This study analysed the relationship between the size of the forward looking blindspot (FLB) produced by vehicles A-post (windshield frame), the speeds of two vehicles approaching an intersection at right angles, and driver behaviour relative to a likely accident event. Researchers observed 28 volunteer participants directly and by four channels of on-board video cameras while they drove in a simulator at the Human Factors Research Laboratory. They noted the way that participants scanned the virtual environment and scored at four levels of scanning activity. They also tracked visual acquisition of the target vehicle and incidence of collision. Only 6.3% of the total fell into type one scanning (eyes fixed). Type II (eyes only) accounted for the highest incident rate at almost 44%. The study considered both as "inactive" forms of scanning. Target vehicle acquisition rate increased with the activity level of the scanning type. The target acquisition rate increased significantly from scanning level one to level two and from scanning level two to level three. There was not a significant increase in the acquisition rate from scanning level three to level four. Not surprisingly, collision rates decreased with increases in scanning level. Collision rates significantly dropped between scanning levels two and three and scanning levels three and four. Yield signs at intersections produced no significant correlation with acquisition rate, collision rate, or scanning level.
Abstract