Abstract
Real time traffic diversion and dynamic route guidance (DRG) are among the key ITS approaches towards improving travel time at both the individual as well as network levels. The impact of such systems on network safety manifested in the number of accidents per unit time is largely unknown. Such impact is presumed to be positive. On the other hand, traffic diversion implies more activity in a traffic network, which might have an adverse impact on safety. The net outcome therefore is yet to be quantified. The objectives of this paper are to explore the variation of network-wide accident numbers caused by traffic redistribution subject to various levels of DRG market penetration, and to examine the potential of a new safety-conscious route guidance system. A microsimulation model was developed and integrated with a set of accident prediction models for links and intersections. Accident estimates were plotted against time to produce temporal accident profiles over a period of peak traffic. These accident profiles, together with the average travel time, were used to explain the relationships between DRG market penetrations and the number of network-wide accidents. The integrated simulation model was also applied to enhance DRG by suggesting routes with the fewest estimated accidents.