IN PROGRESS2025-26

Project Nucleus

Wearable AR/AI System: Haptic Gloves + Custom AR Glasses

A ground-up wearable computing platform built across two physical subsystems. The haptic gloves use a 3D-printed TPU chainmail substrate with ten independently controlled vibrotactile zones, each zone driven by a DRV2605L haptic driver and LRA motor. A V2 layer adds electromagnetic coil actuators beneath each LRA for dual-channel force feedback. The AR glasses are built around a Raspberry Pi 5 with a Hailo-8L neural accelerator (13 TOPS), feeding a custom birdbath optical combiner with OLED display at 1920×1080. The full system runs 6-DOF V-SLAM spatial tracking, AI-accelerated hand recognition, and a local Whisper voice pipeline, all battery-powered from a belt-mounted compute case.

HOLOGRAPHIC RENDER

Interactive 3D visualization — drag to rotate, scroll to zoom

Target Specifications

>10-zone haptic feedback: <10ms BLE command latency
>Dual-layer actuators: LRA vibrotactile + EM coil force feedback
>Birdbath AR optics: 1920×1080 OLED per eye
>13 TOPS edge AI inference on Hailo-8L
>6-DOF V-SLAM spatial tracking
>Local voice pipeline: Whisper STT, no cloud dependency
>3–4 hour wearable runtime from belt-mounted battery

Tech Stack

Raspberry Pi 5 (8GB)Hailo-8L NPU (13 TOPS)ESP32-S3 BLEDRV2605L Haptic DriversTCA9548A I2C MuxBNO085 IMUVL53L5CX ToFTPU 95A Chainmail SubstrateBirdbath Optics (custom build)KiCad 8Fusion 360MediaPipe + Hailo RuntimeOpenCVWhisper (local)Python 3.11

Tags

HardwareCircuitsCodeDesign/CAD

Active Sub-Projects

>Haptic Gloves V1 - 10-zone LRA vibrotactile, ESP32-S3 BLE
-Haptic Gloves V1 - 10-zone LRA vibrotactile, ESP32-S3 BLE
-Haptic Gloves V2 - Dual-layer LRA + EM coil force feedback
-AR Glasses V1 - Birdbath optics, Hailo-8L NPU, V-SLAM
-KiCad Schematics - 7 sheets across gloves + glasses
-Fusion 360 CAD - Glove chassis, lens barrel, frame
-Anamnesis + Tesseract voice integration layer