Researchers at the Massachusetts Institute of Technology (MIT) developed an AI-powered ultrasound wristband that translates internal hand movements into real-time robotic commands [1].
This technology removes the need for external cameras to track hand gestures. By monitoring the body's internal mechanics, the system allows for precise control of machinery in environments where visual tracking is impossible or impractical, such as during surgery or in augmented reality settings [1].
The wristband functions by tracking the movement of muscles, tendons, and ligaments beneath the skin [1]. The system specifically monitors 34 muscles [3], 27 joints [3], and more than 100 tendons and ligaments [3]. This internal data is processed by AI to determine the exact position and intent of the user's hand.
Researchers said the device can track 22 degrees of hand freedom [1]. This level of granularity allows a robotic hand to mirror complex human gestures with high fidelity. The system operates with a latency of 120 ms [1], ensuring that the robotic response occurs almost immediately after the human movement.
The MIT team designed the interface to enable broader human-machine interaction [1]. Because the device is worn on the wrist, it does not obstruct the user's natural range of motion. This makes it a viable alternative to bulky gloves or restrictive sensors that have previously limited the adoption of intuitive robotic control [2].
Potential applications for the ultrasound wristband extend beyond simple robotics. Researchers said the device has utility for virtual and augmented reality, where precise hand tracking is essential for immersion [3]. It may also be utilized in medical fields to allow surgeons to control robotic instruments with higher precision and less invasive hardware [1].
“The system specifically monitors 34 muscles, 27 joints, and more than 100 tendons and ligaments.”
The shift from optical tracking to ultrasound-based internal monitoring represents a significant step toward seamless human-machine interfaces. By bypassing the need for line-of-sight cameras, this technology allows for more reliable control in cluttered or dark environments, potentially accelerating the integration of teleoperated robotics in high-stakes fields like remote surgery and industrial hazardous-material handling.
