CluckTrack: IoT Coop Surveillance
IOT & AI developer
An automated smart chicken coop monitoring system built to track native chicken activity and alert poultry raisers of potential problems. The system uses a budget-friendly ESP32-CAM module mounted inside the coop to snap photos at regular intervals and send them to Firebase storage. A specialized AI model (YOLOv8s) on a laptop analyzes the images fetched from Firebase to keep a count of active chickens. If any unusual inactivity or chicken count is less than specified count, the system updates a mobile app and instantly triggers a cellular module to text an emergency alert directly to the farmer's phone.
- 01Successfully built and trained YOLOv8s model that accurately identified local native chickens with 92.8% precision under normal coop lighting (day light) conditions.
- 02Chose an image snapshot polling method instead of video streaming to save device memory, protect the camera from overheating, and ensure reliable data transmission over weak Wi-Fi.
- 03Integrated a backup cellular text alert layer that operates independently of internet access, ensuring farmers receive immediate emergency notifications even during rural network outages.
- 04Co-authored an IEEE-published research paper documenting the system, the creation of the custom chicken image dataset, and its practical benefits for backyard poultry farming.
System Architecture
The end-to-end data pipeline showing the flow from the edge layer (ESP32-CAM) to Firebase cloud storage. It maps out how the server-side Python environment ingests the images for YOLOv8s processing, and details the hardware-level fallback trigger between the processing laptop and the Arduino/SIM800C GSM module.
Circuit Diagram
A detailed schematic of the coop edge devices. An ESP32-CAM functions as the vision hub. The Arduino Uno orchestrates communication and commands, interface connections with the SIM800C GSM module, establishing a secondary cellular SMS signaling network when standard Wi-Fi is lost.