Data-Driven Mitigation of Wildlife–Vehicle Collisions in the Czech Republic

Wildlife-vehicle collisions (WVC) represent a growing problem in the Czech Republic, with significant impacts on road safety, economic losses and biodiversity. This thesis places WVC in the global and European context and demonstrates that collisions with wildlife are not purely random events, but a form of risk with a distinct spatial and temporal structure. The main objective of the thesis is to develop a data-driven business case for strategic investments in WVC prevention in the Czech Republic. Using Czech Police accident statistics the study analyses spatio-temporal patterns of WVC for the period 2018–2024 and quantifies the degree of risk concentration. The methodology is based on a consolidated ETL process, the KDE+ kernel density estimation tool for identifying collision clusters, a 5×5 km grid and concentration measures such as Lorenz curves, the Gini coefficient and the Herfindahl-Hirschman index. The analysis confirms a strong spatial concentration of WVC: a relatively small share of the road network generates a disproportionately high share of collisions and forms a stable network of hotspots recurring over multiple years. In temporal terms, the results are consistent with known biological and traffic patterns, with seasonal peaks in spring and autumn and daily maxima around dusk and dawn. Building on these findings, the thesis proposes a procedure for selecting suitable road sections for mitigation measures (e.g. olfactory repellents) and evaluates their effectiveness using model-based cost–benefit analysis. A separate chapter focuses on funding options through the Czech motor insurers’ Fund for the Prevention of Damage (Fond zábrany škod) and outlines a practical methodology for preparing successful project applications. Finally, the thesis formulates a business case showing that targeted prevention on a limited number of high-risk sections can be economically efficient while delivering substantial road safety and ecological benefits.