The issue of road accidents is very critical in Greece and internationally, being the cause of death worldwide for more than 1.2 million people per year. The aim of this postdoctoral research is to correlate road accidents with high-resolution traffic and weather data to understand and hence reduce real-time crash risk. To this end, traffic, weather and accident data on a motorway will be used.

Initially, the critical factors that affect the likelihood, frequency and severity of accidents are examined. Particular emphasis will be given to developing (for the first time in this field) specific multi-parametric regression models that take into account the high correlation between the independent variables. In this way, a deeper understanding of the phenomenon will be achieved, since some independent variables will not have to be excluded because of their high correlation with other important independent variables.

The present research will also explore the application of alternative methodologies such as cusp catastrophe theory, especially for the study of accident frequency as the intense non-linearity will allow exploration of critical factors that have a strong influence on road accidents and increase dramatically the risk of an accident. This will enable to implement interventions which will have the most cost-effective effect (at minimum cost).