Knowledge Graphs for Autonomous Robotic Missions

A new methodology integrates knowledge graphs into ROS 2 systems to improve autonomous robotic mission efficiency. The approach defines initial and target conditions, structures tasks and subtasks, sequences them, represents data in a knowledge graph, and designs missions with a high-level language. A simulated search and rescue mission in Gazebo using Aerostack2 framework demonstrates drones autonomously locating a target, showing enhanced decision-making and performance. The paper is available on arXiv.