Abstract:Teleoperated manipulation under the human intelligence is an effective solution to confront complicated tasks in unknown environments. However, the uncertain contact in environments is the main challenge to achieve good teleoperation, and some inevitable issues such as nonlinearities, various uncertainties, constraints, and communication delays in local and remote robots should also be taken into account. In this article, a unified contact model is proposed to be the targeted environment interaction for remote robot, which can cover various conditions such as free motion and rigid contact by setting different model parameters. Subsequently, a hybrid motion/force controller is developed to cope with nonlinearities and various uncertainties by adaptive robust technique, thus guaranteeing system stability and good transient convergence to the unified contact model. Particularly, to handle misteleoperation or sudden change of conditions described in unified contact model, a model predictive control-based contact optimization method is developed to be the outer loop of hybrid motion/force controller, which plans the desired motion and force trajectories that meet the state and targeted interaction constraints. By the estimation and transmission of environment parameter, the environment dynamics is reconstructed in the local side, which provides the effective force feedback of remote contact environment for human operator. Since the signal transmitted is replaced by the estimated environment parameter, the power-cycle in communication channel is eliminated. This design can avoid the passivity stability problem caused by communication delays, so the stability and good transparency under delays can be guaranteed by the separate design of local and remote controllers. The comparative experiments are implemented and verify the effectiveness of proposed framework for teleoperated manipulation.

Model Predictive Control for Fluid Human-to-Robot Handovers

A Human-Robot Collaboration System for Object Handover

Human–Robot Shared Control Based on Locally Weighted Intent Prediction for a Teleoperated Hydraulic Manipulator System

Collaborative Behavior Models for Optimized Human-Robot Teamwork

Collaborative Interaction Models for Optimized Human-Robot Teamwork.

Human Grasp Classification for Reactive Human-to-Robot Handovers

Robot-To-Human Handover with Obstacle Avoidance Via Continuous Time Recurrent Neural Network

Human-Robot Co-Transportation with Human Uncertainty-Aware MPC and Pose Optimization

Unified Contact Model and Hybrid Motion/force Control for Teleoperated Manipulation in Unknown Environments

Probabilistic Motion Prediction and Skill Learning for Human-to-Cobot Dual-Arm Handover Control

Cooperative Distributed Model Predictive Control for Robot In-Hand Manipulation

“Follower” to “Collaborator”: A robot proactive collaborative controller based on human multimodal information for 3D handling/assembly scenarios

Learning Human-to-Robot Handovers from Point Clouds.

Contact-Implicit Model Predictive Control for Dexterous In-hand Manipulation: A Long-Horizon and Robust Approach

Model Predictive Control for Human-Centred Lower Limb Robotic Assistance

Robust tube-based MPC with smooth computation for dexterous robot manipulation

Human-to-Robot Handover Based on Reinforcement Learning

Model Predictive Robot-Environment Interaction Control for Mobile Manipulation Tasks

Multi-contact MPC for Dynamic Loco-manipulation on Humanoid Robots

Controlling Object Hand-Over in Human–Robot Collaboration Via Natural Wearable Sensing

Learning Dynamic Robot-to-Human Object Handover from Human Feedback