Event

Explaining the technical roadmap for integrating contact information into SLAM localization

1 reports · 1 sources

3D Vision Workshop released a technical explanation outlining the roadmap for integrating foot contact and force/tactile information into robot state estimation and SLAM: when vision and LiDAR are limited, use kinematics and foot-ground constraints to supplement pose estimation. The IROS 2025 KD-RIEKF was validated on the Unitree Go2, and the paper reports that under its test conditions, both position and attitude estimation errors were reduced when the robot moved at high speed. The explanation states that contact information is a supplementary cue for multi-sensor state estimation and local environmental understanding, and does not replace vision or LiDAR; one of the studies only demonstrates the role of contact events in local navigation and environmental analysis, and does not achieve full tactile SLAM state estimation—the robot, under safe, low-speed, and controllable preconditions, converts necessary contact into local environmental information.