Sign up to our newsletter here

DEX-EE Overview: The most robust dexterous robot hand on the market

Shadow Robot unveils the DEX-EE, a transformative robot for use in dexterous manipulation and machine learning research, designed in collaboration with Google DeepMind.  

Shadow Hand holding a hammer

Developed with the world’s leading AI researchers

In a groundbreaking project, Shadow Robot’s latest project delivers a significant advancement in robot dexterous manipulation through collaboration with Google DeepMind. 

The new dexterous robot hand is designed to address the specific challenges faced in rigorous and physically demanding machine learning research. 

Google DeepMind approached Shadow Robot with a specific set of problems. The Shadow Hand needed to offer their researchers next-level reliability for use in long-running experiments where it would suffer repeated impacts, wrenching of the fingers and various forceful movements in its quest for learning. All whilst being capable of delicate, precise movement and offering the most advanced tactile sensing possible.

 

Iteration at the cutting edge

Through five years of iterative development, Shadow worked with Google DeepMind, evolving the technology to meet their needs

at the cutting edge of the rapidly evolving machine learning space.

Now available to the market as a standard build, DEX-EE combines long mean time to failure and reduced time to repair with a suite of advanced sensors to deliver real-time data with precise force control and high availability. Features such as fail-safes and a graceful shutdown routine have been implemented to further enhance Shadow Hand’s robustness, and integration with ROS makes it a versatile tool for research. 

 

Breaking down The Shadow Hand

 

Every component has been designed from scratch and rigorously tested for optimal performance.

 

  • High bandwidth torque and position control loops give delicate and precise fingertip dexterity
  • Torque and inertial measurement throughout makes the whole hand sensitive to interactions with its environment
  • Stereo camera-based fingertip tactile sensors provide an unprecedented level of 3D interaction detail in a robust package
  • Multi-taxel, 3 DOF tactile sensors on middle and proximal phalanges give additional information during grasping and manipulation
  • Easy for users to maintain with minimal training and designed to reduce instances of failure and downtime, with fail-safes and a graceful shutdown routine

 

Collaborative Consultancy

 

Whether the technology fits your requirements as a standard build, or you’re looking to adapt components or knowhow to meet your needs, Shadow works with DEX-EE customers long term to facilitate collaborative development and provide support and engineering services. We have vast experience of working with both academic and applied research teams, and welcome the opportunity to discuss your project specifications. 

 

Contact us to discuss how DEX-EE could accelerate your machine learning research.

Back
Share: