ORIGINAL RESEARCH article
Front. Soil Sci.
Sec. Soil Management
Volume 5 - 2025 | doi: 10.3389/fsoil.2025.1652058
Real-Time Soil Fertility Analysis, Crop Prediction, and Insights Using Machine Learning and Deep Learning Algorithms
Provisionally accepted
- 1Vellore Institute of Technology, Chennai, Chennai, India
- 2Botswana International University of Science and Technology, Palapye, Botswana
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Abstract - Sustainable agricultural management depends on the ability to predict soil fertility, and conventional techniques for doing so are labour-intensive and dangerous. However, utilizing soil metrics, meteorological data, and other pertinent variables, it is now feasible to reliably estimate soil fertility thanks to machine learning (ML) and artificial intelligence (AI) approaches. This paper presents a study of ML algorithms for predicting soil fertility. The findings emphasize the significance of using a thorough method to predict soil fertility that considers variables like pH, temperature, humidity, moisture content, NPK (nitrogen, phosphorus, and potassium), carbon content, organic matter, weather, and climatic conditions. The suggested approach provides a rapid and accurate result, empowering farmers to maximize soil fertility and make well-informed decisions. A hardware prototype was developed with the help of appropriate sensors and a micro-controller to detect the soil parameters for the soil samples and real time value from the kit is compared with ML prediction algorithms. The results obtained from the kit were tested with the lab results to check the reliability of the kit. Overall, the outcome demonstrates the significant potential of algorithms of ML predict soil fertility and have practical agricultural management implications. Ensemble models showed Random Forest achieving about 92% accuracy, followed by Extra Trees and other classifiers with decent scores. The crop prediction is done using deep learning algorithms such as multi-layer perceptron and Long Short-Term Memory (LSTM) and the Accuracy, F1-Score, and Recall give high results while maintaining precision constant.
Keywords: soil fertility, machine learning, random forest, decision tree, Support vector machine, Soil nutrients
Received: 23 Jun 2025; Accepted: 11 Sep 2025.
Copyright: © 2025 Gunasekaran, A, Sreevardhan and SamiKannu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Karmel A, karmel.a@vit.ac.in
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.