Learning assistive teleoperation behaviors from demonstration

Havoutis, Ioannis; Calinon, Sylvain

doi:10.1109/ssrr.2016.7784308

Cited by 12 publications

(14 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently [8] demonstrated how virtual fixtures can be learned from data and how new fixtures can be added to the system so that it can adapt to new manipulation examples. In our recent work in shared control [9], we showed how a teleoperation system can be designed to rely on learned probabilistic models of manipulation tasks in combination with online operator's input.…”

Section: Motivation and Related Workmentioning

confidence: 99%

Supervisory teleoperation with online learning and optimal control

Havoutis

Calinon

2017

2017 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

Abstract-We present a general approach for online learning and optimal control of manipulation tasks in a supervisory teleoperation context, targeted to underwater remotely operated vehicles (ROVs). We use an online Bayesian nonparametric learning algorithm to build models of manipulation motions as task-parametrized hidden semi-Markov models (TP-HSMM) that capture the spatiotemporal characteristics of demonstrated motions in a probabilistic representation. Motions are then executed autonomously using an optimal controller, namely a model predictive control (MPC) approach in a receding horizon fashion. This way the remote system locally closes a high-frequency control loop that robustly handles noise and dynamically changing environments. Our system automates common and recurring tasks, allowing the operator to focus only on the tasks that genuinely require human intervention. We demonstrate how our solution can be used for a hot-stabbing motion in an underwater teleoperation scenario. We evaluate the performance of the system over multiple trials and compare with a state-of-the-art approach. We report that our approach generalizes well with only a few demonstrations, accurately performs the learned task and adapts online to dynamically changing task conditions.

show abstract

Section: Motivation and Related Workmentioning

confidence: 99%

Supervisory teleoperation with online learning and optimal control

Havoutis

Calinon

2017

2017 IEEE International Conference on Robotics and Automation (ICRA)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recently, methods have been proposed to assist humans in co-manipulation and teleoperation tasks given demonstrated trajectories (Raiola et al, 2015 ; Havoutis and Calinon, 2016 , 2017 ). Our work contributes to this field by providing a new reinforcement learning algorithm, P earson-Correlation-Based R elevance Weighted Policy O ptimization (PRO), to improve upon demonstrated trajectories when these are suboptimal or when solutions to new situations must be found.…”

Section: Introductionmentioning

confidence: 99%

Learning Trajectory Distributions for Assisted Teleoperation and Path Planning

Ewerton

Arenz

Maeda³

et al. 2019

Front. Robot. AI

View full text Add to dashboard Cite

Several approaches have been proposed to assist humans in co-manipulation and teleoperation tasks given demonstrated trajectories. However, these approaches are not applicable when the demonstrations are suboptimal or when the generalization capabilities of the learned models cannot cope with the changes in the environment. Nevertheless, in real co-manipulation and teleoperation tasks, the original demonstrations will often be suboptimal and a learning system must be able to cope with new situations. This paper presents a reinforcement learning algorithm that can be applied to such problems. The proposed algorithm is initialized with a probability distribution of demonstrated trajectories and is based on the concept of relevance functions. We show in this paper how the relevance of trajectory parameters to optimization objectives is connected with the concept of Pearson correlation. First, we demonstrate the efficacy of our algorithm by addressing the assisted teleoperation of an object in a static virtual environment. Afterward, we extend this algorithm to deal with dynamic environments by utilizing Gaussian Process regression. The full framework is applied to make a point particle and a 7-DoF robot arm autonomously adapt their movements to changes in the environment as well as to assist the teleoperation of a 7-DoF robot arm in a dynamic environment.

show abstract

“…A probabilistic method was developed to construct a task model in the assistive teleoperated operation. The performance of the teleoperation was improved by the method and demonstrated by remotely operated vehicle (ROV) tasks [26,27]. In order to simplify the complexity of the system and to cope up with the varying dynamical interaction, Huang et al [28] developed a hierarchical interactive learning (HIL) algorithm with dynamic movement primitives (DMPs) and locally weighted regression (LWR) to learn the task trajectories for an exoskeleton system.…”

Section: Introductionmentioning

confidence: 99%

A Robot Learning Method with Physiological Interface for Teleoperation Systems

Luo

Yang

et al. 2019

Applied Sciences

View full text Add to dashboard Cite

The human operator largely relies on the perception of remote environmental conditions to make timely and correct decisions in a prescribed task when the robot is teleoperated in a remote place. However, due to the unknown and dynamic working environments, the manipulator’s performance and efficiency of the human-robot interaction in the tasks may degrade significantly. In this study, a novel method of human-centric interaction, through a physiological interface was presented to capture the information details of the remote operation environments. Simultaneously, in order to relieve workload of the human operator and to improve efficiency of the teleoperation system, an updated regression method was proposed to build up a nonlinear model of demonstration for the prescribed task. Considering that the demonstration data were of various lengths, dynamic time warping algorithm was employed first to synchronize the data over time before proceeding with other steps. The novelty of this method lies in the fact that both the task-specific information and the muscle parameters from the human operator have been taken into account in a single task; therefore, a more natural and safer interaction between the human and the robot could be achieved. The feasibility of the proposed method was demonstrated by experimental results.

show abstract

Learning assistive teleoperation behaviors from demonstration

Cited by 12 publications

References 12 publications

Supervisory teleoperation with online learning and optimal control

Supervisory teleoperation with online learning and optimal control

Learning Trajectory Distributions for Assisted Teleoperation and Path Planning

A Robot Learning Method with Physiological Interface for Teleoperation Systems

Contact Info

Product

Resources

About