Towards a user-centred road safety management method based on road traffic simulation

Abstract

One of the most important gaps in road safety management practises is the lack of mature methods for estimating reliability. Road safety performance assessment systems have been developed; however, these provide only historical or retrospective analyses. Effective safety management requires a prospective viewpoint. The main goal of this research is to assist in reducing accident rates in Cyprus by providing ample time to the authorities to react to high risk situations through a safety prediction early warning system. This ultimately will prevent accidents from occurring which subsequently could save lives. Traditional approaches focuses solidly on empirical data concerning road network dynamic properties, despite the fact that the most vulnerable component of the system is the human element. This paper described the integration of agent-based simulation with Bayesian Belief Networks (BBN) for improved quantification of accident probability. The BBN is developed using multidisciplinary influences.