Overview

Chapter 1: INTRODUCTION

1.1 Background of the study

In African countries like Rwanda, the need to increase agricultural productivity is becoming increasingly urgent. Farmers, especially small-scale farmers, face the challenge of deciding which crops to cultivate and where to plant them, taking into account the changing environmental conditions and other factors that influence farm productivity. The central issue is the difficulty farmers encounter in making informed decisions about crop selection based on their location, soil characteristics, and environmental factors.

Agriculture plays an important and significant role in the economies of most countries, especially in less developed countries, where the contribution of agricultural activities toward the economy is substantial [1]. Hence, high levels of production in the agricultural sector is of the utmost importance for the economic well-being of most countries and the welfare of people [2] [1].

This is indeed the case for most countries in Africa, including Rwanda and its neighbouring countries. Besides the need to increase productivity, there are also a variety of factors such as climate change, reduction of farmland due to population growth, and globalization [3] that lead to the need for farmers to adapt in various ways, and as of late, by changing the crops they choose to grow. The greatest challenge for most farmers, and in particular subsistence farmers whose contribution to the agricultural sector is a significant one, is being able to decide what to grow and where [4] [5].

In other words, farmers would be more productive if they knew what types of crops or livestock to grow depending on their location and their attributes, as well as other environmental factors that affect farm productivity [6]. This would result in the improvement of the food security of a country and its overall economic well-being. Our main aim is to come up with a smart crop recommender (SCR) system that will help farmers select crops that will produce the greatest yield, based on several factors, in particular soil type and other environmental parameters [7] [9].

This SCR system should be able to recommend a crop based on the attributes of the soil and environment that would make it more accessible and productive to most farmers, backed by ease of access to such data [8] [9] [7].

An inaccurate SCR system would result in great losses for farmers, the agricultural sector, and the economy as a whole, should the SCR system be adopted by a significant number of farmers [10] [12]. Therefore, it is crucial that the SCR system is highly accurate to avoid such losses. With this in mind, we propose a system, an intelligent system that would consider environmental parameters (temperature, rainfall, geographical location in terms of state) and soil characteristics (N, P, K, pH value, soil type and nutrients concentration) before recommending the most suitable crop to the user [11] [12] [10].

2. Problem statement

In Rwanda, agriculture is one of the most important professions. Many of the people do agriculture, but are unable to determine which types of crops are more suitable for their soil. It is important to understand that there are a variety of crops which are only suitable for wet soil, some require medium humidity in the soil to grow…but this knowledge is less known to farmers as well as newbies who develop some interest in farming. As of now, there are very few resources and software which will help in choosing an informed decision with the goal of improving quality. We propose, thus, such a type of software, the Smart Crop Recommender; an IoT and AI-Integrated platform for real-time soil data acquisition and management with crop recommendation capabilities.

3. Objectives

The main objective of this research project is to develop a Smart Crop Recommender system “An IoT and AI-Integrated Platform for Real-Time Soil Data Acquisition, Management With Crop Recommendation Capabilities" addresses a critical need within the agricultural sector.

Specific objectives

  • Review and assessing current state of Smart Recommender System
  • To develop an IoT-based data collection framework
  • Integrate with AI and machine learning algorithms to predict the best crops to grow based on the collected data.
  • Testing of IoT sensor network and launch a pilot program with local farmers to test the crop recommendation capabilities

When designing a dashboard for a smart crop recommender system, you should consider including the following pages and functionalities:

Design

  1. Homepage: The main entrance to the dashboard, providing an overview of the system's purpose and features.

  2. Soil Analysis: A page where users can analyze the soil nutrient levels for efficient farming, as proposed in[1]. This can include visualizations of soil data and recommendations for appropriate crop types based on the soil's nutrient levels.

  3. Crop Recommendation: A page that uses machine learning techniques to predict crop suitability based on factors such as temperature, rainfall, and other environmental data[2]. This can display a list of recommended crops for a specific region or farm, along with their predicted yields and growth rates.

  4. Fertilizer Recommendation: A page that provides optimal fertilizer recommendations for crops based on historical data and expert systems[5]. This can display the suitable quantities of fertilizers for each crop, helping users to make informed decisions about their crop production.

  5. Data Collection and Management: A page that allows users to collect and manage data related to crop production, such as temperature, rainfall, and soil nutrient levels[1]. This can include data visualizations, charts, and tools for users to input and analyze data.

  6. Machine Learning Models: A page that showcases the machine learning models used in the system, such as the KRR model mentioned in[4]. This can include information about the models' performance, accuracy, and how they are applied to the data.

  7. User Profile and Settings: A page where users can customize their dashboard experience, such as setting thresholds for crop production or selecting specific crops for analysis.

  8. Help and Support: A page that provides guidance on using the dashboard, troubleshooting issues, and accessing additional resources or documentation. Remember to design the dashboard with a user-friendly interface, clear navigation, and visually appealing data visualizations to ensure a positive user experience.

Citations: