A decision support toolkit to inform road safety investment decisions

Abstract

Road safety practitioners are tasked with maintaining safety on their network, primarily by identifying collision hotspots to which resources should be allocated and by ensuring that road safety schemes are operating effectively. Both these tasks often require the analysis of road safety count data (e.g. regarding collisions or casualties); however, these data are frequently bedevilled by confounding statistical factors such as regression to the mean (RTM) and trend. Failing to account for these factors can lead to misallocation of resources as well as sites at risk of high counts not receiving treatment. To overcome this, methods are proposed in this paper which clean data for RTM and trend to allow for more accurate scheme evaluation, and a proactive approach for hotspot prediction. Unfortunately, these techniques require the development of complex, statistical algorithms and so can be inaccessible to some practitioners. To overcome this, user-friendly software applications have been developed and are described here, which implement the aforementioned techniques, with an additional application to investigate the potential collision contributory factors at sites.