Robots have learned to understand people without cameras or gloves: Russia created a system with "invisible vision"

Robots will no longer need cameras, hand sensors, or special gloves to understand humans. RUT MIREA engineers have developed a technology that allows gesture recognition using an invisible radio beam — even in complete darkness.

The system works like a mini-radar. It scans the surrounding space with an ultra-wideband radio signal, captures the reflection from a person's hand, and uses this data to determine movement. No optical equipment is used in the process.

The development is based on a method of generating radio signals using nonlinear distortions. Thanks to this, engineers were able to expand the signal spectrum by more than 20 times. This is what allows robotic systems to "see" hand movements without cameras.

Such a radio radar is capable of recognizing gestures with 96% accuracy, RUT MIREA added. During tests, the system detected wrist rotation, page scrolling, and fist clenching.

The developers believe that the technology can be used not only in robotics. They plan to apply it in drone control, VR headsets, smart home elements, and other automated systems. The research results are published in the Russian Technological Journal.