SYSTEMS AND METHODS FOR ROBOTIC CONTROL USING LIDAR ASSISTED DEAD RECKONING

    公开(公告)号:US20240329653A1

    公开(公告)日:2024-10-03

    申请号:US18622287

    申请日:2024-03-29

    申请人: Brain Corporation

    发明人: Micah Richert

    摘要: Systems and methods for robotic control using LiDAR assisted dead reckoning are disclosed herein. According to at least one non-limiting exemplary embodiment, a robot may accurately localize itself over time by detecting parallelized environmental surfaces, such as walls or shelves arranged in a parallel manner, extracting a primary orientation of those surfaces using LiDAR data, and utilizing the primary orientation to accurately define its heading angle in real time, thereby enabling localization via dead reckoning.

    AUTONOMOUS ROBOT AND ITS POSITION CORRECTION METHOD

    公开(公告)号:US20240219919A1

    公开(公告)日:2024-07-04

    申请号:US18403305

    申请日:2024-01-03

    摘要: An autonomous driving robot includes a driving unit that moves the autonomous robot; a camera; a traveling distance measurement sensor; and a control unit that estimates a location of the autonomous robot using a captured image and traveling distance information. In this case, the operation control program generates a robot viewpoint map based on the image captured by the camera, estimates a location of the autonomous robot based on the robot viewpoint map and the measured traveling distance information, and generates a global map based on the robot viewpoint map and position estimation information, and the operation control program inputs the generated robot viewpoint map and global map into a style-transfer model, and inputs a style-transferred robot viewpoint map and a style-transferred global map output by the style-transfer model into the operation agent to correct the estimated position.