A Robust LiDAR-Based Indoor Exploration Framework For UAVs in Uncooperative Environments

Open Webpage

Wolfgang Bolz, Marco Cella, Noah Emanuel Maikisch, Francesco Vultaggio, Francesco d'Apolito, Felix Bruckmüller, Christoph Sulzbachner, Phillipp Fanta-Jende

Published: 19 Jun 2024, Last Modified: 02 Mar 20262024 International Conference on Unmanned Aircraft Systems (ICUAS)EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: This paper introduces a robust frame-work for autonomous exploration using UAVs in uncooperative indoor environments, i.e. destroyed, unknown, and with limited visibility. The proposed system integrates a LiDAR-based Simultaneous Localization and Mapping (SLAM) algorithm, a Euclidian/Truncated Signed Distance Field (E/TSDF) mapping approach, a frontier-based exploration logic, a B-Spline motion planner, and a parallel Collision Avoidance (CA) module to ensure safe and efficient navigation through unexplored areas in real-time. Successful validation in simulated environments demonstrates the system's robustness in handling real-world issues such as odometry uncertainty, depleted point clouds, and intermittent networking in distributed setups.