Detail Publikasi
Abstrak
Every year 20% of emergency patient death is blamed on traffic jam delays. More than 50% of heart attack cases reach the hospital late. The major problem is that no one is responding until ambulance reaches them, it became a hard time for an ambulance to reach its destination. To avoid such situations, most of the European countries introduced “Police motorcycle Escorts” for clearing the way. To accomplish this, we propose Drones which are designed by neural network that travel through streets of city safely. It is designed as a fast residual network of eight layers, two outputs are produced by the drones for each single input image drone navigates using a steering angle while avoiding obstacles, and dangerous situations are recognized by a collision probability to let the UAV take action accordingly. Deployment in urban environments has so far been limited to certain extent.