| Summary: | This paper looks at stand-alone collision warning systems that are based on information from on-board sensors and evaluates their safety performance relative to system penetration rate. The authors developed an autonomous microscopic traffic simulator for collision warning systems, including both forward vehicle collision warning systems and side collision warning systems, then evaluated such systems through simulation. Safety performance from the perspective of drivers was evaluated using the average distance driven without an accident for both system-equipped and unequipped vehicles. Safety performance from the perspective of road administrators was evaluated using the average interval between accidents. The average distance driven without an accident for system-equipped vehicles, compared to that for a system in which no system-equipped vehicles exist, increases greatly beginning from a low rate of penetration, suggesting that increased rates of penetration are attended by even greater effectiveness. With regard to the average distance driven without an accident for unequipped vehicles, too, increased rates of penetration are attended by increased safety performance due to the collision avoidance effect of warnings produced by system-equipped vehicles. In terms of safety performance from the perspective of road administrators, that is, the average interval between accidents, evaluations indicated that safety performance increases dramatically when the penetration rate exceeds 60%. The above findings illustrate the effect of system penetration rate on the safety performance of stand-alone collision warning systems that are based on information from on-board sensors.
|
|---|
