Modeling real-time driver behavior when overtaking bicycles on rural roads

Despite the voluminous number of studies performed regarding bicycles in urban settings, very few studies of a similar nature have been performed on rural roads. Most rural studies are performed using only crash report data. Other authors indicate that the lack of available field data (especially collected in real-time) or a means to readily collect this field data is the major hindrance to performing similar studies. While rural crashes account for approximately 8 percent of all bicycle-motor vehicle crashes in Wisconsin annually, these crashes account for more than half of all fatalities, and approximately 14 percent of all incapacitating crashes. The disparities in fatal and incapacitating crash rates indicate that rural roads are significantly more hazardous for bicyclists.;Driver behavior has long been researched in the passing maneuver context, but never before in real-time in the field. Prior research efforts have used experimental test tracks and prepared situations to evaluate driver judgment, reaction, and decision-making among other factors for vehicle-vehicle passing. These studies, conducted over the course of several decades, found similar results regarding reaction time and distances, although most interestingly found that drivers often underestimated the distances needed to perform certain passing maneuvers. This misjudgment on the part of drivers is noteworthy as it represents a common cause of collision noted in the analysis of vehicle-bicycle crashes.;This research developed a means to collect this real-time field data, and deployed this methodology through the use of a bicycle-mounted data collection system. The collected data was then used to develop the first-ever model to evaluate lateral clearance distance between vehicles and bicycles. The developed model confirms various field observations, such as when road speed limit increases, so does the lateral clearance distance provided by drivers, and that when an oncoming vehicle is present, the lateral clearance distance provided decreases.;This new model will enable stakeholders to identify those roadway segments where a paved shoulder would prove an effective safety countermeasure based on existing roadway conditions. This model will also enable roadway designers to better predict, during the design phase, those roadway segments that should be constructed with a paved shoulder.