This paper presents an autonomous Unmanned Aerial Vehicle (UAV)-enabled control system to handle accident incidents occurred by physical disasters in Smart Cities (SCs). Specifically, after a physical disaster like an earthquake or a tsunami wave several areas in SC are damaged. In these areas citizens are trapped and most of the cases hurt thus not being able to rescue themselves. However, there is not need to serve each incident with the same priority since there are cases of low, medium or high significance. In addition, SC infrastructure has a limited amount of UAVs to handle such cases. Concretely, after a trigger, like a phone call,the nearest UAV reaches a certain incident where a doctor make a remote diagnosis based on the video and audio provided by the camera and the microphone embedded in the UAV. After evaluating a high significance incident an ambulance is invoked to collect the injured citizen and transport her/him to the nearest hospital. Certain use cases are presented, which are evaluated with proposed metrics incorporated in the control system to infer the optimum use case. Such a use case is proposed to adopt by a SC infrastructure to handle a real physical disaster scenarios.