A Methodological Approach to Identifying Unsafe Intersections for Micromobility Users: A Case Study of Vilnius

Cities are increasingly integrating micromobility, which heightens the need for robust analytical methods to identify high-risk intersections. This study presents a three-stage methodological approach that combines six years of accident data, spatial hotspot analysis, and calibrated floating-car traffic data to estimate exposure and calculate intersection crash rates in Central Vilnius. Testing the proposed approach identified eight high-risk intersections, with intersection crash rates (ICR) ranging from 0.044 to 0.151, indicating substantial differences in exposure-adjusted risk across the network. The validation of floating-car data (FCD) produced a determination coefficient (R 2 ) of 0.87, confirming reliable exposure estimates where traditional traffic counts are not available. One selected intersection was analyzed in greater depth using drone-based observations and conflict assessment, leading to two redesign alternatives. Both reduced conflicts, though the signalized option eliminated uncontrolled conflict points and offered the strongest expected safety improvement. The suggested methodological approach demonstrates how integrating accident data, exposure estimation, and behavioral analysis can support evidence-based scalable interventions to improve micromobility safety. Despite certain limitations, it enables the rapid identification of problematic intersections, provides site-specific safety diagnosis, and facilitates the development of data-driven design improvements to enhance the safety of micromobility users. As the world strives to shift towards greater sustainability, the concept of micromobility, defined as the use of lightweight, short-distance modes of transport, has gained growing attention among users and policymakers.