2021/03/03 by David D. Fan, Fan, David D., Kyohei Otsu +11 · 6 citations
Computer Science · Engineering · #Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #FOS: Electrical engineering #Robotic Locomotion and Control #Robotic Path Planning Algorithms #Robotics (cs.RO) #Robotics and Sensor-Based Localization #Systems and Control (eess.SY) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.2103.02828
openalex publication_date 2021/03/03 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Although ground robotic autonomy has gained widespread usage in structured\nand controlled environments, autonomy in unknown and off-road terrain remains a\ndifficult problem. Extreme, off-road, and unstructured environments such as\nundeveloped wilderness, caves, and rubble pose unique and challenging problems\nfor autonomous navigation. To tackle these problems we propose an approach for\nassessing traversability and planning a safe, feasible, and fast trajectory in\nreal-time. Our approach, which we name STEP (Stochastic Traversability\nEvaluation and Planning), relies on: 1) rapid uncertainty-aware mapping and\ntraversability evaluation, 2) tail risk assessment using the Conditional\nValue-at-Risk (CVaR), and 3) efficient risk and constraint-aware kinodynamic\nmotion planning using sequential quadratic programming-based (SQP) model\npredictive control (MPC). We analyze our method in simulation and validate its\nefficacy on wheeled and legged robotic platforms exploring extreme terrains\nincluding an abandoned subway and an underground lava tube.\n