A Construction Method of Remote Operating System for Position and Pose Mapping of Master-Slave Heterogeneous Robots

Robots exhibit significant advantages over humans in complex environments due to their high adaptability. However, the development of fully autonomous robots remains an ongoing challenge. This study presents a remote operation system for master-slave heterogeneous robots, addressing potential mismatches in degrees of freedom and workspaces between the master hand controller and the slave robot. A novel motion control strategy based on virtual robotic arm position and posture mapping was designed, enabling precise synchronization between the two through a control chain linking the force hand controller, virtual robotic arm, and physical robotic arm. The system architecture was analyzed, and kinematic models for both master and slave robots were constructed to obtain accurate position data. To enhance the perception accuracy and efficiency of the slave robot's environment, a semi-structured workspace construction strategy was employed. This strategy integrates predefined models with real-time environment point cloud data captured by an RGB-D camera, followed by processes such as filtering, texture mapping, registration and segmentation to accurately reconstruct a virtual model of the unstructured environment. Additionally, an innovative “pseudo-contact” error correction mechanism was introduced to optimize the human-robot interaction process. This research demonstrates a significant step towards improving remote robotic operation in complex settings.