The teleoperation of robots with human-like capabilities may pose significant challenges to the human operator due to the kinematic complexity and redundancy of these robots. Bimanual telemanipulation represents such a challenging task that requires precise coordination of the two arms to perform a stable bimanual grasp on an object and eventually transport the object while maintaining the grasp.In this work, we present a shared control telemanipulation interface to facilitate the bimanual grasping and transportation of objects of unknown mass. With the proposed method, the robot is able to transport the object maintaining autonomously a sufficient amount of grasping force while accepting commands from the operator to reach the desired location. As humans do, it is not necessary to know the weight of the object in advance; instead, the robot estimates it during the lifting phase. On the basis of the estimated weight, the required amount of grasping force is computed. During object transportation, the robot autonomously regulates the grasping forces in a shared control fashion, allowing the operator to seamlessly command only the trajectories of the object.The proposed method has been implemented and validated on the CENTAURO robot, a quadrupedal platform with a humanoid dual arm upper body, performing experiment where objects of different weights and dimensions must be picked up and transported.
A Shared Telemanipulation Interface to Facilitate Bimanual Grasping and Transportation of Objects of Unknown Mass
Torielli, D;De Luca, A;
2022-01-01
Abstract
The teleoperation of robots with human-like capabilities may pose significant challenges to the human operator due to the kinematic complexity and redundancy of these robots. Bimanual telemanipulation represents such a challenging task that requires precise coordination of the two arms to perform a stable bimanual grasp on an object and eventually transport the object while maintaining the grasp.In this work, we present a shared control telemanipulation interface to facilitate the bimanual grasping and transportation of objects of unknown mass. With the proposed method, the robot is able to transport the object maintaining autonomously a sufficient amount of grasping force while accepting commands from the operator to reach the desired location. As humans do, it is not necessary to know the weight of the object in advance; instead, the robot estimates it during the lifting phase. On the basis of the estimated weight, the required amount of grasping force is computed. During object transportation, the robot autonomously regulates the grasping forces in a shared control fashion, allowing the operator to seamlessly command only the trajectories of the object.The proposed method has been implemented and validated on the CENTAURO robot, a quadrupedal platform with a humanoid dual arm upper body, performing experiment where objects of different weights and dimensions must be picked up and transported.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.