Probabilistic Terrain Mapping for Mobile Robots With Uncertain Localization

Abstract

Mobile robots build on accurate, real-time mapping with onboard range sensors to achieve autonomous navigation over rough terrain. Existing approaches often rely on absolute localization based on tracking of external geometric or visual features. To circumvent the reliability-issues of these approaches, we propose a novel terrain mapping method which bases on proprioceptive localization from kinematic and inertial measurements only. The proposed method incorporates the drift and uncertainties of the state estimation and a noise model of the distance sensor. It yields a probabilistic terrain estimate as a grid-based elevation map including upper and lower confidence bounds. We demonstrate the effectiveness of our approach with simulated datasets and real-world experiments for real-time terrain mapping with legged robots and compare the terrain reconstruction to ground truth reference maps.

Mine Tunnel Exploration Using Multiple Quadrupedal Robots

“…The problem of local planning is then broken into two primary pieces: quantifying terrain traversability and planning feasible trajectories on the generated cost map. Due to imperfect odometry and our decision to minimally couple planning and mapping, we adopt the approach of [11] and continually center the costmap on the robot's current position. We additionally model traversability as a function of the magnitude of the terrain gradient, which makes costmap integration extremely simple.…”

Actively Mapping Industrial Structures with Information Gain-Based Planning on a Quadruped Robot

“…Both phases rely on an elevation grid map generated from LiDAR measurements of the environment. We used the approach of [21] to compute the slope and normal of each cell and in turn a measure of the traversability of the terrain. The traversability is used to determine which states planned by the RRT and RRT* are valid and reachable.…”

Autonomous Free-Form Trenching Using a Walking Excavator

“…The mapping framework for autonomous excavation is based on robot-centric elevation mapping for mobile robots by Fankhauser et al [16]. The algorithm creates a 2.5 dimensional elevation map from any distance sensor and a corresponding pose estimate.…”