KeyboardHand vs Apple Vision Pro.

Apple Vision Pro introduced spatial computing to the masses, but its approach to hand-tracking is optimized for GUI navigation (pinching and dragging), not sustained, high-speed text input. Here is why dedicated hardware beats HMD optics for typing.

The Problem with Head-Mounted Tracking

The Apple Vision Pro (AVP) uses downward-facing RGB cameras and IR illuminators on the headset to track your hands. This is brilliant for passive UI control (looking and pinching). However, for dense text throughput, HMD-based tracking suffers from fundamental occlusion and field-of-view issues.

If you look away from your hands, or place your hands close to your body in a relaxed ergonomic posture, the HMD cameras lose sight of your phalanges. KeyboardHand solves this by separating the tracking hardware from the display hardware. By placing our 850nm IR array on the desk, looking *up* at your hands, we maintain an unbroken 120Hz view of all 10 fingers regardless of where your head is turned.

Metric KeyboardHand Apple Vision Pro OS
Sensor Location Desktop (Static Reference) Head-Mounted (Moving Reference)
Text Input Paradigm Spatial Chording (180+ WPM) Hunt & Peck Virtual QWERTY (~30 WPM)
Occlusion Risk Zero (Upward looking) High (Looking away breaks tracking)
Compute Overhead On-board Cortex-M7 (Zero Host load) Shared with R1/M2 host processors
Host Compatibility Any OS (Standard USB HID) VisionOS Only

Virtual QWERTY is a Dead End

Currently, typing in VisionOS relies on a floating virtual keyboard where users must look at a key and pinch, or physically poke floating keys in the air. This "hunt and peck" paradigm is incredibly slow and fatiguing (often called "gorilla arm" syndrome).

KeyboardHand is explicitly designed for Spatial Computing environments. By plugging KeyboardHand into your Mac (or routing it via Bluetooth for mobile setups), you can write code in VisionOS at 200 WPM while your hands rest comfortably in your lap.