Enhancing cycling is an increasingly important challenge, especially as it is promoted for its economic, environmental, and health benefits. However, ensuring safety of cyclists is crucial to support this shift in mobility. In this context, machine learning offers promising avenues. This study proposes a novel approach to identifying high-risk locations by dynamically incorporating spatio-temporal patterns and environmental conditions. The method was tested using comprehensive data from Germany, and its design suggests strong potential for generalization to different countries. This work can support urban planners, policymakers, and navigation systems in improving road safety and informing smarter mobility decisions
Data-Driven Prediction of High-Risk Situations for Cyclists through Spatio-Temporal Patterns and Environmental Conditions
Sarah Di Grande
;Mariaelena Berlotti;Salvatore Cavalieri;
In corso di stampa
Abstract
Enhancing cycling is an increasingly important challenge, especially as it is promoted for its economic, environmental, and health benefits. However, ensuring safety of cyclists is crucial to support this shift in mobility. In this context, machine learning offers promising avenues. This study proposes a novel approach to identifying high-risk locations by dynamically incorporating spatio-temporal patterns and environmental conditions. The method was tested using comprehensive data from Germany, and its design suggests strong potential for generalization to different countries. This work can support urban planners, policymakers, and navigation systems in improving road safety and informing smarter mobility decisionsI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.