AI Smart Glasses based on BL602

Client

An AR wearables startup pushing for sub-50g smart glasses that provide contextual AI assistance without a tethered smartphone.

Challenge

Integrating a camera, display, audio, and Wi-Fi/BLE connectivity into a glasses frame weighing under 50g, with acceptable thermal distribution on the face.

Product Concept & Architecture

A distributed architecture splits processing: BL602 handles Wi-Fi/BLE connectivity and wake-word detection, while a BES2600 manages camera capture and audio. A single-mode waveguide display (480x320) is embedded in the right lens. The 5MP camera sits in the bridge.

Engineering Solution

The glasses spend 99% of time in deep-sleep wake-word listening mode (<50µA). On "Hey ViewPort" detection, the BES2600 boots and captures a 5MP photo. The BL602 forwards the image to a paired phone for cloud AI processing. Results are displayed on the waveguide as subtle text overlays. Thermal modeling guided component placement: all hot ICs are on the right temple, away from skin contact.