Telescope-based scanning LiDAR system for eye-safe long-range UAV localization and tracking

Area of Science:

• Robotics and Automation
• Optical Engineering
• Aerospace Engineering

Background:

• Accurate localization and tracking of small uncrewed aerial vehicles (UAVs) are crucial for various applications, including surveillance, inspection, and delivery.
• Existing tracking systems often face limitations in range, accuracy, or eye-safety, especially for long-distance operations.
• Developing an eye-safe, long-range localization and tracking system for UAVs is essential to overcome these challenges.

Purpose of the Study:

• To design and experimentally evaluate a novel telescope-based scanning light detection and ranging (LiDAR) system.
• To achieve eye-safe, long-range localization and tracking of small uncrewed aerial vehicles (UAVs).
• To validate the system's performance in both laboratory and field environments.

Main Methods:

• Implementation of a prototype system featuring an eye-safe laser transmitter, a one-dimensional avalanche photodiode (APD) detector array, and a telescope.
• Laboratory experiments for static and dynamic localization of a DJI Phantom 3 UAV at 300 meters.
• Field tests involving distance measurements to a DJI Matrice 30 up to 50 meters and hybrid target tracking at 300 meters with velocities up to 15 m/s.

Main Results:

• Robust static and dynamic localization of a UAV was achieved at a range of 300 meters in laboratory settings.
• Successful distance measurements to a UAV were demonstrated up to 50 meters in field tests.
• Feasible hybrid target tracking, combining LiDAR for horizontal and image-based processing for vertical movement, was shown at 300 meters and 15 m/s.

Conclusions:

• The developed telescope-based scanning LiDAR system provides an effective solution for eye-safe, long-range UAV localization and tracking.
• The system demonstrates significant potential for real-world applications requiring precise UAV monitoring.
• Hybrid tracking strategies integrating LiDAR and vision-based methods enhance tracking accuracy and versatility.