EcoFarm is an advanced smart farming ecosystem that combines IoT sensors, machine learning, and cloud computing to optimize agriculture. It provides real-time environmental monitoring and crop predictions, enabling farmers to make data-driven decisions, ensuring maximum yield and sustainability.
✅ Real-Time Environmental Monitoring: Tracks soil nutrients, temperature, humidity, pH, and air quality.
✅ IoT-Driven Data Collection: Uses ESP32 and specialized agricultural sensors.
✅ AI-Powered Crop Prediction: Hoeffding Adaptive Tree Classifier (River package) ensures continuous learning and adaptation.
✅ 24/7 Cloud-Backed System: Backend operates seamlessly to ensure uninterrupted performance.
✅ Mobile Application for Farmers: Flutter-based app providing real-time monitoring & action control.
✅ Incremental Learning Model: Updates continuously without requiring manual retraining.
✅ LLM-Based Chat Assistant: AI-powered chatbot for agricultural guidance & problem-solving.
✅ Proven & Tested PCB: Successfully validated through 24-hour testing with 12,000+ real-world observations.
✅ Azure-Based ML Deployment: Ensuring seamless AI-driven decision-making.
📌 DHT22 - Monitors temperature & humidity.
📌 NPK Soil Sensor - Analyzes essential soil nutrients (Nitrogen, Phosphorus, Potassium).
📌 Analog pH Sensor - Measures soil acidity levels.
📌 MQ-135 Gas Sensor - Detects air pollutants affecting crop health.
📌 ESP32 Devkit-V1 - Microcontroller for IoT integration & data transmission.
🔹 Backend: Flask-based with FastAPI, integrating MQTT & REST APIs.
🔹 Database: Supabase (PostgreSQL) for seamless data storage & retrieval.
🔹 Frontend: Flutter-based mobile app for data visualization & control.
🔹 Cloud Services: Azure VM powering the machine learning & analytics.
| 🔹 Component | 🔹 Technology Used |
|---|---|
| IoT Devices | ESP32, DHT22, MQ-135, NPK Sensor, pH Sensor V2 |
| Backend | Flask, MQTT (HiveMQ), FastAPI |
| Frontend | Flutter via FlutterFlow |
| Database | Supabase (PostgreSQL) |
| ML Models | Hoeffding Adaptive Tree Classifier (River package) |
| Deployment | Azure VM, Docker, MLflow |
- Sources: IoT sensors, historical climate data.
- Features: N, P, K, temperature, humidity, pH, rainfall.
- EDA Actions: Feature engineering, correlation analysis, and pattern discovery.
- Hoeffding Adaptive Tree Classifier: Built for real-time, incremental learning.
- Evaluation Metrics: Accuracy, Precision, Recall, F1-score.
- MLflow for experiment tracking & versioning.
- Supabase integration for real-time data updates.
- Live analytics via Flutter App.
| Name | Role |
|---|---|
| Mahmoud Essam | Leader - Machine Learning Engineer - IoT Dev - Flutter Dev - Cloud Architect |
| Marwan Ali | IoT Developer - Cloud Architect |
| Muhammed Hamdi | Flutter Developer |
| Abdullah Ibrahim | Cloud Architect |
| Abdelrhman Ragab | Data Scientist |
| Ziad Ashraf | Data Scientist |
| Safy Fathy | Flutter Developer |
| Muhammed Hassan | Cloud Architect |
📩 Connect with Ziad Ashraf on 
🚀 Transforming agriculture, one smart farm at a time! 🌱