Exploratory Review of the Literature on Computer Simulation Models of Safety on Roads

From time to time it is necessary to assess the level of development in a research area as a basis of determining how far we have come. Recent decades have seen considerable growth in computer capabilities, data collection technology and communication mediums which has had considerable impact on our ability to replicate and understand failures in the traffic system. This paper sets out to review the use of computer models in simulating the level of safety in the traffic system. In particular, it reviews computer models of driver and vehicle behaviour within a road context. It focuses on stochastic numerical models of traffic behaviour and how reliable these are in estimating the level of safety on the traffic network. Several modelling studies have utilised existing traffic simulation software (VISSIM, PARAMICS, AIMSUM) calibrating or changing parameters to better represent the dynamics of traffic behaviour. These models appear to provide a flexible enough platform for the modelling of safety and the developers appear amenable to assisting developers in their pursuits. Most of the studies have focused on areas where video imaging can be used to provide data to calibrate and validate the models. These are parking lots, intersections and isolated road lengths. In-vehicle data (more commonly termed naturalistic data sets) have also found application in developing these models. The increase in this type of data will enable a closer link with accidents and traffic characteristics. The measures of performance used in these models are also explored. The paper concludes with some areas for future research.