ESP32 IoT system with a C++ firmware for Zigbee sensors and an AI voice assistant frontend with real-time dashboards.
Robin started as a Zigbee sensor dashboard and evolved into an AI voice assistant. The ESP32 firmware is written in C++ and handles communication with Zigbee-compatible sensors — temperature, humidity, motion, door/window contacts.
The frontend provides real-time dashboards showing sensor readings with historical charts. But the interesting addition is the AI voice assistant layer. You can ask Robin about your home status, set up automation rules, and get alerts through natural language interaction.
The system uses MQTT for sensor data transport and WebSockets for real-time updates to the dashboard. The ESP32 acts as a Zigbee coordinator, processing sensor reports and forwarding them to the backend.
This is where hardware meets software — dealing with the quirks of ESP32 memory constraints, Zigbee protocol timing, and the challenge of making an AI assistant that actually understands home automation context.