Testing Unmanned Aircraft for Roadside Snow Avalanche Monitoring

Abstract

An important part of the Norwegian Public Roads Administration’s (NPRA’s) mission is to monitor and react to snow avalanche hazards in steep areas above their roads. Small unmanned aircraft systems (UASs) are increasing capable and commercially available. The NPRA wanted to evaluate if this technology could support their avalanche program but had concerns about UASs’ ability to function and to provide usable information in typical field conditions. The NPRA hired vendors to fly nine different UAS aircraft on challenging avalanche surveillance missions in cold, windy weather in steep mountains. An evaluation team’s conclusions were that different aircraft could be effectively used for different aspects of avalanche monitoring and that cameras on UASs (both visible light and infrared) provided images useful for avalanche surveillance. A second test further evaluated if sensors and cameras carried on UASs could provide NPRA’s avalanche staff with usable snow depth, layering, and surface information. Data from UAS carrying visual cameras, used for both real-time viewing and photogrammetry (structure from motion), and ground penetrating radar (GPR) were compared with snow pack information collected using hand dug pits. The review team found the GPR output could identify the weak snow layers that cause avalanches but the raw data required time-consuming post-processing. Photogrammetry had considerable potential as it mapped snow surface conditions and, with a baseline survey, measured snow depth, both of which are valuable for avalanche assessment. Overall, the review team felt UAS technology held considerable promise to support transportation agencies’ roadside avalanche monitoring programs.