Psst… We received a tip…

It is 9pm and you have achieved peak couch. Blanket on, snacks within reach, the kind of horizontal that took real commitment to build. The movie is too loud but the remote is on the kitchen bench one full sit-up away.

So you do the modern thing. You say it out loud. "Hey, volume down." Nothing. You say it again, louder, the way you would to someone pretending not to hear you. Now your flatmate is awake, the assistant has cued up a podcast about volcanoes, and you are sitting fully upright in a state of defeat.

There is a version of this where you never said a word. You mutter "volume," barely above a breath, and lower your hand a few centimeters, like turning down an invisible dial. The TV obeys. The band on your wrist felt the nerves in your forearm fire and heard the one word you said, put the two together, and worked out what you meant.

The patent comes from Wearable Devices Ltd, a company out of Yokneam Illit in northern Israel that makes the Mudra line of neural wristbands.

You may know the format already. A band sits on your wrist, reads the tiny electrical signals your nerves send to your fingers, and turns a pinch or a flick into a command, with no controller and or camera pointed at your hands. 

The new idea in this week’s patent is pairing that gesture reading with your voice, and treating the two together as a single instruction.

Picture the couch scene again. You want to change the aircon. You press your index finger to your thumb to wake the system up, which the patent calls an initiating cue. Then you say "set room temp." The word selects the function. Now you lift or drop your hand, and that movement sets the value, one degree per flick of the thumb. A second gesture, fingers tapped together, tells the system you are done.

The patent is careful that these have to land inside the same short window to count, so a stray hand movement and an unrelated sentence do not accidentally combine into a command you never meant.

Underneath, the band is reading surface nerve conductance, the signals travelling down the median, ulnar and radial nerves toward your hand, alongside a motion sensor tracking how your wrist rolls and tilts. The band is listening to the instruction your brain already sent your hand. Sensitive enough, and it can read the intention a beat before the motion is even visible.

The filing also describes verifying that a gesture was performed by one specific person, by comparing the live signal to a stored "gesture signature" and "voice signature," and refusing to act if the two do not match. It describes estimating how much force your fingers are applying. It describes pinch-to-zoom performed in mid-air, with the band inferring where your fingers started and stopped. The through-line is a wristband that reads intent precisely enough to stand in for a touchscreen, a remote, and a password at once.

Two ways of talking to our devices both break in ordinary rooms. Voice assistants need you to speak out loud, which is fine alone and awful everywhere else, such as the quiet carriage, shared office, around a sleeping baby, in the meeting you are pretending to pay attention to… the list goes on. Camera-based gesture control needs your hands in frame and good light, and it still struggles to tell a real command from someone scratching their nose.

Wrist-worn gesture control fixes the camera problem and keeps your hands free, but it carries its own gap. A wristband is sensing your arm all day, and arms move constantly. Working out when a gesture begins, when it ends, and whether you meant anything by it at all is the hard part.

The category we’re looking at is neural input, involving reading the electrical traffic between brain and muscle to control a device, usually from the wrist. It serves the smart-glasses crowd well as they need a way to click that doesn’t involve anti-social movements or commands in public.

The closest and largest name is Meta, whose Neural Band ships bundled with its Ray-Ban Display glasses. It reads surface electromyography, muscle-level signals rather than nerve-level, to drive subtle pinches and swipes. Meta has been at this since it bought the startup CTRL-labs in 2019, and its models were trained on data from roughly 200,000 research participants, which is the part a smaller rival cannot simply copy.

Boston-based Pison built EMG wrist sensing and has since leaned toward cognitive and fitness tracking, licensing its sensor to brands like Timex through a tie-up with STMicroelectronics. Doublepoint, a Helsinki team doing gesture recognition in software on ordinary smartwatches, was bought by the smart-ring maker Oura in early 2026. Ultraleap, the Bristol hand-tracking and haptics firm, was absorbed by the instrument maker ROLI in late 2025. The pattern is hard to miss, that the input layer keeps getting acquired into someone else's product.

