Shadow Robot unveils the world’s most robust dexterous robot hand, developed in partnership with Google DeepMind.

Shadow Robot unveils the world’s most robust dexterous robot hand for AI research, developed in partnership with Google DeepMind.

The new Shadow Hand was built by Shadow Robot and developed with research and insights from the Google DeepMind robotics team.

A key challenge in AI and robotics is to develop hardware that is dexterous enough for complex tasks, but also robust enough for robot learning. Robots learn through trial and error which requires them to safely test things in the real world, sometimes executing motions at the limit of their abilities. This can cause damage to the hardware, and the resulting repairs can be costly and slow down experiments.

Shadow Robot collaborated with Google DeepMind to develop a robotic hand for state-of-the-art learning research. The hand was designed to have fast, flexible and precise finger movements mimicking some properties of the human hand, but also built to be reliable enough for long-running experiments. It is also equipped with advanced tactile sensors to enable the robot to better understand its environment through the sense of touch. The hand was developed in an iterative process during which Shadow Robot repeatedly incorporated feedback from research to improve the hardware.

The new Shadow Hand has precise torque control, achieved with internal 10 kHz force control loops, giving compliance to every joint. It’s also strong and fast, going from fully open to closed in 500 milliseconds, and able to give a 10 N fingertip pinch. Each finger is rich in ground-breaking tactile sensing technology, with dozens of 3-DOF taxels on the middle and proximal phalanges and hundreds on the fingertip, with a massive dynamic range. All sensing data is available in real-time.

The Hand’s technology represents a transformative step for machine learning robots. It provides researchers with a reliable platform, freeing them to focus on the development of artificial intelligence software rather than constant maintenance.

The new Shadow Hand will be demonstrated to the public for the first time at ICRA in May 2024. It is available for purchase as a standard build. Shadow offers a consultancy service, supporting academic and applied research teams to modify the robot or further develop specific components for their project requirements.

Rich Walker, Shadow Robot: “This Hand is an example of how collaboration across cutting-edge research and robotics hardware can create transformative innovation. This is a new way to deliver manipulation ability that will directly contribute to significant advances in how robots learn. I’m excited to show this to the world and ask, ‘what can you do now that you couldn’t do before?’ ”

About Shadow Robot

The Shadow Robot Company has over 20 years of experience in designing and developing robotic systems for research and industry. The company has worked with clients such as NASA, ESA, OpenAI, Google Brain, MIT, the University of Oxford, the University of Cambridge and many others. The company has also won several awards for its innovation and excellence in robotics, such as the Queen’s Award for Enterprise for Innovation in 2019, the AIconics Award for Best Innovation in AI Hardware in 2019 and the E&T Innovation Awards for Excellence in R&D, Intelligent Systems and Tech For Good in 2020.

For press enquiries, contact press@shadowrobot.com

Back
Share:

Press Contact