Design and Development of Autonomous Drone to Detect and Alert Intruders in Surveillance Areas

Abstract

The design and development of Autonomous Drone to detect and alert intruders in the surveillance area is a technology to support law enforcement personnel involved. To assess the situation, and to detect persons entering the surveillance area to alert relevant authorities. The structure is made of plastic and carbon fiber in the shape of a hexagon Autonomous Drone. The control unit is divided into flight controller and intruder detection camera control unit. The flight controller uses the Pixhawk board with: GPS module, Telemetry module, Receiver module and ESC module through Mission Planner control program. The intruder detection camera uses the Roboflow platform for object detection to detect people. Aviation is divided into 2 operation types. Manual is using: Stabilizer mode and Loiter mode. Automatic operation uses Auto mode to fly by waypoint working for human detection. In testing an Autonomous Drone to detect and alert intruders in the surveillance area, the test was divided into 3 parts: manual flight, automatic flight, and automatic flight with detection of people in the surveillance area. It was found that the manual flight has the highest accuracy of 97.50%. The automatic flight at 100 meters has the highest accuracy of 98.50%. Lastly, the automatic flight with human detection, in detecting and alerting at a height of 3 meters, has the highest accuracy of 97.00%. In the design and development of Autonomous Drone, there were problems in the structure designed with a 3D Printer to support the weight so that fastening of the parts could not be achieved as intended, resulting in the design of a new structure. The camera detection of the Autonomous Drone had problem in that the range of the active camera and the speed of the Autonomous Drone must be appropriate for the detection to work properly.