Wearable Devices sits among these as a cross-platform play. Its bands work with Apple, Android, Windows and a growing list of glasses makers, including a Rokid pairing due to bundle in 2026, where Meta's band only talks to Meta's glasses. The patent's biometric layer is the genuine differentiator. Meta's band is an input device, whereas this filing describes an identity device, one that can tell you apart from everyone else by the electrical signature of your nerves.

The wristband has no use as a standalone product. 

The most promising carrier is smart glasses. Grand View Research puts the category at around US$2.46 billion in 2025, heading toward US$14.4 billion by 2033. Unit shipments matter more than dollars here, and the Bank of America Institute notes they were expected to clear 10 million in 2025, up from a couple of million the year before. Meta's Ray-Ban line reportedly passed 2 million cumulative sales entering 2026, which means a control problem now exists at scale, that there are millions of people with a display on their face and no obvious way to click.

More broadly, the band is one node in the broader wearable-sensors market, valued near US$1.9 billion in 2024 and forecast toward US$13.2 billion by 2034. The behavior being reshaped is the input layer of computing itself, the thing the mouse did for the desktop and the touchscreen did for the phone.

Here is where leverage gets interesting. Meta's strategy is vertical, that they own the glasses, give the band away, and lock the two together. Wearable Devices' strategy is the opposite, as it works with everybody's hardware. Every glasses maker without its own band is a potential customer. 

Neural input matters. The question is whether the world pays for a separate band on the wrist, or expects it bundled and free the way Meta already trained them to.

The money says neural input is real and that the companies building it rarely stay independent.

Meta bought CTRL-labs in 2019 for a figure Bloomberg reported somewhere between US$500 million and US$1 billion, years before there was a product to attach it to. Meta is extending its band beyond glasses, showing demos at CES 2026 with Garmin for in-car control and research partners for smart-home and accessibility uses. 

Oura acquired Doublepoint in March 2026, folding wrist-gesture software into a smart ring. ROLI acquired Ultraleap in late 2025. Pison has raised about US$25.2 million across its history, including a US$5.35 million round in January 2025 with Samsung Ventures among its backers, and has pointed itself at cognitive-health licensing rather than going head to head with Meta on glasses control.

The sharp risk is the login.

The patent describes refusing to act unless the live gesture matches a stored signature of one specific person's nerve signals and voice. Read as a security feature, that is appealing. Your car will not start for a thief wearing your band, your payment will not clear for someone who stole your band. Read as a data question, it is concerning. A nerve-signal signature is biometric data, and biometric data does not reset. You can change a leaked password in thirty seconds. You cannot reissue your motor neurons.

Imagine a delivery driver is required to wear a band that both controls the warehouse system and confirms their identity all shift. That signature, the electrical fingerprint of their nervous system, now sits in a database somewhere. Who holds it, can an insurer or employer ever ask for it, and what happens when the database leaks, are open questions with unclear answers. 

There is a feasibility risk sitting beneath the privacy one. Reading nerve signals reliably across different people is genuinely hard, which is why Meta needed data from 200,000 participants to get a band that works for a stranger out of the box. A smaller player's system may lean on heavy per-user calibration, and the failure modes cut both ways. Too strict, and the real owner gets locked out of their own front door. Too loose, and the signature stops meaning much at all.

Gesture-plus-voice makes a tidy remote, but the part of this filing with teeth is the claim that a band can confirm it is you by the signature of your nerves and your voice together. Imagine how it could be used for cars, payments, restricted machinery, or your front door.

This week's patent is US 20260023439 A1, titled "Gesture and Voice Controlled Interface Device," published by Wearable Devices Ltd.

Read the filing, hit reply with where you think a nerve-reading band actually belongs, and come find us on Instagram and LinkedIn for the between-issue stuff.