S. Mitra, K. Utsav

Pages: 65-76

Abstract
Driving in fog is a challenging task. ITS application through Changeable Message Sign (CMS) is one of the popular ways to assist drivers in fog. However, an evaluation done on California’s Automated Warning System (CAWS) showed that “a generic advisory speed” for a traffic stream is not effective in modifying driver’s behavior for better safety. On the other hand existing literature shows that “in-vehicle headway feedback” may be effective in modifying driver behavior. However, to provide such feedback, it is essential to build understanding about safe headways in various visibility conditions. Estimate of such safe headways may be obtained by analyzing drivers’ car-following behavior in reduced visibility. In this study real life data on fog driving, containing records such as time of arrival and departure, speed, headway and vehicle lengths are used to study driver’s headway maintenance with speed. Investigation is done to check headway variation with respect to speed at the study location. The results indicate no variation across sites but clear variation of headway with respect to speed across lanes. It is also observed that for a particular speed under similar visibility condition, headway value is the least in the fast lane. Graphs with cumulative distribution of headways are plotted to find 85th percentile headways across various sites and lanes for a particular visibility level. To measure the propensity and severity of collision Time to Collision (TTC) and Potential Collision Speed (PCS) are computed and used as surrogate safety measures. Also, recommendations for safe headways are made for specific visibility condition.

Keywords: car-following; reduced visibility; fog; safe headway; TTC; PCS