🤖 AI Summary
To address the safety navigation challenge posed by unstable or陷ing terrain—such as soft soil, puddles, and shrubbery—in unknown rough environments for legged robots, this paper proposes a collapse-proneness perception and traversability analysis framework integrating vision, geometry, and tactile sensing. We introduce a force-controlled active ground probing mechanism and, for the first time, model ground mechanical response as a quantitative collapse risk metric—overcoming the limitations of purely visual or geometric approaches under high terrain uncertainty. Leveraging RGB-D semantic-geometric joint segmentation and multi-source sensor fusion, we construct both global and local traversability occupancy grids, enabling safe and optimal path planning. Extensive validation in simulation and on a real quadrupedal robot demonstrates significant improvements in gait stability and navigation success rates across complex field scenarios, including soft soil, standing water, and dense vegetation.
📝 Abstract
Inspired by human behavior when traveling over unknown terrain, this study proposes the use of probing strategies and integrates them into a traversability analysis framework to address safe navigation on unknown rough terrain. Our framework integrates collapsibility information into our existing traversability analysis, as vision and geometric information alone could be misled by unpredictable non-rigid terrains such as soft soil, bush area, or water puddles. With the new traversability analysis framework, our robot has a more comprehensive assessment of unpredictable terrain, which is critical for its safety in outdoor environments. The pipeline first identifies the terrain’s geometric and semantic properties using an RGB-D camera and desired probing locations on questionable terrains. These regions are probed using a force sensor to determine the risk of terrain collapsing when the robot steps over it. This risk is formulated as a collapsibility metric, which estimates an unpredictable region’s ground collapsibility. Thereafter, the collapsibility metric, together with geometric and semantic spatial data, is combined and analyzed to produce global and local traversability grid maps. These traversability grid maps tell the robot whether it is safe to step over different regions of the map. The grid maps are then utilized to generate optimal paths for the robot to safely navigate to its goal. Our approach has been successfully verified on a quadrupedal robot in both simulation and real-world experiments.
