In-vehicle decision support systems for collision avoidance at rural thru-stop intersections.

Abstract

In 2002 there were 590 fatal traffic accidents in Minnesota, and an estimated 17% of these occurred at rural thru-stop intersections. A thru-stop intersection presents a challenge to a driver attempting to cross or enter the highway because he must stop at the stop sign and wait for a gap in highway traffic. If a driver misjudges whether a gap is large enough, a high-speed collision with highway traffic may result. Previous studies have developed a prototype decision support sign (often referred to as the Icon Sign) that is aware of highway traffic and warns drivers when it is not safe to cross. The current study conducted a driving simulator experiment to test two invehicle decision support systems and compare them to the Icon Sign. The first in-vehicle system, called the Side Mirror Displays, consisted of two visual displays located on the vehicle’s side mirrors. The second system, called the Vibrotactile Seat, was a driver’s seat with left and right vibrational pads. No system clearly outperformed the others in terms of promoting safer driving behavior, nor did any improve driving performance compared to the control condition. The questionnaire and usability data showed that the Icon Sign was most preferred, with 50% of participants rating it as their top choice. The Icon Sign may have been preferred because drivers are more accustomed to gaining information from traffic signs and signals than from other systems. The Side Mirror Displays and Vibrotactile Seat were preferred nearly equally, with each rated as top choice by 25% of participants. The Side Mirror Displays were comprehended by 83.3% of participants, the Icon Sign by 62.5%, and the Vibrotactile Seat by 58.3%. The high comprehension rate for the Side Mirror Displays may have been due to the display continuously changing, giving the driver more opportunities to interpret the displays and deduce what information the displays were providing. In future work, the Side Mirror Displays should incorporate advisory messages, telling the driver when to wait instead of merely indicating the distance to approaching traffic. This could aid interpretation of the Side Mirror Displays. For the Vibrotactile Seat, the vibrational pads should be spaced farther apart so that drivers can better distinguish the left and right vibrations instead of mistaking them for a single vibration. Results so far indicate that a visual display would be easier to comprehend than a vibrotactile display when no training or explanation is provided. No results of this study indicate that in-vehicle systems are an inherently poor means of presenting traffic gap information to the driver.