Modeling the Risk of Wrong-Way Driving on Freeways and Toll Roads

Abstract

Wrong-way driving (WWD) is dangerous and poses a significant legal and safety risk when it occurs on limited access facilities. Most previous studies focused on WWD crashes to develop countermeasures. The combined risk of WWD citations and 911 calls, however, has been overlooked. Furthermore, because WWD crashes are rare and represent less than 3% of all crashes, such crashes are difficult to analyze. WWD prediction is an important assessment because it can help mitigate and reduce future occurrences. This paper builds on previous work pioneered by the authors in which WWD crashes were predicted with the use of WWD noncrash events (e.g., citations and 911 calls). These WWD noncrash events occur more frequently, and their data are widely available. The paper demonstrates how WWD 911 calls and citations, along with route characteristics, can be linked to WWD crashes and so target corridors for countermeasures. Two models were developed and applied in South Florida to identify WWD hot spots. The first model shows that WWD citations and 911 calls positively affect yearly crash prediction. The second model identifies hot spot segments in a route and predicts crashes during a 4-year period. This second model predicts crashes with the use of several variables, such as major interchanges per mile, directional interchanges per mile, and WWD 911 calls along the segment. The findings showed high WWD risk values on SR-821 (Homestead Extension) from Exits 20 to 39 and on SR-9 (I-95) from Exits 0 to 6B and Exits 7 to 14.