SubT-MRS

CMU AirLab · 2024 · datasets.bot · datasets.bot page

One-liner. Real-world all-weather multi-robot SLAM dataset spanning the DARPA Subterranean Challenge and extensions (2019-2023), with LiDAR, fisheye/depth/thermal/RGB cameras and IMU on aerial, legged, wheeled and RC-car robots in fog, dust, smoke and snow.

Setup

Schema

Per-sequence ROS bags and extracted folders containing time-synchronized multimodal sensor streams (up to 4 RGB + fisheye cameras, LiDAR point clouds, IMU, thermal camera). Ground-truth trajectories in TUM format (timestamp x y z q_x q_y q_z q_w) plus initialization poses.

Links