Paper presented at ISTVS 2025 | 55th Conference of the International Society for Terrain-Vehicle Systems https://doi.org/10.56884/GPMZPXQX
Authors: Brian Brady, Aaron Meyer, Sergey Vecherin, Mike Parker
Keywords: LiDAR; snow; simulation; autonomous; vehicle
Abstract:
The sensors used by a robust, off-road autonomous vehicle must be capable of detecting obstacles in any weather conditions, but this becomes especially important during inclement weather with reduced visibility. Snowfall is no exception to this and poses extremely challenging conditions to vehicles exclusively employing Light Detection and Ranging (LiDAR) systems for object detection. Snowflakes may occlude “hard” targets in the vehicle’s path of travel or introduce noise from “soft” detections of the snowflakes themselves. This study highlights preliminary results from a simulation that reproduces these effects within the U.S. Army Corps of Engineers’ Virtual Autonomous Navigation Environment (VANE). The simulation accurately models snowflake size, concentration, and velocity while accounting for sensor-specific modes of operation. VANE simulations provide vehicle designers with a rapid and affordable method to evaluate LiDAR perception and autonomy response in snowy conditions.
Last updated
