AI Based IoT Framework for Soil Analysis and Fertilization Recommendation for Smart Coconut Farming

India is identified as one of the top two farm producers in the world, and the agriculture sector plays a key role in the Indian economy. The major revenue of Kerala state is from coconut farming. Kerala is the top-ranked producer of coconut in India in terms of both area and production. It is observed that coconut farmers are going through a tough situation due to the decline of coconut production over time. They find it difficult to manage the crop efficiently. Proper irrigation and fertilization will help farmers improve the crop yield. In this paper, we have developed analytical models for predicting the fertilizer requirements based on the soil analysis and macronutrients present in the soil. This research work proposes an IoT framework for improving the yield of coconut production by effectively predicting fertilizer recommendations based on the crop’s current fertilization requirements, weather forecasts, and environmental conditions. Based on this analysis an efficient ML model is derived that helps in predicting fertilization requirements based on soil parameters. A model called Linear Regression is built to predict the amount of fertilizer needed using the values of soil parameters such as soil pH, potassium, nitrogen, phosphorus, boron, zinc, soil carbon, manganese, and so on as input features. This model demonstrates promising performance with an accuracy of 93.5% in predicting the fertilizer and can be used by farmers as a tool for scientific farming for enhancing crop production.