Sudha et al. (2026) A review on machine learning-based precision agriculture techniques for crop farming monitoring with IOT

Identification

• Journal: Discover Environment
• Year: 2026
• Date: 2026-01-08
• Authors: S. P. Sudha, J. B. Shajilin Loret
• DOI: 10.1007/s44274-025-00305-8

Research Groups

• Department of Computer Science and Engineering, St.Joseph’s College of Engineering, Chennai, Tamilnadu, India
• Department of Information Technology, Francis Xavier Engineering College, Tirunelveli, Tamilnadu, India

Short Summary

This review synthesizes the integration of machine learning (ML) and Internet of Things (IoT) in precision agriculture for enhanced crop monitoring, resource management, and yield prediction, while also identifying key challenges and future research directions for robust, scalable, and secure smart farming solutions.

Objective

• To review and synthesize the integration of machine learning (ML) techniques with Internet of Things (IoT)-based precision agriculture systems to enhance crop health monitoring, soil analysis, irrigation management, and yield prediction, while also identifying challenges and future research directions for robust, scalable, and secure smart farming.

Study Configuration

• Spatial Scale: Review covering applications from individual plants/fields (e.g., pest detection, soil moisture) to broader farm/regional scales (e.g., yield prediction, climate impact assessment) using in-field sensors, drones, and satellites.
• Temporal Scale: Review covering real-time monitoring (e.g., sensor data, immediate interventions), short-term predictions (e.g., yield forecasts, disease detection), and long-term sustainability strategies.

Methodology and Data

• Models used: Support Vector Machines (SVM), Random Forests, K-Means, DBSCAN, Deep Learning (DL), Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), XGBoost, Gradient Boosting, k-Nearest Neighbors (k-NN), Naïve Bayes, Multiple Linear Regression (MLR), Reinforcement Learning, Fuzzy Logic Control, Chaotic Jaya Optimization Algorithm (CJOA), Generative Adversarial Networks (GANs).
• Data sources: IoT sensors (temperature, soil moisture, humidity, pH, nutrient), drones (UAVs), satellites (satellite imagery, remote sensing, hyperspectral images, multispectral images), ground-based cameras, smartphone cameras, historical yield records, climate records, weather stations, market analytics, genetic information, field data (soil, crop, weather), time-series data, spatial data.

Main Results

• The integration of ML and IoT significantly enhances precision agriculture across crop health monitoring, soil analysis, irrigation management, and yield prediction.
• Various ML algorithms (e.g., CNN, LSTM, XGBoost, Random Forest) demonstrate high accuracy in tasks like pest/disease detection, yield forecasting, and resource optimization, with performance varying based on data complexity and environmental context.
• IoT sensors provide real-time environmental and crop data, enabling data-driven decision-making and automated farm operations.
• Key challenges include data heterogeneity, sensor reliability, computational complexity, high implementation costs, lack of technical expertise, cybersecurity threats, interoperability issues, and class imbalance in datasets.
• Edge computing and cloud architectures optimize data processing and decision support, reducing latency for real-time interventions.
• Blockchain-based security models, lightweight cryptography, and federated learning are crucial for enhancing data integrity, privacy, and secure data sharing in IoT-enabled agriculture.

Contributions

• Provides a holistic and multidisciplinary review integrating conventional ML and IoT frameworks with recent advancements in edge computing, blockchain, AI-based image processing, and predictive analytics.
• Uniquely focuses on the synergistic impact of combining IoT-enabled real-time data acquisition with intelligent decision-making powered by machine learning.
• Includes an in-depth analysis of temporal data, environmental cost assessments, and crop-specific data augmentation strategies.
• Incorporates a discussion of security protocols such as homomorphic encryption, federated learning, and lightweight cryptographic models, enhancing relevance for both research and practical implementation.
• Establishes a unified framework spanning technical, operational, and security aspects, serving as a benchmark for future studies and smart farming applications.

Funding

No funding was received for this study.

Citation

@article{Sudha2026review,
  author = {Sudha, S. P. and Loret, J. B. Shajilin},
  title = {A review on machine learning-based precision agriculture techniques for crop farming monitoring with IOT},
  journal = {Discover Environment},
  year = {2026},
  doi = {10.1007/s44274-025-00305-8},
  url = {https://doi.org/10.1007/s44274-025-00305-8}
}