Hinook —AI-Powered Remote Hardware Lab

Hinook is a remote hands-on electronics lab platform that solves a real problem in hardware education: students can't learn embedded programming without physical hardware, but hardware is expensive, fragile, and inaccessible outside the classroom. With Hinook, each student gets browser-based access to a real physical ESP32 kit — not a simulation. The kit lives in a data center on a Raspberry Pi 5, which runs a full VS Code development environment, a live WebRTC camera stream of the hardware workspace, and a fleet management layer powered by Balena Cloud. Students write real embedded C code in VS Code, flash it to the ESP32 over USB, and watch the hardware respond in real time through the camera. The entire experience happens in a browser tab — no installation, no hardware ownership, no physical presence required. The platform includes 9 progressive labs covering: LED blinking, servo motors, IR sensors, LED strips, light sensors, OLED displays, and Wi-Fi/cloud connectivity. What makes Hinook different is the AI layer. A Gemini-powered agent acts as a live lab instructor. It reads the student's code directly from their VS Code editor buffer — including unsaved changes — and guides them step by step through each lab. When the agent wants to point the student to a specific line, it highlights it directly in their editor using a custom VS Code extension that bridges the AI to the code-server running inside the container. The agent also detects the student's language and responds in either English or Arabic. The infrastructure is built for scale: Balena Cloud handles over-the-air updates to the entire device fleet from a single command, and Cloudflare Zero Trust tunnels expose each kit securely to the internet without port forwarding.