LEARNING THROUGH DOING: LESSONS FROM LiDAR-ACQUISITION DRONE FLIGHTS IN MARYSVALE, UTAH

Autonomous drone flight has emerged as a revolutionary technology with diverse applications including emergency response, and geoscience outreach and education. The Geologic Hazards Program (GHP) at the Utah Geological Survey (UGS) utilizes drone technology as we aim to reduce risk by responding to various geologic hazards, including landslides, post-wildfire debris flows, and surface-rupturing earthquakes. However, there remains a learning curve to becoming a competent and useful drone pilot. In this presentation, we describe our experience learning to fly drones in an emergency response, including our goals, challenges, and insights.

We recently responded to Marysvale, Utah, where the town’s mayor requested assistance with flooding and debris flows caused by the recent Silver King wildfire. Three geologists from the GHP, one FAA-certified pilot and two pilots-in-training, conducted 48 flights with our DJI Matrice 300 RTK drone equipped with a Zenmuse L1 lidar sensor. After planning the flight path, the programmed drone flew its course autonomously, with manual intervention as needed for launch and landing.

The flight planning process was challenging due to weather, terrain, and limited power supply and viable launch locations. Assistance from the local community provided knowledge of the study area’s weather and terrain, helping tackle unexpected challenges. Temporary flight restrictions also needed to be navigated by consulting with the local incident command of the fire area. Several flight locations were in steep, narrow canyons requiring special consideration for the environment. Pilots and visual observers were staggered throughout the flight path to maintain line of sight and clear communication. We learned to adjust flight missions to shorter, canyon-parallel paths to accommodate the steep canyon topography and prevailing winds while optimizing battery power.

Our flights resulted in lidar point clouds and structure from motion models for four canyons in Marysvale, providing information that the town can use to mitigate hazards and valuable experience for the pilots in training.