
Sentinel Turns Old Phones Into AI Cameras
Sentinel is an MIT-licensed, LAN-first monitoring framework that uses the CamFlow Android client to stream JPEG frames from spare phones to a PC dashboard for live viewing, recording, and optional AI event analysis.
This is a compelling hardware-arbitrage project, but it is still closer to an extensible prototype than a production-ready security system.
- –Repurposes old Android phones as inexpensive, distributed camera nodes.
- –Keeps capture and monitoring on the local network, reducing cloud dependence.
- –PC-side Flask/OpenCV processing enables recording, dashboards, and structured AI events.
- –Multimodal inference may still rely on an online model service, so “local” does not guarantee end-to-end privacy.
- –Manual pairing, missing production hardening, and debug-signed Android builds raise deployment friction.
DISCOVERED
39m ago
2026-11-08
PUBLISHED
57m ago
2026-11-08
RELEVANCE
AUTHOR
GithubProjects