| Title |
Multi-sensor Collaborative Remote Control of a Robotic Arm Reflecting User Grasping Intention |
| Authors |
정성훈(Seong-Hoon Jeong) ; 신기원(Gi-Won Shin) ; 오도창(Do-Chang Oh) |
| DOI |
https://doi.org/10.5573/ieie.2026.63.6.131 |
| Keywords |
Electromyography(EMG); ArUco marker; Multi-sensor fusion; Robot operating system; Human-robot interaction(HRI) |
| Abstract |
Conventional user interfaces, such as joysticks, often impose high cognitive loads on operators and lack intuitiveness, leading to operational fatigue and reduced task efficiency. To address these limitations, this paper proposes a novel hybrid teleoperation interface that integrates electromyography (EMG)-based grasp recognition with vision-based automatic target estimation. The proposed system utilizes a wearable EMG armband to intuitively detect the user's grasp intention via Root Mean Square (RMS) signal analysis, which is directly mapped to the robot's gripper control. Concurrently, a monocular camera system employing ArUco markers and precise Charuco board calibration estimates the target object's 6-DoF pose, automatically guiding the robot’s end-effector to the target coordinates through a robust Robot Operating System (ROS)-based coordinate transformation framework. Comparative experiments involving a pick-and-place task demonstrate that the proposed EMG-Vision hybrid system significantly outperforms conventional joystick control, reducing the average task completion time by 44 % and decreasing the task failure rate by over ten-fold (1.7 % vs. 18.3 %). These results validate the system's efficacy in providing a more intuitive and reliable human-robot interaction (HRI) method, suggesting potential applications in rehabilitation robotics, remote handling in hazardous environments, and collaborative robot systems. |