Embedded Engineer - Wearable Smart Glasses + Pocket AI Unit (Prototype) Project Overview We are looking for an experienced embedded systems engineer (or a small team) to design and build a working prototype of a wearable visual reading assistant. The system is composed of: • 1. Smart Glasses (Head Unit): Image capture + bone conduction audio. • 2. Pocket Processing Unit: AI processing, 4G/Bluetooth communication, and power management. The system must be silent, reliable, and discreet, suitable for everyday use. This is a hardware + firmware project requiring industrial-quality execution. Scope of Work A. Head Unit (Smart Glasses) • Base Frame: Carrera 8825 or similar - Matte Black Acetate. Size: 55-16- 145. • Optical Prescription: The lenses must incorporate the prescription details provided below. o Future Replaceability: The design must allow for the prescription lenses to be changed in the future without issues (lenses must not be permanently sealed or glued to the electronics). o Pupillary Distance (PD): 67 mm. o Right Eye: Sph: -2.50, Cyl: -1.00, Axis: $180^{\circ}$. o Left Eye: Sph: -2.50, Cyl: -1.50, Axis: $175^{\circ}$. • Camera: OV5640 (5MP), pinhole lens, mounted in the bridge area. o Apertura $\le1.0$ mm. o Optical insert using black PMMA (IR-pass acceptable). o Matte surface finish (sanded or bead-blasted) to match frame texture and minimize reflections. o All LEDs disabled or removed. • Audio: Stereo bone conduction transducers integrated in temples. • Power: $\sim 350$ mAh LiPo battery with Qi wireless charging (no exposed USB ports). • Controls: Power via Hall sensor + magnet (hinge-based). No accelerometer. B. Pocket Unit • Core: Raspberry Pi Zero 2 W + Integrated 4G/LTE Modem/Module (HAT or USB) with physical SIM slot. • Cellular Connectivity: The 4G/LTE module is mandatory for all Cloud AI connectivity (Gemini, ChatGPT, etc.). The chosen hardware must be fully compatible and tested with Spanish cellular networks. • Enclosure: 3D printed ABS or PETG. o Passive cooling only (no fans). o The design must accommodate the 4G/LTE module and provide an accessible slot/tray for the physical SIM card. • Interface: Two recessed physical buttons (Volume +/-). • Battery Indicators: Incorporate a 3-color LED scheme (Green: High, Yellow: Medium, Red: Low) on the enclosure to indicate the charge status of: o One LED system for the Head Unit (Glasses) battery level. o One LED system for the Pocket Unit (Raspberry Pi) battery level. • Power: 5,000 mAh battery, BMS, USB-C charging. C. Communication & Firmware • Internal Connectivity: Bluetooth Classic only (SPP profile) between Head Unit and Pocket Unit. Wi-Fi disabled on head unit. • Image Preprocessing (Head Unit): Grayscale conversion and JPEG compression. • Target Latency: $<5$ seconds end-to-end. D. Software Architecture • Function: OCR + AI-based text understanding. • AI: Cloud AI supported (API-based, e.g., Gemini / OpenAI / equivalent). No API keys hardcoded. • System must read from SD card: o config.txt → system parameters. o prompt.txt → AI prompt text (user-editable). Deliverables • Fully assembled hardware prototype (glasses + pocket unit). • Firmware for head unit. • Software stack for pocket unit. • SD card configuration format documentation. • Power consumption report (must include 4G/LTE impact). • Assembly and maintenance documentation. Required Skills • Embedded systems (ESP32 / SBC, RPi Zero 2 W). • Integration of 4G/LTE Cellular Modules (with SIM slot). • Design and Implementation of Battery State Indicators (3-color LEDs). • Bluetooth Classic (SPP). • Computer vision & OCR. • Low-power hardware design. • 3D printing & enclosure design. To Apply Please include: • Relevant past projects. • Which parts you would handle personally. • Estimated timeline. • Estimated budget range.