A New 'Intent Layer' for AI: Inside Wearable Devices' White Paper on Neural Wristbands, LMMs, and the Future of Human-Machine Interaction

On June 23, 2026, Wearable Devices published a white paper positioning its wrist-worn neural sensing technology — centered on the Mudra wristband and the proprietary Large MUAP Model (LMM) — as a foundational 'intent layer' designed to bridge the gap between human intention and the comprehension of AI agents, augmented reality (AR) systems, and robotics .

Core Thesis: The Missing 'Intent Layer' in Computing

The white paper, titled "From Intention to Action: How Brain–Computer Interfaces Are Removing Friction from AI & AR Interaction," argues that as agentic AI, AR, and spatial computing advance, a missing layer exists between human intent and machine action . Wrist-worn neural sensing using surface electromyography (EMG) to decode neural signals before any physical movement occurs can fill that gap, enabling intent-based rather than gesture-based or voice-based interaction .

The company's Founder, President, and Chief Scientist Guy Wagner stated: "Neural input can become a practical control layer for AI and extended reality ('XR') devices, just as controllers and camera-based gestures have shaped today's XR interfaces" .

How the Large MUAP Model (LMM) Works

Analogous to large language models (LLMs) for text, the Large MUAP Model (LMM) is an AI model trained on Motor Unit Action Potential (MUAP) data — the neural signals that govern muscle movement . The LMM translates these neural signals into a 'neural token' representation that AI agents, AR/XR devices, and robotic systems can interpret as a user's intent . The company describes this as "just as LLMs unlocked the power of language for AI, LMMs unlock the power of human intent" .

Unlike camera-based motion tracking that requires line-of-sight and waits for physical action, the Mudra technology decodes neural impulses before movement occurs, offering a faster, discreet, and socially acceptable interaction method .

Hardware Tiers and Product Ecosystem

The white paper and related announcements describe a multi-tier hardware ecosystem:

• Mudra Pro (flagship tier): Launched in May 2026, this is a high-performance neural sensing platform integrating synchronized EMG, optical photoplethysmography (PPG), and inertial measurement unit (IMU) sensors, powered by an ARM Cortex-M33 processor with Bluetooth Low Energy. It is designed for developers, enterprise innovation teams, OEMs, and researchers .
• Mudra Band and Mudra Link (consumer tier): Existing consumer products enabling touch-free gesture control across multiple operating systems, which received a CES 'Best of CES Innovation Award' .
• Mudra Studio: A development environment enabling rapid prototyping, testing, and validation of new interaction models, supporting cross-platform XR and AI experiences .
• Mudra Ultimate: A higher-resolution enterprise tier mentioned for more demanding applications .

A key supporting infrastructure is the ai6 Labs ecosystem (launched February 2026), a closed-loop neural AI ecosystem integrating three pillars: a foundation layer (LMM-based intent decoding), a product layer (Mudra hardware), and an experimentation layer for rapid AI breakthrough testing . The company describes this as building a "Brain-AI Bus — a high-speed neural data highway connecting biological intent to AI" .

Commercial Applications

The white paper identifies five primary commercial targets:

1. Passwordless Authentication and Payment Authorization

A patent allowance issued in April 2026 by the U.S. Patent and Trademark Office covers biometric authentication using biological signals and gestures . The technology would use biometric neural signatures unique to each user as a continuous authentication mechanism, enabling secure payments and user verification without passwords .

2. AI Agent Control

Providing real-time intent signals to agentic AI systems, allowing users to command and redirect AI agents through thought-linked neural commands rather than text or voice . This positions the Mudra technology as the "critical front door to agentic AI," as one analyst described the broader intent-layer concept .

3. AR/XR Interaction

Replacing traditional controllers and camera-based gesture tracking with wrist-worn neural input that works without line-of-sight and prior to physical action . The white paper specifically targets AR glasses and spatial computing devices, arguing that neural input becomes essential as smart glasses with spatial awareness enter the market .

4. Robotics Control

Enabling intuitive, intent-based control of robotic systems, including industrial and service robots, by decoding the user's intended action at the neural level . Wearable Devices also holds a U.S. patent for neural-based measurement of weight, torque, and applied force from the wrist, with potential robotics and sports technology applications .

5. Healthcare and Rehabilitation

A clinical pilot with Soroka University Medical Center is validating Mudra Link as a rehabilitation tool for patients with impaired grip-force control following motor-cortex brain injuries .

Broader Market Context

Wearable Devices operates through a dual-channel model: direct-to-consumer sales of Mudra hardware and enterprise licensing and collaborations . The company positions itself in the rapidly growing XR, AI wearables, and spatial computing markets, arguing that neural input can become as standard a control layer as touchscreens and mice became for earlier computing eras .

It has received a U.S. patent for a "Gesture and Voice-Controlled Interface Device" integrating gesture recognition, voice control, and biometric authentication , and a separate patent for neural-based measurement of physical forces from the wrist .

Important Caveat

All of the above information comes from company press releases, white papers, and similar corporate communications, which contain forward-looking statements subject to uncertainty. The white paper is a positioning document proposing a vision rather than a description of a commercially deployed standard. The Large MUAP Model and intent-layer concept remain in development and have not been independently validated or adopted by third-party AI or AR platforms .