds) closer and started braking 20% (2.35 seconds) later when presented with a stop sign during the following task. The current results suggest that drivers exposed to cold environmental conditions are more likely to display aggressive driving behavior