RTISSLAM: In-air Imaging Sonar SLAM using a Robust Pose Graph and Occupancy Grid Mapping

Key points

• Sonar was one of the first sensors with which mobile robots built maps, and in-air sonar keeps working where cameras and lidar fail, in darkness, smoke, dust and next to glass.1
• Classical sonar SLAM, however, relied on geometric features or occupancy grids built from sparse range readings, and both struggle with the specular reflections that dominate indoor sonar.1
• Energyscapes from delay-multiply-and-sum beamforming with a range-wise CFAR serve as place descriptors for sequence-verified loop closures, acoustic odometry calibrates the wheel odometry online, recognized revisits are registered metrically, and the grid is stored per keyframe and anchored to the pose-graph nodes, so that it follows every loop closure.1
• In simulated worlds with specular walls, RTISSLAM reduces the trajectory error of the wheel odometry from \SI{8.2}{\meter} to \SI{0.31}{\meter} in an office, while lidar-style grid SLAM on the same sonar data fails.1

Sources (1)

• [1]RTISSLAM: In-air Imaging Sonar SLAM using a Robust Pose Graph and Occupancy Grid Mapping
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 13, 02:40 PM

  In this letter, we introduce RTISSLAM, a novel sonar-only SLAM system for an embedded 3D imaging sonar, which combines the robust pose graph of BatSLAM 2.0 as back-end with an occupancy grid as map.

  Sonar was one of the first sensors with which mobile robots built maps, and in-air sonar keeps working where cameras and lidar fail, in darkness, smoke, dust and next to glass.

Extractive summary: sentences quoted from the sources.