Research Focus
Assistive hand movement control based on low-density sEMG
This research investigates a low-density sEMG-based control framework for personal assistive hand movement, targeting post-stroke or hand-impaired users. Instead of relying on many EMG channels or highly generalized cross-subject models, this work focuses on a practical three-sensor setup placed on selected forearm muscles. The system aims to decode the user’s intended hand or wrist movement from the healthy side and translate it into control signals for a robotic hand, gripper, or future exoskeleton device. The main research challenge is not only gesture classification accuracy, but the reliability of sEMG interpretation under realistic conditions, including inter-subject variability, session-to-session signal changes, electrode placement differences, and real-time constraints. Therefore, the proposed direction emphasizes personal calibration, robust segmentation, compact feature representation, and stable control-oriented recognition. The expected contribution is a practical proof-of-concept architecture showing that minimal sEMG sensing can support personalized assistive control for improving hand function and quality of life.
Grasp assistance device for daily living support
My research focuses on a grasp assistance method using a tendon-driven robotic hand/exoskeleton powered by a single motor. The study aims to support people with reduced grip strength, especially in aging societies, by developing a lightweight, low-cost, and user-friendly system. In this system, a single DYNAMIXEL motor controls multiple fingers cooperatively through a wire-driven mechanism that enables flexion and extension motions. In addition, the system performs grasp control without using external sensors by utilizing motor current values and position information, achieving both structural simplicity and operational safety. The research also investigates switching control modes according to different objects and estimating grasp states for stable grasping performance. In the future, this system is expected to be applied to daily life support and rehabilitation fields.