Novel Mapping Algorithm For Bilateral Teleoperation Of Bipedal Robot Using A Manipulator

Abstract

Research on the Bilateral teleoperation of biped robots is limited. So far, researchers have only studied bilateral teleoperation of bipedal robots using exoskeletons due to the kinematic similarities. This is, however, not practical as it requires considerable resources for the purchase, maintenance, and operator safety training. This thesis presents a novel mapping algorithm for bilateral teleoperation of a bipedal robot using a robotic manipulator. This research aims to control the gait of a biped robot and receive haptic feedback representative of its movements along its center of mass (CM). Two mapping methods are presented and compared: 1- a cartesian-based mapping that relies on forward and inverse kinematics to map desired trajectories; 2- a deep-learning approach that reduces complexity and computation time. Simulation and experimental results on a 10-degrees of freedom (DoF) biped robot and 3-DoF robotic manipulator are carried out to validate the feasibility of this proposal.