Completed
Voice-Controlled Home Automation System
Embedded Systems
ESP32-based voice automation with TinyML wake-word detection
C/C++
Python
TensorFlow Lite
ESP32
JavaScript
This project brings voice control to home automation through an embedded system built around the ESP32 microcontroller with on-device wake-word detection using TinyML.
TinyML Wake-Word Detection
- Privacy: Audio processing happens entirely on-device
- Latency: Sub-100ms wake-word detection
- Reliability: Operates independently of internet connectivity
- Accuracy: >90% detection in typical household noise conditions
WiFi-Enabled Device Control
- HTTP REST APIs for device control
- MQTT messaging for event coordination
- WebSocket connections for real-time status updates
Web Interface
Mobile-responsive web interface with real-time device status, manual override controls, and system health monitoring.