Viewpoint Selection for the Efficient Teleoperation of a Robot Arm Using Reinforcement Learning
Open Access
- 26 October 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 11, 119647-119658
Abstract
In this study, we developed a novel method to determine the optimal viewpoint from which an operator could realize faster and more accurate robot teleoperation using reinforcement learning. The reinforcement learning model was trained using images obtained from several candidate viewpoints from scratch, and the viewpoint at which the model achieved the highest rewards was considered the optimal viewpoint. The experimental results validated the choice of the world model as the reinforcement learning model. Moreover, subject experiments wherein subjects operated a robot arm remotely to reach the target position were conducted. The experiments produced results that strongly aligned with the performance obtained through computer simulations, indicating that the proposed method is capable of selecting the optimal viewpoint without handcrafted design and subject experiments.Keywords
Funding Information
- Japan Atomic Energy Agency (JAEA) Nuclear Energy Science and Technology (S&T) and Human Resource Development (JPJA19H19210047)
This publication has 13 references indexed in Scilit:
- Free viewpoint image generation system using fisheye cameras and a laser rangefinder for indoor robot teleoperationROBOMECH Journal, 2020
- Automated Image Presentation for Backhoe Embankment Construction in Unmanned Construction Site2020
- Viewpoint selection for grasp detection2017
- The Code for Facial Identity in the Primate BrainCell, 2017
- Technology of Unmanned Construction System in JapanJournal of Robotics and Mechatronics, 2014
- An autonomous multi-camera control system using situation-based role assignment for tele-operated work machines2014
- Emergency response by robots to Fukushima‐Daiichi accident: summary and lessons learnedIndustrial Robot: the international journal of robotics research and application, 2012
- Redesign of rescue mobile robot Quince2011
- Beyond Usability Evaluation: Analysis of Human-Robot Interaction at a Major Robotics CompetitionHuman–Computer Interaction, 2004
- Effects of experience and processing demands on visual information acquisition in driversErgonomics, 1998