USN Based Early Vehicle Fire Detection System in the Tunnel

This is a study on a system wherein vehicles are detected in tunnels with various sensors by recognizing patterns through acoustic, infra-red and RFID tag analyses; recognizing emergency situations by identifying atypical patterns; and forecasting accidents that are very likely to immediately follow such patterns to enable early response. Cognition was realized by assuming two lanes in a tunnel and analyzing the vehicle progress patterns to mirror real conditions. Before operating an existing automated fire warning system that recognizes secondary accidents such as fire in tunnels, a combination of Arduino, sensing equipment and wireless equipment was used to form a ubiquitous sensor network that warns of disasters for disaster control and furnishes early warning data.