Wildlife-vehicle collisions (WVCs) threaten the health of road users and wildlife alike. Furthermore, they cause high insurance claims every year. Although influencing factors have already been investigated in various studies, developing a robust model for risk assessment, especially for individual journeys and road sections, remains challenging. 

Challenges are, in particular, the data availability, heterogeneous area-specific conditions, and a large amount of influencing factors. In our research we address those challenges by evaluating the feasibility of AI-based systems for the risk assessment of WVCs by discussing requirements, available data, and methods. 

In this presentation we give an overview over our current and future research. Examples will be based on the WVC data provided by the German federal states.