AgriGuardian : AI Based Smart Crop Disease Dectector and Treatment Advisor

Agriguardian is a React Native mobile app that uses an ESP32-CAM module and environmental sensors (temperature, humidity, soil moisture, etc.) to capture real-time plant images and data for rapid crop-disease detection, then delivers tailored treatment recommendations with dosage guidelines; it also analyzes live market trends to suggest optimal selling prices, offers a global chat network for peer-to-peer support and expert advice, and features a location-aware shop finder to help farmers quickly locate nearby agro-suppliers.

Table of Contents:

• AgriGuardian : AI Based Smart Crop Disease Dectector and Treatment Advisor
• Table of Contents:
• Technologies Used
• Hardware Requirements
• Software Requirements
• Workflow
• Setup
  • For Supabase
  • For React Native
  • For Python (AI Processing of Data)
  • For ESP32
  • For ESP32 Cam Module
• Running the App
• Advantages
• PPT

Technologies Used

• React Native
• Expo
• Supabase
• Node.js
• Python
• FastAPI
• ESP32
• ESP32-CAM and Sensors
• Arduino IDE
• Embedded C
• NativeWindCSS
• Langchain
• Gemini
• Open Meteo

Hardware Requirements

• ESP32‑CAM module (with onboard OV2640 camera)
• ESP32 development board (for Servo control library and extra I/O)
• FTDI module
• 128×64 I²C OLED display (SSD1306)
• DHT22 temperature & humidity sensor
• Soil‑moisture sensor (analog output)
• Rain‑detection sensor (analog output)
• Flame‑detection sensor (digital output)
• LDR Sensor
• Two hobby servos (for scanning and tilt: e.g. SG90)
• 5 V water pump
• Relay module (to drive the pump; active‑LOW relay)
• Piezo speaker or buzzer (for alerts/tone output)
• Four momentary push‑buttons (LEFT, RIGHT, UP, DOWN; wired with INPUT_PULLUP)
• I²C wiring (for OLED: SDA, SCL, VCC, GND)
• Jumper wires & breadboard or PCB
• 5 V power supply (capable of driving pump, servos, ESP32, sensors)
• Arduino Cable
• Micro USB Cable
• Power Source

Note : OLED and buttons are optional.

Software Requirements

• Node.js (Latest LTS Version)
• Python (3.11.4)
• Arduino IDE
• Expo GO in Mobile (Expo GO 52)

Workflow

Setup

Clone the repository.

For Supabase

1. Go to Supabase.
2. Create a new project named Agriguardian.
3. Go to Authentication > Policies > Create Policy.
4. Create a new policy with the following configuration:

   Policy Name : Enable All for authenticated users only
   Table : public.Users
   Policy Command : All
   Target Roles: authenticated

For React Native

1. Navigate to Root directory.
2. Open Terminal.
3. Run npm install.
4. This will install the required dependencies.
5. Create a .env file with the following environment variables:

    EXPO_PUBLIC_SUPABASE_URL=
    EXPO_PUBLIC_SUPABASE_API_KEY=
    EXPO_PUBLIC_ESP32_IP=
    EXPO_PUBLIC_ESP32_CAMERA_IP=
    EXPO_PUBLIC_CHARTS=https://api.open-meteo.com/v1/forecast? 
    EXPO_PUBLIC_PYTHON_SERVER=

For Python (AI Processing of Data)

1. Navigate to Models Directory.
2. Create a virtual environment with Python 3.11.4 with name .venv.
3. Activate the virtual environment using command .\venv\Scripts\activate.
4. Run pip install -r requirements.txt to install the required dependencies.
5. Create a .env file with the following environment variables:

   GOOGLE_API_KEY = 

   which you can find from here.

For ESP32

1. Navigate to ESP32 directory.
2. Follow the tutorial to install the ESP32 development board.
3. Upload the code to your ESP32.
4. Connect all the components to the ESP32 in following manners.

Components	VCC	GND	Signal Pin(s)	Notes
OLED 128×64 (SSD1306)	3.3 V (or 5 V)	GND	SDA → GPIO 21 SCL → GPIO 22	I²C bus; reset (–) left unconnected
DHT 22	3.3 V	GND	Data → GPIO 4	10 K Ω pull‑up on data line
Soil Moisture Sensor	3.3 V (or 5 V)	GND	Out → GPIO 34	ADC1_6; use analogRead()
Rain Sensor	3.3 V (or 5 V)	GND	Out → GPIO 35	ADC1_7; use analogRead()
Flame Sensor	3.3 V (or 5 V)	GND	Out → GPIO 15	Digital, active LOW
LDR Sensor	3.3 V (or 5 V)	GND	Out → GPIO 32	ADC1_4; e.g. photoresistor
Water Pump Relay	5 V	GND	IN → GPIO 26	Active‑LOW relay, VCC = 5 V
Servo (Scan)	5 V (or 3.3 V)	GND	PWM → GPIO 25	Use ESP32PWM timer; 500–2400 µs pulse width
Servo (Tilt)	5 V (or 3.3 V)	GND	PWM → GPIO 33	Use separate ESP32PWM timer
Buzzer / Speaker	3.3 V	GND	PWM → GPIO 17	Use tone() for beeps/alerts
Left Button	-	GND	→ GPIO 13 (INPUT_PULLUP)	Wire other leg to GND; reads LOW when pressed
Right Button	-	GND	→ GPIO 12 (INPUT_PULLUP)
Up Button	-	GND	→ GPIO 14 (INPUT_PULLUP)
Down Button	-	GND	→ GPIO 27 (INPUT_PULLUP)

For ESP32 Cam Module

1. Navigate to ESP32_Camera directory.

2. Connect the ESP32-CAM module to the FTDI in following manner.

   ESP32 Cam Module	FTDI
   5 V	VCC
   GND	GND
   UOT	RX
   UOR	TX

3. Connect the IO0 and GND of the ESP32-CAM module. (This is only done while uploading the code)

4. Upload the code to your ESP32-CAM module.

5. Disconnect the IO0 and GND of the ESP32-CAM module.

6. Restart the ESP32-CAM module.

Running the App

1. Connect the ESP32, ESP32-CAM module, Mobile and Laptop (or PC) to the same network (Common WiFi or can be Mobile Hotspot).
2. Open the terminal.
3. Type ipconfig to get the IP address of your network.
4. Copy the IP address in IPv4 Address and paste in the EXPO_PUBLIC_ESP32_CAMERA_IP environment variable and apiEndpoint (line no. 21) of ESP32_CAMERA.ino and upload the code.
5. Connect the ESP32 and ESP32 CAM module to Laptop and check Serial Monitor. Note the IP address on which these devices are running and paste these IPs in EXPO_PUBLIC_ESP32_IP and EXPO_PUBLIC_ESP32_CAMERA_IP respectively.
6. Navigate to Models folder and open terminal and run uvicorn prediction_server:app --host 0.0.0.0 --port 8000 to start the python server. You can also check if the python server is running by going to http://<YOUR_IP_ADDRESS>:8000/ in your browser. If not working, check the IP address again.
7. Now go to root directory and run npx expo start -c to start the app. You will get a QR code there.
8. Open the Expo Go App in your mobile and scan the QR code.
9. Wait for the app to load and bundle everything.
10. You are all set to go!

Advantages

• Upto 94% accuracy in crop disease detection.
• Upto 99% accuracy in weather prediction.
• Use of Hardware for real time data collection.
• Cross Platform Application.
• Can also suggest optimal selling prices.
• Can also be used as a remote monitoring system.

PPT