Project Overview
Farmers in Indonesia frequently face harvest failures and volatile market prices, leading to significant financial losses and reduced welfare. This project offers a technological solution by integrating IoT sensors with AI models to analyze soil content and predict optimal crop suitability. By combining real-time environmental data with market price information, the system empowers small-to-medium scale farmers to make informed, data-driven planting decisions. The ultimate goal is to optimize agricultural yields and improve the economic sustainability of farming practices.
System Architecture
ESP32 Microcontroller
Acts as the central hub for developing sensors and managing data transmission from the agricultural field.
Multi-Sensor Array
Includes DHT22, pH, and soil moisture sensors to collect comprehensive environmental and soil data.
MongoDB Atlas
Provides a scalable cloud-based NoSQL database for centralized storage of all captured sensor information.
AI Model (Python)
Analyzes soil content data and market price trends to predict the most suitable crops.
Streamlit Web App
Serves as the interactive user interface to display sensor data and crop recommendations to farmers.
Key Features
Real-Time Soil Analysis
Continuous monitoring of moisture, pH, and temperature through an integrated IoT sensor network.
AI-Powered Predictions
Utilizes machine learning to compare real-time sensor data with ideal crop growth parameters.
Market Price Integration
Real-time food price data is incorporated to ensure recommendations are economically viable for farmers.
Unified Dashboard
A centralized website displays detailed sensor metrics (NPK, Humidity, Temp) alongside specific crop results.
Multi-Device Compatibility
Designed as an improvement over manual methods, allowing integration with various hardware devices.
System Flow
Data Acquisition
IoT sensors deployed in the field collect real-time soil and atmospheric data via ESP32.
Data Classification
The collected raw sensor metrics are processed and categorized for analysis.
AI Comparison
The system compares categorized sensor data against AI models and integrated market price datasets.
Recommendation Generation
Based on the analysis, the system identifies and recommends the best crop for the current conditions.
Web Visualization
Final results and crop suggestions (e.g., "coffee") are displayed on the farmer's web dashboard.
Project Outcome
The project successfully delivered an end-to-end smart farming solution that replaces manual analysis with automated AI insights. Real-world testing demonstrated the system's ability to ingest data to accurately suggest crops like coffee. This tool provides farmers with a measurable unique value proposition through improved decision-making accuracy and integrated market awareness.
Screenshots
