Influence of infrastructure parameters on the frequency and severity of accidents in Germany and Great Britain in 2016, -using OpenStreetMap data

While accident databases of European countries are heterogenous, an enlargement by infrastructure factors provides fundamental data on road safety and permits better data comparison. This paper provides a method to connect infrastructure parameters and accident databases and examines the safety effects of infrastructure factors on predicted accident occurrence along motorway and primary road sections in Germany and Great Britain. The infrastructure variables of OpenStreetMap data are linked to the crash information of the accident databases, accessible at the Fraunhofer Institute. The accident data of 2016, traffic characteristics, road infrastructure parameters are used in the investigation. Starting with a Negative binomial regression, the estimation results of the coefficients indicates that the influence from infrastructure parameters on the number of accidents vary among road categories and the accident severities. The results of the analysis show that the most important variables associated with accidents are the length of a section, the number of lanes and the annual average daily traffic, followed by the tempo limit, which is a significant parameter only on primary roads. The regression model predicts in average more accidents than in Great Britain, except the number of accidents with slight injuries on rural motorways.