Riku Kiboshi

Study on the advancement of snow removal patrols using LiDAR sensing

Kazuyoshi Takahashi

Snow removal is a necessary part of life in areas with heavy snowfall. However, with the declining birthrate and aging population, the number of snow removal operators is decreasing, making it necessary to improve the efficiency of snow removal operations as a whole. In Nagaoka City, snow removal patrols survey the road surface conditions and measure the depth of snow accumulation to make decisions on dispatching snow removal vehicles. While snow removal operations are becoming more efficient through research into ICT, snow removal patrols also need to become more efficient and labor-saving. 
In this study, we examine whether LiDAR sensing technology can acquire data that can simplify and improve snow depth measurement for snow removal patrols. We focused on iPhone LiDAR because of its low cost and ease of data sharing, and compared the cross-sectional shape of snow cover recorded by iPhone LiDAR with that recorded by N-Quick at three locations. The difference in elevation between the two point clouds increased with distance from the measurement location at the two points, suggesting that the iPhone LiDAR may not reproduce the horizontal plane correctly because N-Quick uses GNSS/INS for attitude correction. When the difference in snow depth was calculated within a range of 1 m around the road surface, the difference was 7 cm at one point and 1 to 3 cm at two points. According to the results of an interview with a snow removal contractor, the allowable error when comparing the snow depth calculated from the point cloud recorded by N-Quick and the actual measured snow depth was approximately 3 to 5 cm, suggesting that it is possible to grasp the snow cover situation over an area within 1 m of the measurement position. However, it is necessary to devise a measurement method that enables the iPhone LiDAR to correctly measure the horizontal plane.