Artificial Intelligence–Assisted Soil Fertility Evaluation and Fertilizer Recommendation Using Laboratory Soil Testing Data

Abstract

Soil fertility assessment is important for sustaining agricultural productivity and environmental health. Conventional soil testing measures parameters such as pH, electrical conductivity, organic carbon, and macronutrients (N, P, K), but interpreting these results often requires expert knowledge. This study presents an Artificial Intelligence (AI)–based decision-support system that combines laboratory soil test data with machine learning algorithms to predict soil fertility and recommend nutrient management practices. Soil samples were analyzed using standard procedures, and the data were used to train Artificial Neural Network and Random Forest models. The system successfully classified soil fertility and identified nutrient deficiencies, particularly low phosphorus and potassium levels, with results comparable to laboratory interpretation. The proposed approach helps farmers understand soil reports, apply balanced fertilizers, and improve crop productivity, demonstrating the potential of AI for precision agriculture and sustainable nutrient management