1. Why we are interested in robotic hands?

Hands are essential for almost every activity of daily living. For people who have lost hand function due to spinal cord injury, stroke, or other conditions, even simple tasks such as holding a cup become difficult. For people who have lost a hand, a prosthesis is the only way to regain it. A wearable robot that assists grasping, or a prosthetic hand that replaces it, could give them back part of their independence. Robots face the same challenge: to work in human environments, humanoids and robot arms also need hands that can grasp a wide variety of objects.

To be used in daily life, however, these hands should be light, compact, easy to put on, and simple to control. Conventional rigid designs with many motors are often too heavy, bulky, and complex for this purpose.

Our approach is to use soft, tendon-driven mechanisms with under-actuation: a small number of actuators pull tendons routed along the hand, and the routing itself determines how the fingers and thumb move. In this way, we move complexity from the motors and controllers into the mechanical design, so that a simple system can still produce useful grasps.

2. What have we done so far?

2.1. Hand-wearable robots

We have developed a series of soft and hybrid hand-wearable robots, the Exo-Glove family, together with the actuators and sensing methods that drive them.

  • Exo-Glove Pinch – a soft hand-wearable robot designed through the analysis of constrained tendon routing. IEEE RA-L, 2025. Paper, Project
  • Dual-Tendon Routing – a tendon routing method for under-actuated, tendon-driven soft hand-wearable robots. IEEE RA-L, 2025. Paper, Project, Video
  • Exo-Glove Shell – a hybrid rigid-soft wearable robot for thumb opposition with an under-actuated tendon-driven system. Soft Robotics, 2025. Paper, Project, Video
  • Exo-Glove Poly III – grasp assistance by modulating the thumb and finger motion sequence with a single actuator. Soft Robotics, 2025. Paper, Video
  • Slider-Tendon Linear Actuator – a compact actuator with under-actuation and fast connection for soft wearable robots. IEEE/ASME T-MECH, 2021. Paper, Video
  • Joint angle estimation of a tendon-driven soft wearable robot through tension and stroke measurement. Sensors, 2020. Paper
  • GRIPIT – development and assessment of a hand assist device. Journal of NeuroEngineering and Rehabilitation, 2017. Paper

2.2. Prosthetic hands

We are extending our tendon-driven, under-actuated design principles to prosthetic hands. The same goals apply: a light and compact hand that produces useful grasps with a small number of actuators and simple control. Stay tuned for results.

2.3. Robot hands and grippers

The same tendon-driven and under-actuated principles also apply to robot hands and grippers. We are currently working on AI-based design optimization of task-specific robot hands for humanoids and robot arms.

  • Strong compliant grasps using a cable-driven soft gripper. IROS, 2024. Paper
  • Under-actuated robotic gripper with fiber-optic force sensing tendons. IEEE RA-L, 2024. Paper, Video

2.4. What’s next?

We plan to combine our tendon-driven design principles with our computational robot co-design research so that tendon routings and mechanisms can be designed automatically for a given task.

This page was written with the help of large language model (LLM).