LiDAR-Based Exploration and Mapping Strategy for Autonomous Multi-Drone Forest Operations

Kurzfassung

In recent years, autonomous Unmanned Aerial Vehicles (UAV) have become increasingly important for applications such as environmental monitoring, forest inspection, search and rescue operations, and autonomous exploration. In particular, forest environments present significant challenges due to the presence of dense vegetation, complex three-dimensional (3D) structures, and limited visibility, making reliable environment perception essential for safe and efficient operation. To address these challenges, Light Detection and Ranging (LiDAR) sensors have become one of the most widely adopted technologies for outdoor robotic applications because they provide accurate depth measurements and detailed 3D representations of the environment. However, the cost, weight, and power requirements of these sensors often make it impractical to equip every drone in a multi-UAV system with a dedicated LiDAR sensor. In this context, this thesis presents a framework that enables a LiDAR-equipped drone to generate a 3D occupancy map of the environment and share it in real time with other drones through a wireless mesh network. The proposed solution integrates LiDAR sensing, MavLink-based positioning information, 3D map generation, ROS2 communication, and map synchronization mechanisms to provide a common representation of the environment across multiple UAV. The work also includes a comparative evaluation of different 3D mapping frameworks in order to identify the most suitable solution for real-time map sharing under the computational and communication constraints of embedded aerial systems. Based on the obtained results, OctoMap was selected as the mapping framework used in the final implementation. Finally, a series of laboratory and field experiments were conducted to validate the proposed architecture. The results demonstrate the successful generation, transmission, reconstruction, and synchronization of occupancy maps between multiple drones, confirming the feasibility of distributed map sharing for cooperative multi-UAV operations in forest and outdoor environments.