Autonomous UAV Navigation and Exploration in Tunnel Environments

This research develops novel algorithms for autonomous UAV navigation and exploration in tunnel environments, addressing critical challenges of GPS-denied navigation, geometric constraints, communication limitations, and environmental hazards in underground spaces. The work creates advanced SLAM-based navigation systems with visual-inertial pose estimation, LiDAR-based mapping and localization, multi-sensor fusion for robust state estimation, and loop closure detection in repetitive tunnel environments. Key technical innovations include frontier-based exploration with tunnel-specific heuristics, safe trajectory planning in constrained spaces, adaptive speed control, emergency return procedures, and compact quadrotor platforms with multi-modal sensor suites optimized for confined space operation and extended mission duration.

Infrastructure inspection and emergency response applications benefit from autonomous transportation tunnel monitoring, utility assessment capabilities, mining safety inspections, search and rescue operations, and disaster assessment mapping. The work demonstrates significant improvements in navigation accuracy, exploration coverage efficiency, collision avoidance performance, and computational resource utilization through comprehensive validation in underground parking structures, transportation tunnels, and mining environments. Industry partners and emergency response agencies can leverage this expertise for developing next-generation autonomous inspection systems, advancing GPS-denied navigation technologies, implementing robust SLAM algorithms for challenging environments, and creating reliable UAV platforms that enable safe tunnel exploration, hazardous environment assessment, and communication relay deployment for critical infrastructure monitoring and emergency response applications.

Acknowledgments

This work was supported by research grants and collaborations with transportation authorities and emergency response agencies. We thank our partners for providing access to testing facilities.

---

Note: Full content extraction from PDF required for complete details. Please refer to the original publication: 23-ral-zhefan-xu-tunnel.pdf