A first investigation of perceived safety patterns in Athens, Greece

This study performs an in-depth investigation into the existing road network system of Athens, Greece that currently does not facilitate the use of popular micro-mobility modes. To prove it, this study utilizes a subjective factor called perceived safety. Perceived safety is influenced by the road environment, while it affects various travel decisions, like mode or route choice. An already-developed perceived safety ordinal logit model is applied to estimate the safety scores per road link and transport mode: car, e-bike, e-scooter and walk. The model independent variables are the cross-section type road, the condition of the pavement, the presence of obstacles on the sidewalk as well as the presence of pedestrian crossings. This analysis revealed diverse spatial patterns of perceived safety that have not been investigated yet. Safety perceptions are at high, sufficient levels for car drivers and pedestrians, while at below average levels for e-bike and e-scooter riders. By performing a kernel density analysis in perceived safety scores, the few safe corridors that can be followed by e-bike users are pointed out. High-high spatial autocorrelation But there is a notable absence of safe routes suitable for using e-scooter. The spatial autocorrelation analysis showed some high-high areas in the city center of Athens, where e-bikes can be utilized. However, the perceived safe network is still heterogenous and far from being complete.