A Crop Recommendation System Based on Nutrients and Environmental Factors Using Machine Learning Models and IoT

With the ever-increasing population of the world, enough crop production is the biggest concern for the human race. This issue is more pressing than ever as the world population has surpassed the 8 billion mark. Smart farming has become a popular option as it solves the problem by suggesting ways to increase the quality and quantity of crop yield. It is a term associated with the practice of automating farm-related activities. This paper proposes a crop recommendation system based on machine learning algorithms for agricultural fields in India. A sensor system is also prepared to collect first-hand data from fields. These IoT sensors are then used to record levels of soil moisture content, Temperature, and the three most important macro-nutrients required for soil growth: Nitrogen (N), Phosphorus (P), and Potassium (K), from different fields. Additionally, other variables such as rainfall, sowing season, and pH value of soil are also considered to build the proposed crop recommendation system that recommends the best-yielding crop based on the other environmental factors. Multiple machine learning algorithms including Artificial Neural Networks (ANN), Random Forest, Logistic Regression, and K-Nearest Neighbor (KNN) are used and compared to identify the most efficient algorithm for the crop recommendation system. The proposed system aims to develop a model that can help farmers increase their crop yield and quality by providing personalized recommendations based on environmental variables.