Locating and evaluating fixed safety cameras in South Australia

Abstract

Fixed safety cameras that detect red light running and speeding vehicles are known to be an effective method of controlling driver behaviour and producing road safety benefits. While no definitive best practice for choosing safety camera locations was found in the literature, there are a number of criteria that are frequently used and that make sense: locations with a high number of crashes (particularly injury crashes); locations with a high proportion of speeding vehicles; locations with high traffic volumes; locations with large numbers of unprotected road users (pedestrians and bicyclists); different camera types covering different areas of the road network; and deploying cameras widely throughout the road network. Evaluation of the effects of safety cameras on particular sites and as a whole can be attempted using a number of methodologies: changes in crash numbers before and after installation; changes in vehicle speeds before and after installation; and changes in offence detections from the time of installation onwards. There are limitations with each of these methodologies such that evaluating the effectiveness of an individual safety camera is often not possible. By tracking many safety camera sites for a long period of time, the effect of the safety cameras can be examined but there will always be other factors that may explain any observed changes. A true experiment could be conducted to determine safety camera effects but it would involve deliberate non-treatment of good candidate sites for many years. This study gives a reasonable set of principles for selecting safety camera sites and evaluating their effectiveness. However, the detailed processes for South Australia will depend on what data is available, the resources and funding that can be applied, and political and other considerations.