Shadow Dexterous Hand

ROS Software Maintainer: <>

The Shadow Dexterous Hand, often shortened to Shadow Hand, is part of the Shadow Hand Dexterous Series.

Shadow Dexterous Hand Holding Light Bulb.jpg

The Shadow Dexterous Hand is a truly anthropomorphic approach to robot manipulation. With 20 actuated degrees of freedom, position and force sensors, and ultra-sensitive touch sensors on the fingertips, the Hand provides unique capabilities for problems that require the closest approximation of the human hand. Researchers and developers in industrial labs, research institutes and universities who are looking for the most human-like robotic hands with advanced grasping and manipulating abilities, can use Shadow’s Dexterous Hand to yield data-driven results and make key discoveries in their field.

It has been used for space research by NASA, The European Space Agency (ESA) and Japan Aerospace Exploration Agency (JAXA)

Open AI has used the Hand to further advance their research into artificial intelligence.

The Human Brain Project has used the Hand in simulation.

The Shadow Dexterous Hand uses industry-standard interfaces and can be used as an end-effector such as a teleoperation tool or mounted on a range of robot arms as part of a robot system.

For more information, check out the Dexterous Hand page on our website.


ROS Kinetic and Melodic

For a typical user, we recommend installing Shadow software using Docker. Docker is a container framework where each container image is a lightweight, stand-alone, executable package that includes everything needed to run it. It is similar to a virtual machine but with much less overhead.

Follow the instructions here to get the latest Docker container of the hand driver and interface up and running for both a real Shadow hand and a simulated hand.

Packages Overview

Our code is split into different repositories:

  • sr_common: This repository contains the bare minimum for communicating with the Shadow Hand from a remote computer (urdf models and messages).

  • sr_core: These are the core packages for the Shadow Robot hardware and simulation.

  • sr_interface: This repository contains the high level interface and its dependencies for interacting simply with our robots.

  • sr_tools: This repository contains more advanced tools that might be needed in specific use cases.

  • sr_visualization: This repository contains the various rqt_gui plugins we developed.

  • sr_config: This repository contains the customer specific configuration for the Shadow Robot Hand.

Tutorials and Examples

We are providing a high level wrapper around our robots to make the interfacing easier for non-ROS people. The sr_example package contains examples on how to use it.


Our Tactile Telerobot showreel will give you a good idea around the dexterity of our Hands.

The Human Brain Project uses our Hand in simulation.

OpenAI trained our Hand to manipulate objects and advance research into machine learning and AI:

OpenAI solve a Rubik’s cube with our Robot Hand:

See our Hands and systems in action on our YouTube channel.


Please contact us:

Wiki: Robots/Shadow_Hand (last edited 2020-10-23 11:06:10 by georgiablanco)