Sistem Prediksi Iklim Berbasis AI dan IoT untuk Optimalisasi Pertanian

Name of Project

Sistem Prediksi Iklim Berbasis AI dan IoT untuk Optimalisasi Pertanian

Project Domain

This project falls under AI and IoT based agricultural optimization, which focuses on climate and weather prediction by integrating sensors, actuators, cloud and processing units using ESP32 microcontroller.

Meet Our Team

             

Problem Statements             

  • Climate Change.
  • Farmers cannot predict climate change or obtain information related to climate change.

Goals

  • Developing an AI and IoT based climate prediction system to help farmers predict climate and weather conditions accurately.
  • Displays prediction results on the website and on the Lcd.

Solution Statements

  • Using DHT22 sensor to measure temperature and humidity.
  • Using Rain sensor MD0127 to measure rainfall.
  • Using LDR sensor to measure the duration of sunlight.
  • Using Anemometer to measure wind speed.
  • Use Esp32 Devkit V4 as the processing unit, AI integration and send sensor data to the Cloud
  • The climate prediction process will be done on Aws Cloud the results will also be displayed on the website.
  • The weather prediction results and sensor recordings will be displayed on the LCD.

Prerequisites – Component Preparation

Hardware :

  • DHT22 : Measure temperature and humidity.
  • Rain sensor MD0127 : Measure rainfall.
  • LDR sensor : Measure the duration of sunlight.
  • Anemometer : Measure wind speed
  • Esp32 Devkit V4 : Central controller and AI integration.
  • LCD : Display weather prediction results and sensor recordings.
  • 12v Adapter : Power source

Aws Cloud :

  • S3 : Used as a place to store data (csv files), ML models and present prediction results.
  • Lamda : Used to read dependency libraries installed on ec2 via efs.
  • EC2 : Used for library and dependency storage

AI :

  • Long Short Term Memory (LSTM) : Models used for climate prediction.
  • Multilayer Perception : Models used for weather prediction.

Datasheets

Datasheet : using data from databmkgonline with several parameters such as temperature and humidity, rainfall, duration of sunlight and wind speed

Esp Datasheet 

Schematic

Schematic Diagram

 

Demo and Evaluation

  • Setup : 

Preparing the components 

The components needed :

  • Esp32
  • DHT22
  • MD0127 Rain sensor
  • LDR
  • Anemometer
  • LCD + I2C

First test 

 

Combine components with others

 

Making the program code

Trial and Error

Config the Aws Cloud 

The Prototype

  • Demo : 
  • Evaluation :

MLP Model Performance test 

LSTM Model Performance test 

Test Running for 1 hour 

 

Result after 24 hours sensing and prediction

 

 

Conclusion

This project presents a smart and practical solution to address climate-related issues that often change, so that farmers can be aware of climate change and can take action when the climate changes. By using the LSTM model for climate prediction for the next 7 days, MLP for weather prediction for the next 1 day, Esp32 and Aws Cloud as processors, Aws Cloud and LCD as outputs, and several sensors as inputs, this solution can support farmers to be able to identify climate changes that occur in the future. Future improvements can develop prediction targets, improve model and tool performance, and replace power sources

Contact us!

Linkedin :

  • Tamim | www.linkedin.com/in/tamim-habibi-9b6448336
  • Fernanda | www.linkedin.com/in/fernanda-risqy-septian-2b3344370
  • John | https://www.linkedin.com/in/john-nicholas-85bbb5298?utm_source=share&utm_campaign=share_via&utm_content=profile&utm_medium=android_app
  • Usamah | https://www.linkedin.com/in/usamah-miharjo-1bb061343
  • Veda | www.linkedin.com/in/perlitaveda
  • Yasa  | https://www.linkedin.com/in/yasa-harissavanno/?originalSubdomain=id

 

Similar Posts