Skip to content

RAPID Hand

A low-cost, modular dexterous robotic hand platform with hardware design and vision-guided teleoperation software for manipulation research.

  • Other
  • ROS
  • 3D model (URDF)
  • 3D printable

Suggested by Robotheca, created by @SYSU-RoboticsLab (opens in a new tab), from github.com/SYSU-RoboticsLab/RAPID-Hand (opens in a new tab). Pinned to commit 1767d5b (opens in a new tab).

Hardware license
CC-BY-NC-SA-4.0 (opens in a new tab)
Software license
CC-BY-NC-SA-4.0 (opens in a new tab)
Commercial use
Not allowed

A non-commercial license (CC-BY-NC) does not allow selling kits or parts.

This repository contains executable files that Robotheca doesn't check. Only run them if you trust the creator.

Which files (2 scripts)

Photo and 3D model

Its license doesn't let Robotheca show its photos.
See the photos on GitHub (opens in a new tab)
Try it in the browser

3D model loaded from the original files on GitHub at 1767d5b, under CC-BY-NC-SA-4.0. View the original files from GitHub (16.3 MB) (opens in a new tab)

What it can do

Written by Robotheca from the project's documents.

  • Dexterous robotic hand with four fingers and a thumb, each joint driven by a Dynamixel motor.
  • Predefined gesture sequences such as finger curling, run from a motion sequence controller script.
  • Vision-guided teleoperation of the hand using motion data streamed from an Apple Vision Pro headset.
  • Optional real-time simulation of the teleoperated motion in the Sapien simulator.
  • Optional control of a UR10 arm together with the hand, with joint and sensor data published on ROS topics.

Bill of materials

This repository has no parts list Robotheca can read yet.

Build guide

The guide appears here as you scroll.

Guide by @SYSU-RoboticsLab (README.md). Its license (CC-BY-NC-SA-4.0) doesn't let everyone reuse it. To reuse it, ask the creator on GitHub (opens in a new tab). Creators can ask Robotheca to remove it at contact@robotheca.org.

More guides

Wiring and electronics

Notes written by Robotheca from the project's documents.

Main parts

  • Dynamixel servo motors, 20 in total, five per finger and thumb, controlled over a serial bus.
  • Tactile sensors, one per finger and one for the thumb, read through a PCB sync module.
  • Orbbec Gemini 335 depth camera, connected by USB 3.0.
  • Vision Pro headset streaming hand tracking data over the local network.
  • Host computer running ROS and Python.

Connections

  • Dynamixel motors connect to the host computer through a serial port such as /dev/ttyUSB0, configured in rapid_args.yaml.
  • Tactile PCB appears as a serial port such as /dev/ttyCH341USB0, also set in rapid_args.yaml.
  • The tactile signal order from each finger is mapped in the configuration to match the sensor sequence.
  • The Orbbec camera connects to the host over USB 3.0 for reliable data transfer.
  • The Vision Pro and host computer must share the same local network, with the headset IP set in visionpro.yaml.