Researchers at the University of California San Diego have introduced a new AI powered wearable that enables people to control machines, robots and digital tools using ordinary hand and arm gestures, even during heavy movement. Gesture controlled wearables have traditionally failed in real life scenarios because running, shaking and vehicle vibrations produce motion noise that overwhelms sensor readings.
This new device solves that challenge by combining soft stretchable electronics with an AI model trained specifically to filter out that noise. Tests show that the system continues to identify gestures accurately even in situations involving running, physical impact and simulated ocean waves.
The wearable is built as a soft patch attached to the upper arm. Inside the patch are motion sensors that track direction and acceleration, muscle activity sensors that detect subtle biological signals, a stretchable battery that moves with the body and a Bluetooth enabled microcontroller that sends data instantly to connected devices.
“Wearable technologies with gesture sensors work fine when a user is sitting still, but the signals start to fall apart under excessive motion noise,” said co-first author Xiangjun Chen. “Our system overcomes this limitation.”
“This work establishes a new method for noise tolerance in wearable sensors,” Chen said. “It paves the way for next-generation wearable systems that are not only stretchable and wireless, but also capable of learning from complex environments and individual users.”
These hardware components work alongside a custom deep learning system trained on a large dataset of gestures performed in noisy environments. By learning the difference between intentional gestures and chaotic movement, the AI model is able to isolate meaningful signals and classify them quickly enough for real time control of robotic systems and digital interfaces.
This research represents a major advancement for gesture based interfaces. Until now, most gesture wearables only worked reliably when the user stayed still, limiting their usefulness in everyday life. With this new system, people can control digital tools while walking, running, traveling in vehicles or working in physically demanding settings. The reliability of this technology also creates new opportunities across multiple fields.
Although the prototype is highly promising, further work is needed before it reaches consumer devices. The researchers aim to reduce the size of the electronics, improve battery performance, enhance accuracy across diverse body types and integrate the technology into products such as smartwatches, fitness bands and AR glasses.
Nevertheless, experts note that the combination of stretchable sensing materials with noise resistant AI represents a significant milestone that could influence the entire field of wearable technology.
The findings are detailed in the journal Nature Sensors, which outlines the system design, the training process and the results from real world motion testing.