ACCIDENT REDUCTION POTENTIAL OF VISION ENHANCEMENT SYSTEM - BASED ON ACCIDENT RECORDS IN MICHIGAN

The most futuristic and hence the least defined components of the Intelligent Transportation System (ITS) are the Automated Vehicle Control Systems (AVCS). These systems extend beyond the function of transferring information among drivers and vehicles to the more challenging functions of relieving some of the driver's responsibilities for controlling the vehicle, which has the potential for a dramatic increase in safety benefits. However, there is also the potential of an increase in risks. An early candidate technology of AVCS that has little risk, is a vision enhancement system, which improves the driver's vision at night or in severe weather conditions. This study investigates the number and types of crashes that may be prevented by a vision enhancement system. The analysis was done for both freeway and intersection crashes. For each crash type, the number of day and night crashes were extracted from the Michigan accident database (1993 and 1994), and compared, to identify the crash types that have a higher rate at night. The vision enhancement system is hypothesized to enhance night vision so that the crash rate at night would be equal to that during the day. To predict the number of crashes that would potentially be prevented, the causal effect of each crash type was examined. Only crashes that had no violation (such as speeding or driving under the influence of alcohol) are assumed to be preventable. The data showed that overturning and hit fixed object crashes on freeways and hit fixed object crashes in intersections are the crash types that are likely to be reduced using a vision enhancement system. The potential crash reduction ranged from 6.2% to 12.6% of night crashes.