Abstract: In India, approximately 70% of the population depends on agriculture. However, crop productivity has not kept up with the increasing demand, necessitating the use of advanced technology. This project focuses on smart farming and precision farming, utilizing IoT, sensors, controllers, and AI/ML to predict suitable crops based on soil and environmental parameters.

B. Objectives of the Project

  • Study different soil types and crops suitable for each type.
  • Identify and select sensors for soil and environment parameters.
  • Create classification and prediction models for soil and crops.
  • Train, test, and deploy the system on a suitable cloud platform.
  • Build an Android app for farmers to interact with the system.

Project Architecture: Architecture Design

C. Scope & Limitations

Scope:

  • Decision-making support for farmers.
  • Soil health awareness.
  • Economic benefits in the agriculture sector.

Limitations:

  • Dynamic nature of soil.
  • Changing weather conditions.
  • Uncertainty in environmental factors.

GitHub Repository: IOT-based-Soil-Classification-and-Crop-Prediction

2.1 Technological Used

  • AWS IoT Analytics
  • Amplify Framework (Serverless Application)
  • Arduino IDE
  • GitHub
  • Git
  • AWS IoT Core
  • Amazon SageMaker
  • AWS API Gateway
  • Amazon Elastic Compute Cloud (EC2)
  • Jupyter Notebook
  • Android Studio
  • NodeMCU.

8. Performance Analysis

  • The project’s performance is based on the classification and prediction model accuracy.
  • Integration of IoT devices with Android Application and AWS ensures accurate sensor readings.
  • Soil classification and crop prediction models provide accuracy between 65% to 85%.
  • Predicted crop list is displayed on the Android application.

9. Application

  • Increase crop production yield rate.
  • Maximize profit for farmers.
  • Ensure crop availability for a healthy society.
  • Reduce uncertainty in farming.
  • Analyze resources and infrastructure for a specific time span.
  • Reduce excessive chemical usage with precision agriculture.

10. Conclusion

  • The developed system successfully predicts crops based on soil and environmental parameters.
  • Android application integration with IoT sensors and Machine Learning deployed models was achieved.

11. Screenshots

Login Page

  • Login Page

OTP Received

  • OTP Received

Connect to IOT Kit

  • Connect to IOT Kit

Get Longitude and Latitude

  • Get Longitude and Latitude

Soil Classification Page

  • Soil Classification Page

Crop Prediction Results through Postman

  • Crop Prediction Results through Postman
  • Crop Prediction Results through Postman

12 Literature Review

  • F. Tseng et al. [4]: Big data analysis of farms based on Intelligent Agriculture.
  • S. Liu et al. [5]: IoT and cloud-based monitoring systems for agriculture management.
  • R. Priya et al. [6]: Efficient crop recommendation system.