System for Detecting Urban Areas with High Density of Vehicle Incidents and Issues

This paper presents an on-board perception and pre-collision system, aimed at detecting traffic risky situations in urban areas, as well as trying to avoid collision with other traffic participant deploying automatic braking and/or evasive steering maneuver. Information related to the detected risky situations is recorded in a vehicles' local database. Local data are then transferred later into a global database that can be exploited for finding out remarkable traffic hotspots. This analysis is aimed at analyzing risky situation patterns, optimize resource allocation, and improve emergency call responses.