An IoT Based Traffic Management System Using Drone and AI

Management of ground traffic on both urban streets and highways in a smart city setting requires collecting a huge amount of logistical data. Accessing real-time information of the traffic is essential in the event of an emergency. It requires the traffic control center to regularly monitor flows of vehicles and take suitable actions to reduce traffic jams. Several tiny devices are needed to collect and transmit real-time data from different locations. However, the bandwidth and power consumption of each device is very limited. Therefore, it is essential to utilize an efficient algorithm which reduces the bandwidth, as well as power consumption. In this paper, an efficient method is proposed to reduce the transmission bandwidth while keeping the quality of the videos acceptable for image processing on the server end. To evaluate the performance of the algorithm, a framework to monitor and control the traffic on highways is developed. This framework uses a drone to fly over the traffic to capture the logistical information, and then send real-time video to the server. An object detection algorithm empowered by artificial intelligence (AI) is implemented on the cloud server that detects the number of type of vehicles, and accordingly makes decisions to manage traffic flow.