Neuroprosthetics Powered by EMG: Restoring Motor Function with Muscle Signals
Senior Hardware Engineer Akshay Yembarwar spoke about how neuroprosthetics are moving from fragile lab systems to practical tools people can use in daily life. He reflected on the engineering hurdles, breakthroughs, and the promise these technologies hold for restoring independence.
For decades, neuroprosthetics seemed like a field trapped in the lab—full of promise, but far from everyday life. Muscle signals were too weak, sensors too rigid, and systems too fragile to survive beyond controlled experiments. Today, however, flexible designs, smarter hardware, and machine learning are helping these technologies break free from research settings. At the center of this shift is a new generation of engineers, including Akshay Yembarwar, whose work bridges the gap between prototypes and real-world tools.
Akshay’s journey began as an intern at CTRL-labs, a pioneering startup later acquired by Meta Reality Labs. “I started by learning how tiny muscle signals could be amplified into something meaningful,” he recalls. Within just a few years, he advanced to a senior hardware engineer role, contributing to muscle-signal platforms that redefined the field.
In 2025, Akshay was part of a global team that published a landmark study in Nature, “A generic non-invasive neuromotor interface for human-computer interaction.” The work demonstrated that a simple wristband could decode human intent without calibration—something experts had long dismissed as impossible. “The true test of such technology,” Akshay explains, “is not how it performs in the lab, but how it adapts to the real lives of people.”
At Meta, he focused on making systems faster, cleaner, and wearable. He designed circuits that cut down noise and power use, while also helping create wristbands with flexible fabrics that allowed natural movement. One of the toughest challenges was electrical interference—everyday environments are full of signals from power lines and wireless devices that drown out muscle activity. By combining smart filters with machine learning, his team found ways to extract clear signals even in noisy conditions.
But Akshay’s contributions weren’t limited to prototypes. He helped scale hardware for studies involving hundreds of participants and released extensive open datasets of muscle activity. These recordings, spanning hundreds of hours and countless tasks, are now used globally to train new models. “Sharing data means the field moves forward together,” he says.
His work highlights a bigger story: transforming delicate lab systems into durable tools for daily life. From motion artifacts caused by wrist movement to differences in skin properties, each challenge demanded painstaking engineering. Yet, with each hurdle cleared, neuroprosthetics moved closer to being truly practical.
The possibilities ahead are striking. Wearable bands could soon help people with spinal cord injuries, ALS, or stroke regain function. Early trials suggest that simply intending to move may allow users to write words at everyday speeds. Looking further, Akshay believes these systems may reshape human-computer interaction, much like the touchscreen did. “Comfort, privacy, and scale are the next big hurdles,” he reflects. “But the direction is clear.”
What was once confined to research labs is now inching toward clinics and homes. And as engineers like Akshay push the boundaries, neuroprosthetics is steadily transforming from an experimental science into a powerful tool for independence and human connection.
