Meta · Filed Dec 31, 2025 · Published Jul 9, 2026 · verified — real USPTO data

Meta Files Patent for Wearable Tech That Learns to Shorten Typing Sequences

Typing in mid-air is slow and awkward. Meta is working on a wearable that watches you type, notices which motions you struggle with, and then offers you a shorter version of the same sequence.

Meta Patent: Wearable Device That Learns Shortcut Typing — figure from US 2026/0194968 A1
Figure from the official USPTO publication.
See all 14 drawings from this filing ↓
Publication number US 2026/0194968 A1
Applicant Meta Platforms Technologies, LLC
Filing date Dec 31, 2025
Publication date Jul 9, 2026
Inventors Rishi Rajalingham, Wei Lwun Lu, Nikhil Nagraj Rao, Niru Nahesh, David Sussillo
CPC classification 345/156
Grant likelihood Medium
Examiner CENTRAL, DOCKET (Art Unit OPAP)
Status Docketed New Case - Ready for Examination (Feb 20, 2026)
Parent application Claims priority from a provisional application 63741792 (filed 2025-01-03)
Document 20 claims

How Meta's wearable figures out your typing shortcuts

Imagine you're wearing a pair of smart glasses or a wrist sensor and trying to type a word by tapping your fingers in the air. You have to hit several specific finger positions in a specific order, and it feels clunky. Meta's patent is designed to fix exactly that problem.

The wearable's sensors track your finger movements as you attempt a sequence of keystrokes. The system then compares what you actually did against what you were trying to type. If it notices a pattern where you're consistently struggling with a multi-step sequence, it suggests a shorter, simpler version of those same gestures that still produces the same typed result.

Think of it like autocorrect, but instead of fixing your spelling after the fact, it reroutes your finger motions before you make the error. Over time, the device essentially learns a custom shorthand for your hands.

How the system spots and suggests a shorter keystroke path

The patent describes a system built around a wearable device (most likely a wrist band or smart glasses with finger-tracking sensors) that captures fine-grained data about how your fingers move when you attempt to type.

When you try to input something, the device records your keystroke gesture sequence and compares it against the target input you were aiming for. If there's a mismatch or if your sequence is unnecessarily complex, the system identifies a 'devolved' sequence (the patent's term for a shorter, simpler alternative path) that produces the same typed output with fewer finger movements.

The key steps are:

  • Sensors capture your attempted typing motions in real time.
  • The system identifies the gap between what you did and what you intended.
  • It surfaces a simplified gesture sequence that achieves the same result with fewer moves.
  • A representation of that simpler sequence is shown to the user, presumably as a visual or haptic prompt.

The phrase 'devolved sequence' is doing specific work here: it doesn't mean degraded, it means deliberately simplified. The system is essentially building a personalized typing shorthand for each user based on their actual movement patterns.

What this means for typing on AR glasses and wrist devices

Right now, typing on a wearable device, whether that's a gesture-based input on a wrist band or finger-tracking on AR glasses, is one of the biggest friction points stopping these devices from replacing a phone or laptop. If Meta can make that experience feel natural by learning and adapting to your movement style, it removes a major barrier to everyday use.

This also fits directly into Meta's push with its Ray-Ban smart glasses and rumored AR headset work. A system that makes eyes-free, hands-in-air typing practical would matter enormously to anyone who wants to interact with an AR device without pulling out a keyboard.

Editorial take

This is genuinely interesting work because it tackles the right problem: air-typing on wearables is currently frustrating enough that most people give up. The approach of letting the device learn shorter gesture shortcuts specific to each user is a sensible fix. Whether it works well in practice depends entirely on execution, but the concept is sound.

The drawings

14 drawing sheets from US 2026/0194968 A1 · click any drawing to enlarge

Patent filing page

Which company should we read for you?

We track 17 companies here. Pro is the same weekly breakdown for any company you choose, delivered privately. Type a name and we'll scope it and send you a quote.

Get one Big Tech patent every Sunday

Plain English, intelligent commentary, no hype. Free.

Source. Full patent text and figures from the official USPTO publication PDF.

Editorial commentary on a publicly published patent application. Not legal advice.