Safety Evaluation of Road Characteristics

Abstract: Accidents are rare and widely distributed on the road network, which poses a challenge for road safety improvement. Traditional methods alone, such as accident analyses and experimental studies, do not explain all accident causation factors. Traditional methods are well-established and are still important for specific contexts of road safety analyses. The objective of this doctoral thesis is to obtain an improved understanding and a safety evaluation of road features, in particular horizontal curves on non-residential roads. To explain the tendency of curves to trigger accidents is also another goal. It is essential to better understand interactions between vehicle, road, and driver components in a system which can explain possible mechanisms underlying accidents. The overall methodology of the proposed system includes the development, implementation, and integration of complementary research approaches. The approaches are based on different kinds of data sources that together produce more detailed results. In the methodology, accident and simulation analyses provided a framework for combining road, vehicle, and human behavior data collected from Field Operational Tests (FOT). The accident analysis in the system identified a set of Critical Road Parameters (CRP) with associated accident types. PC-Crash was used to simulate the influence of CRPs on vehicle response during overtaking maneuvers in curves. The CRPs and vehicle-road interaction results were used as a platform for devising a tool that could identify horizontal curve components and lane change maneuvers from FOT data. The accident analysis found more frequent overtaking crashes on right (inner) than on left (outer) curves. The simulations of road-vehicle interaction showed that the risk of a lane changing maneuver differed in curves, depending on curve direction and the lane to which the maneuver occurs. The simulations showed that requirements for drivers to remain safely within the road boundaries are greater if there is a lane changing maneuver in the curve. Despite this risk, the FOT analysis observed frequent overtaking and lane change maneuvers on the curves, of which 20% more lane changes occurred on right curves than on left curves. The curve entrance was found to be the most dangerous segment of a curve. Current design practice assumes the safety risk is constant when driving along horizontal curves. The results also showed that drivers consider curve radius in choosing their driving speed rather than the posted speed limit of the curves. The use of different data sources and approaches started with an open-ended analysis of accident data, as the first layer in a top-down process, and proceeded to the more specific and important findings of road curves. The system approach in the thesis made possible a safety tool whereby the empirical and simulation analyses together yielded more innovative and detailed results. Road feature analysis gave insight into how road geometry factors affect vehicle motion relevant to safety and driving strategy through curves. The findings are useful inputs for applications such as curve design reviews, selection of appropriate countermeasures, and the improvement of active safety devices.