Publication

Abstract : Navigating urban environments safely remains a pressing challenge, particularly in high-density regions with varying safety conditions. This work proposes a machine learning-based, safety-aware route optimization system tailored for South Bengaluru. Using OpenStreetMap (OSM) data, the system integrates historical crime records, streetlight density, and crowd dynamics to compute safety scores across route segments. A Random Forest classifier predicts risk levels in real time. Experimental results show the system achieves 96percent classification accuracy and adjusts route recommendations with an average computation time increase of 12–15 percent over tra-ditional routing. The platform, powered by folium for interactive map visualization, allows users to prioritize safety or travel time. This paper demonstrates a scalable, data-driven approach for safer urban navigation and highlights potential for real-world integration with smart city systems.