Automation of a Drone Swarm Surveillance System

Sammanfattning

This thesis addresses the need for a scalable and flexible drone-based surveillanceplatform for the autonomous vehicle testing industry. The platform should supportcooperation between drones from different manufacturers through a unified communicationinterface and automatically generate missions in response to detectedobjects. The project was conducted as a collaboration between Chalmers Universityof Technology, Sweden, and The Pennsylvania State University, USA, for AstaZero,a research and testing facility for automated and connected vehicles. The developedplatform integrates a backend system responsible for drone coordination, missionplanning, and communication between system components with a web-based operatorinterface. The interface allows operators to define surveillance areas on aninteractive map, monitor live video streams and telemetry data, and manage thegenerated missions when objects are detected. The system also evaluates availabledrones based on hardware capabilities and operational status to autonomously selectthe most suitable one for a selected mission, and determine appropriate flightparameters for effective surveillance coverage. Experimental validation, includingon-site testing at AstaZero, demonstrated the platform’s ability to communicatewith multiple drones, dispatch missions, stream real-time video, and integrate objectdetection into the surveillance workflow. Limitations regarding object detectionreliability and hardware dependencies were identified during the project. Nevertheless,the project shows the feasibility of a scalable multi-drone surveillance platformfor autonomous vehicle testing environments.