top of page
Current Research



Designing Versatile Human-Robot Interfaces
[Related Publications: ICRA'20, ICRA'22, IROS'22, IROS'23, THRI'22, RA-L'23, THRI'24]
Remote manipulation and perception present challenges due to the mismatch between human and robot capabilities, as well as unfamiliar viewpoints. To address these limitations, I design systematic solutions that help users adapt to diverse scenarios.
-
Direct Operation (remote manipulation | complex teleoperation)
-
hared Control (shared autonomy | human intent inference)
-
Supervised Robot Autonomy (reliability full autonomy | error recovery)
Modeling User States for Workload Adaptation
[Related Publications: IROS'19, ICRA'20, ICRA'21, THRI'22, THRI'23, RA-L'23, THRI'24]
As operational workload significantly impacts both human-controlled robot performance and the acceptance of robots in everyday life, understanding and monitoring workload—both physical and cognitive efforts—in real-time is crucial for designing workload-adaptive interfaces that enhance control ergonomics.
-
Offline Workload Analysis (muscle effort | physical fatigue | mental workload)
-
Online Workload Estimation (predictive models for physical and cognitive workload)
Developing Effective Training Protocols for End-Users
[Related Publications: IROS'24] [Submitted: RA-L'24, ICRA'25]
Conventional robot control training is largely free-form and self-guided, leading to inconsistencies in skill development among users with diverse abilities and backgrounds. This lack of structured training poses significant challenges in achieving uniform competence levels and accurately calibrating acquired skills. Varying proficiency levels can undermine task performance and hinder productive research on robot control, making it difficult to reliably compare new methods against established baselines.
-
Curriculum-Based (comprehensive training modules)
-
Targeted-Based (identify specific skill gaps)
Completed Research


Wearable Walking Assistive Exoskeleton Robot (2014-2018)
[Won an R&D 100 Award in 2016]
This project began with a question: Can paraplegics stand up and walk again with modern technology? Many rely on wheelchairs for mobility, yet face numerous challenges. Now, imagine if they could stand and walk again—they could accompany their families to department stores and, most importantly, maintain eye-level interactions with others. To make this vision a reality, we undertook the task of developing an exoskeleton robot, funded by the Ministry of Economic Affairs, to fulfill the hopes of physically challenged individuals.
-
Minimizing energy consumption during walking, standing, and sitting transitions
-
Increasing the movement and load-bearing capability of the lower limbs
-
Enhancing independence and social participation
Hip Joint Based Walking Assistive Device (2016)
The mobility challenges faced by elderly individuals are a long-term issue that requires appropriate solutions. We developed a lightweight, hip joint-based walking assistive device with strength control to facilitate easier wearing and optimize energy consumption efficiency. The user-fitting structure is designed to reduce load and friction between the thigh frame and the skin. Motors are installed in each hip joint to assist users in lifting each leg at the thigh while walking and climbing stairs. The adjustable torque of the motors outputs is based on the user's strength contribution and operates in the same direction as the user's intended movement, whether forward or backward.
-
Improvement of gait patterns based on motion analysis
-
Optimization of the device structure for greater lightweight and endurance
-
Involvement in clinical research to gather user feedback
bottom of page