Application of AI in Precision Soil Quality Assessment for Sustainable Agriculture

Soil quality assessment is a critical component of precision agriculture and sustainable land management, as soil physicochemical and fertility-related properties directly influence agricultural productivity and environmental sustainability. Conventional soil evaluation methods are often labor-intensive, time-consuming, and difficult to scale across large agricultural regions. This study aims to develop an accurate, scalable, and interpretable artificial intelligence (AI)–based framework for precision soil quality assessment. A comprehensive secondary soil dataset containing physicochemical and fertility indicators was used to classify soil quality into Good, Medium, and Poor categories. Four AI models—Random Forest, XGBoost, Multilayer Perceptron (MLP), and one-dimensional Convolutional Neural Network (1D-CNN)—were implemented and evaluated using an 80:20 stratified train–test split. Model performance was assessed using accuracy, balanced accuracy, precision, recall, F1-score, confusion matrices, Receiver Operating Characteristic curves, and Precision–Recall curves. Results indicate that deep learning models outperform traditional machine learning approaches, with the MLP achieving the highest accuracy, followed by the 1D-CNN. Explainable AI techniques were applied to identify key soil parameters influencing classification outcomes, enhancing model transparency. The proposed AI-based framework demonstrates strong potential for supporting precision agriculture practices and sustainable soil management.