Collaborative Bimanual Manipulation Using Optimal Motion Adaptation and Interaction Control

Abstract

This work developed collaborative bimanual manipulation for reliable and safe human-robot collaboration, which allows remote and local human operators to work interactively for bimanual tasks. We proposed an optimal motion adaptation to retarget arbitrary commands from multiple human operators into feasible control references. The collaborative manipulation framework has three main modules: (1) contact force modulation for compliant physical interactions with objects via admittance control; (2) task-space sequential equilibrium and inverse kinematics optimization, which adapts interactive commands from multiple operators to feasible motions by satisfying the task constraints and physical limits of the robots; and (3) an interaction controller adopted from the fractal impedance control, which is robust to time delay and stable to superimpose multiple control efforts for generating desired joint torques and controlling the dual-arm robots. Extensive experiments demonstrated the capability of the collaborative bimanual framework, including (1) dual-arm teleoperation that adapts arbitrary infeasible commands that violate joint torque limits into continuous operations within safe boundaries, compared to failures without the proposed optimization; (2) robust maneuver of a stack of objects via physical interactions in presence of model inaccuracy;(3) collaborative multi-operator part assembly, and teleoperated industrial connector insertion, which validate the guaranteed stability of reliable human-robot co-manipulation.

Achieving Dexterous Bidirectional Interaction in Uncertain Conditions for Medical Robotics

“…The Motion Adaptation (MA) and Interaction Control (IC) adapts the desired motion to the robot's physical capabilities and the task requirements and, subsequently, generate the torque command for the Replica. The FC is an admittance controller that modifies the trajectory input by the user to account for the interaction at the robots' end-effectors, and it is based on the approach used in [22].…”

“…where ∆ ẋmax is the maximum twist change allowed; thus, it constraints the admittance control to the reachable space around the robot state [22].…”

“…In the eventuality that any of these adverse conditions occurs, SEIKO returns the feasible solution which is closest to the desired state. The problem formulation follows, but the reader should refer to [22] to better understand the algorithm.…”

“…It is executed at the MA level, where the optimisation constraints are added to the conditions required to maintain the grip on the object. These constraints evaluate the contact forces with the object and relative pose of the arms to maintain the grasp, derived by the grasp matrix and a simplified dynamic model of the object (i.e., geometry and mass) [22]. Additionally, a fifth controller is turned on in the IC, which enforces the relative pose between the arms.…”

“…The passivity also allows multiple controllers to be superimposed without affecting their stability, enabling decoupling the control problems and reducing the computational complexity [19]. Earlier teleoperation experiments showcase how the proposed architecture enables the remote operator to collaborate with another person interacting with the replica robots [18], [20]- [22].…”