Wearable Devices positions wrist sensing as AI intent layer
Wearable Devices Ltd. published a white paper titled 'From Intention to Action', positioning its Mudra neural interface as the foundational 'intent layer' for agentic AI, augmented reality, and robotics. The paper introduces the proprietary Large MUAP Model (LMM) to translate wrist-based nerve signals into machine-readable 'neural tokens', aiming to reduce the alignment gap between human intent and AI understanding. Key applications include password-less identity verification and training robotic hands using human muscle force.

*this image is generated using AI for illustrative purposes only.
Wearable Devices Ltd. announced the release of a white paper positioning its Mudra neural interface as the foundational 'intent layer' for agentic AI, augmented reality (AR), and robotics. The document, titled 'From Intention to Action: How Brain–Computer Interfaces Are Removing Friction from AI & AR Interaction', argues that wrist-worn neural sensing can close the 'intent gap' between human intent and AI agent understanding. This development aims to enable more natural interaction with AI agents and AR systems across consumer, enterprise, and industrial markets.
The white paper introduces the Company’s proprietary Large MUAP Model (LMM), a data-compounding 'neural-token' representation designed as a differentiated data asset. It outlines near-term, high-value applications, including password-less identity verification, payment authentication, and the training of robotic hands on real human muscle force and intent. The paper establishes a multi-sensor framework for intent-aware agentic solutions, describing how the LMM translates wrist-based nerve and muscle signals into machine-readable patterns to reduce per-user setup and improve prediction accuracy.
Key Themes and Commercial Use Cases
The paper defines intention within a practical framework connecting intention, agency, and human–agent collaboration. It highlights 'Passive Identity', a neuromuscular-signature-based approach to continuous authentication and transaction authorization. Additionally, it explores how wrist-worn sensing may address a major robotics bottleneck by allowing robotic hands to be trained on actual human muscle force and anticipatory tension rather than just visual paths.
Mudra Platform Configurations
The white paper details two commercial configurations of the Mudra platform designed to serve as the physical layer for this intentional interface:
| Configuration | Specifications | Target Application |
|---|---|---|
| Mudra Pro | 3 EMG channels, IMU, PPG | Context-aware consumer interaction |
| Mudra Ultimate | 8 EMG channels | Enterprise applications with high spatial resolution to power the LMM |
Physiological Context and AR Feedback
Beyond action, the paper explores how movement, muscle tone, and heart-related signals may reveal attention, engagement, and internal state. This physiological context helps systems understand not only what a user is trying to do but how the user is responding. The document presents AR as a feedback channel where visual, spatial, and haptic cues allow users and AI to adjust to each other continuously.
Guy Wagner, Founder and Chief Scientific Officer at Wearable Devices, stated that the interface remains the bottleneck as AI and AR become more capable. He emphasized that combining wrist-worn neural sensing with physiological context and real-time AR feedback may help systems understand users faster and act with less friction. Wagner noted that leveraging proprietary neural tokens could unlock zero-friction applications across security, robotics, and digital health.
How will the introduction of the Large MUAP Model (LMM) impact the competitive landscape for neural interface data assets?
What are the potential regulatory hurdles for implementing 'Passive Identity' as a primary method for financial transaction authorization?
Could the integration of Mudra's neural sensing with robotics accelerate the adoption of human-centric automation in industrial manufacturing?


























