The Role and Context of Road Geometry in Rural Driving Crashes in NSW

Abstract

<p>While Australia’s <i>National Road Safety Strategy 2010-2020</i> aims to reduce the number of road deaths and serious injuries by at least 30%, recent years have seen an increase in crashes involving casualties. Road safety problems are complex and multidimensional and some of the contributing factors - especially road geometry, speed and traffic volume - have been found to have mixed effects. These issues underscore the importance of further research in this area, including the development of new methods. Operating within a Geographic Information System framework, this research develops an empirical approach to road safety studies using exploratory data analysis and regression modelling. It considers interactions between geometrical aspects of road curvature (sinuosity index and grade) and behavioural proxies (critical visual points), and examines associations with the geographical distribution of casualty crash rates in different driving environments. Case studies include two major arterial roads (Kings Highway and Waterfall Way) and two main highways (Pacific Highway and Princes Highway) in New South Wales, Australia. This research aims to identify the role of road geometry, information complexity, road hierarchy, traffic volume and changes in driving location and time in affecting casualty crash rates. The visualisation and segmentation methods employed in this research are useful for summarising relationships between crash data and road geometry factors, and highlight some key findings.</p> <p>On arterial roads, sinuosity and complexity of visual information are the main factors affecting crash distributions during day-time, but crashes at night are randomly distributed. This is likely a result of the headlight effect filtering and simplifying visual information relating to the driving environment. By contrast, the distribution of day-time crashes on main highways does not vary significantly, whereas high crash rates at night appears to be linked to the homogeneity of information, higher traffic volumes, driver fatigue and speed. While varying according to day/night, total crash rates decrease with increasing traffic volumes, decreasing sinuosity, and simplification of driver visual cues. On the Pacific Highway, road upgrades and resultant simplified driver visual cues have had a more significant effect on reduction of crash rates at night-time. In contrary, on the Kings Highway, road upgrades and safety improvements have had a more substantial effect on reduction of crash rates during the day-time. Some outcomes are location-specific such as the possible effect of sun on the Kings Highway, fog on the Waterfall Way and speed cameras on the Pacific Highway. </p> <p>The similarities and differences in the case studies support the idea of evidence-based planning. There are two scales at which public policy formulations could operate under this framework: first, at a macroscale level where generalisations can be applied on a state-wide basis; and second, location-specific developments which apply only to particular roads. In summary, road geometry and complexity of driver visual information are the main factors underpinning crash clusters on curves during the day on arterial roads, and speed and fatigue on straight and monotonous roads at night are the main factors underpinning high crash rates on main highways. The outcomes suggest that a balance is required between road geometry and complexity of driver cues (geometrical and behavioural proxies of road curvature) for day and night environments. Different public policies, signage and speed limits might be required for day-time and night-time on different type of roads to provide the required balance and to reduce incidences of crashes on Australian roads.</p>