A Distributed Software Framework for Vision-Based Drone Swarm Applications

Wei Li, Taoying Liu

Abstract: This paper presents a distributed software framework for drone swarm applications, focusing on the OWL-Swarm framework. The framework addresses challenges in task allocation and coordination inherent in swarm robotics. It introduces a modular architecture based on the Robot Operating System (ROS), enabling seamless integration of various algorithms and functionalities. The Captain service acts as the central orchestrator, managing mission execution, task allocation, and data sharing among drones. The framework supports both single-drone and collaborative tasks, with mechanisms for distributed scheduling and dynamic task re-allocation. The effectiveness of the framework is demonstrated through use cases involving cooperative obstacle avoidance, coordinated search and target tracking. The modular and extensible design of the framework allows for easy customization and adaptation to various drone swarm applications, making it a valuable tool for research and development in this field.