Robust Hand

Project Duration:
October 2021 – March 2022

In my semester thesis for the Master’s in Robotics and Control Systems at the Soft Robotics Lab (SRL) of ETH Zurich, I aimed to build a robust gripper. The demand for this research came from the Robotic Systems Lab (RSL) and their research centering around four-legged robot ANYmal. The robot should gain the ability to grasp daily objects and open doors. I recognized that our civilization is designed with the human hand in mind. Therefore, having an artificial hand would be a great cornerstone in solving robotic grasping. While studying many different robotic gripper and hand designs, I realized that currently used grippers have difficulty doing both; they either miss the dexterity or the robustness. With this project, I worked on a finger design accomplishing both. 

Rolling Contact Joint

My design was inspired by the design described in the paper “Fluid Lubricated Dexterous Finger Mechanism for Human-Like Impact Absorbing Capability”. It introduced the rolling contact joint. Such joint consists of two circular rolling surfaces connected by crosswise ligaments. They have very low friction and are super robust to impacts. Instead of breaking, the joint simply dislocates – similarly as biological joints do.  

Finger design

The developed finger design is cheap to produce and can be manufactured on standard FDM 3D printers. Driven by four tendons, it has a similar range of motion as the human hand has. I proofed the capabilities of the design with a two-finger gripper, which is driven by eight Dynamixel servo motors.

High Load Carrying Capability

The two-fingers can easily carry 12.5 kg safely and repeatedly.

Door Opening

I also mounted the gripper on a Franka robot arm, and it could open and close doors similarly as a human would do it. 

Conclusion

Building up on my foundation, the research continued at the Soft Robotics Lab, and it led to the development of the Faive Hand and the publication of the paper “Getting the Ball Rolling: Learning a Dexterous Policy for a Biomimetic Tendon-Driven Hand with Rolling Contact Joints”. It also got covered on Swiss TV SRF1 here.
The project was one of the most challenging I have ever done. But it was also fascinating and kindled my fascination with robotic manipulation and hand design.