Hands for robots.

https://cdn.steemitimages.com/DQmVYShtEGeaRgvWotcrngPtpstYtqjsGYYM5nvyXiKtTfZ/image.png
https://cdn.steemitimages.com/DQmcM9vDyChXWWfaBhnqxXD5scZn7pnHcZrWiwyZ8HMwmbX/separador%201.png

It is not a lack of intelligence that still makes humanoid robberies clumsy, it is a lack of sensitivity, for years we taught machines to copy the movement of human hands, but without allowing them to feel the world they were touching, now a Dutch company has just attacked exactly that blind spot, Manus launched the Metagloves Pro Haptic, professional gloves designed not for games or virtual reality, but for something much more serious, Humanoid teleoperation built-in artificial intelligence, the one that needs to learn with the body, not just with data abstract.

https://www.youtube.com/watch?v=-KiAKI5oOPk
https://cdn.steemitimages.com/DQmcM9vDyChXWWfaBhnqxXD5scZn7pnHcZrWiwyZ8HMwmbX/separador%201.png

The heart of the system is a rare technology outside of industrial environments, electromagnetic field tracking, unlike cameras that lose fingers when there is occlusion or inertial sensors that accumulate error over time, this method creates a local field and measures at the millimeter level each micro variation caused by the movement of the fingers.

The result is a digital twin of the human hand with 25 degrees of freedom. And it's not just the position of the fingers, but the rotation, orientation, and subtle movements of the pasterns—exactly the kind of data that neural networks need to learn fine manipulation in the real world. But capturing motion is not enough, the biggest problem with modern teleoperation is something that researchers call the “sleeping hand.”

The operator sees the robot touch an object, but feels absolutely nothing, without tactile feedback, the human brain applies unreal forces, squeezing too much, passing through virtual objects, creating physically impossible data, that is the bottleneck that Meragloves try to solve, by integrating optical feedback, the gloves give the operator back the sensation of contact, resistance and pressure.

Subtle vibrations inform the exact moment of the touch and help the brain to modulate the applied force, the sensorimotor cycle finally closes, curiously, the launch occurs in the midst of a confusion of names, despite there being some hands linked to the world of AI agents and that was purchased by Meta, the Dutch Manus remains independent focused exclusively on the physical layer of robotics.

And when humans transmit strength, precision and sensitivity to robots, the limit is no longer the software, the next step is for the technology itself to learn to adapt to the world.

https://cdn.steemitimages.com/DQmcM9vDyChXWWfaBhnqxXD5scZn7pnHcZrWiwyZ8HMwmbX/separador%201.png

Sorry for my Ingles, it's not my main language. The images were taken from the sources used or were created with artificial intelligence

Posted Using [INLEO](https://inleo.io/@rebe.torres12/hands-for-robots-7th)