# IoT Based Smart Agriculture System: Architecture & Implementation
The agricultural sector faces immense challenges with water scarcity and unpredictable weather. Building a **Smart Agriculture System** is one of the most impactful final year projects for Electronics (ECE), Electrical (EEE), and Computer Science students.
This guide dives deep into the genuine technical architecture required to build a robust, industry-standard IoT farming system.
## 1. System Architecture
A professional IoT architecture consists of three layers:
1. **Edge Layer (Hardware)**: Microcontrollers and sensors deployed in the soil.
2. **Fog/Network Layer**: The Wi-Fi or LoRaWAN gateway transmitting the data.
3. **Cloud Layer**: The remote server hosting the database and user dashboard.
## 2. Hardware Components Needed
- **Microcontroller**: ESP32 (Preferred over NodeMCU due to dual-core processing and better ADC channels).
- **Sensors**:
- **Capacitive Soil Moisture Sensor (v1.2)**: Crucial! Do not use resistive sensors; they corrode in the soil within days. Capacitive sensors measure dielectric permittivity and last for years.
- **DHT22 Temperature/Humidity Sensor**: More accurate than the standard DHT11.
- **LDR (Light Dependent Resistor)**: To measure sunlight intensity.
- **Actuator**: A 5V DC Submersible Water Pump driven by a 5V Relay Module.
## 3. Wiring Diagram Overview
1. Connect the **VCC** of the Capacitive Soil Sensor to the 3.3V pin of the ESP32.
2. Connect the **Analog Out (AOUT)** of the soil sensor to **GPIO 34** (an ADC pin on the ESP32).
3. Connect the Relay IN pin to **GPIO 5**.
4. *Important Safety Note*: Always power the water pump using a separate external power supply (like a 9V battery or adapter). Powering a motor directly from the ESP32's 5V pin will cause voltage drops and reset the microcontroller!
## 4. Firmware & Logic Complexity
Writing the firmware in C++ for the ESP32 is where many students get stuck. You need to handle:
- **Analog to Digital Conversion (ADC)**: The raw sensor values must be accurately mapped to a percentage (0-100%).
- **Debouncing & Noise Filtering**: Raw sensor readings fluctuate wildly. You must implement software averaging to prevent the water pump from turning on and off erratically.
- **Fail-Safe Logic**: If the sensor disconnects, the system must recognize the fault and immediately shut off the water pump to prevent flooding.
## 5. Cloud Integration
To make this a true IoT project, you must send this data to the cloud.
- You need to manage **WiFi.h** and **HTTPClient.h** libraries without blocking the main loop.
- The ESP32 must send a secure REST API request every 15 minutes with a JSON payload.
- This data is then stored in a database and visualized on a live dashboard.
## Scope for Upgrades (To Impress Examiners)
If you want an "A" grade, add **Machine Learning**. Instead of hardcoding the 30% threshold, send historical weather data and soil moisture data to the cloud. Train an ML model (like Random Forest) to *predict* the optimal watering schedule based on tomorrow's weather forecast!
---
### Need the Complete Project?
Writing the complex C++ firmware, building the cloud infrastructure, and debugging hardware noise can take months. At **Cybotyx Labs**, we provide the complete, ready-to-upload source code, wiring schematics, and the full cloud dashboard for this exact project. Reach out to us for a custom build!