Bayar et al. (2025) Artificial intelligence of things (AIoT) for precision agriculture: applications in smart irrigation, nutrient and disease management

Identification

• Journal: Smart Agricultural Technology
• Year: 2025
• Date: 2025-11-18
• Authors: Jalal Bayar, Nawab Ali, Zhichao Cao, Yidong Ren, Younsuk Dong
• DOI: 10.1016/j.atech.2025.101629

Research Groups

• Biosystems and Agricultural Engineering, Michigan State University, East Lansing, Michigan, USA
• Computer Science and Engineering, Michigan State University, East Lansing, Michigan, USA

Short Summary

This review comprehensively examines the applications of Artificial Intelligence of Things (AIoT) in smart irrigation, nutrient, and disease management within precision agriculture, highlighting its transformative potential, current challenges, and future opportunities for sustainable and climate-smart agricultural practices.

Objective

• To examine the current state of AIoT applications in precision agriculture, focusing on how combined AI and IoT systems improve the efficiency of smart irrigation, nutrient management, and disease control.
• To analyze various AI algorithms and IoT-based sensing and communication technologies used in real-time agricultural monitoring and automated decision-making.
• To highlight the environmental, economic, and technical challenges limiting the widespread adoption of AIoT technologies in agricultural systems.

Study Configuration

• Spatial Scale: Global, reviewing applications across diverse agricultural systems (e.g., greenhouses, vineyards, cherry orchards, rice, potato, mango, apple farms).
• Temporal Scale: Review of current (past decade) and emerging AIoT technologies, advancements, and future directions in precision agriculture.

Methodology and Data

• Models used: Artificial Neural Networks (ANN), Support Vector Machine (SVM), Random Forest (RF), Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), Decision Trees, Gradient Boost Regressor, Neural Network Regressor, K-Nearest Neighbor (KNN), Logistic Regression, YOLOv5, AlexNet, INAR-SSD. Digital Twin models.
• Data sources: Interconnected sensor networks (soil moisture, temperature, humidity, pH, NPK, leaf wetness, light intensity, water flow, air temperature, wind speed, rain), edge and cloud computing platforms, satellite imagery, Unmanned Aerial Vehicles (UAVs), weather stations, real-time field images, spectral data (350–2500 nanometres), historical irrigation data, crop growth data, soil tests, historical yield records, spore germination data.

Main Results

• AIoT systems significantly optimize water distribution in irrigation through real-time soil moisture sensing, predictive analytics (using ML/DL models like LSTM, CNN, Reinforcement Learning), and dynamic scheduling, leading to substantial water savings (e.g., up to 90% in some case studies, 20.5% energy/water savings in strawberry orchards).
• Precision nutrient management leverages UAVs, soil sensors, and AI-powered data analysis to monitor nutrient availability and recommend optimized fertilization strategies, improving nutrient efficiency and reducing environmental degradation (e.g., 98.75% prediction accuracy for fertilizer application, yield increases of approximately 1.3 to 3.5 tonnes per hectare for cereals).
• AIoT in disease management enhances surveillance and predictive capabilities by integrating sensor data with AI models (e.g., CNN, AlexNet, YOLOv5) for early detection of abiotic stresses, enabling timely and targeted interventions (e.g., 98% accuracy for plant disease diagnosis, 89.4% prediction accuracy for rice blast, 98.33% for potato and mango leaf diseases).
• Key challenges identified include high setup costs, data connectivity issues in rural areas, inconsistent sensor reliability, cybersecurity risks, limited scalability for smallholder farms, technical knowledge gaps, and inadequate infrastructure.
• Opportunities for scaling AIoT in agriculture are driven by advancements in 5G technology, edge computing, and sensor miniaturization, promoting sustainable agriculture and climate change adaptation.

Contributions

• Provides an integrated, system-level perspective of AIoT-driven precision agriculture, emphasizing real-time decision-making, multi-sensor data fusion, and adaptive control systems across smart irrigation, nutrient management, and pest and disease control.
• Synthesizes insights from peer-reviewed literature and real-world implementations, offering a comprehensive understanding for researchers, policymakers, and agricultural practitioners on how AIoT is transforming modern agriculture.

Funding

Not explicitly stated in the provided text.

Citation

@article{Bayar2025Artificial,
  author = {Bayar, Jalal and Ali, Nawab and Cao, Zhichao and Ren, Yidong and Dong, Younsuk},
  title = {Artificial intelligence of things (AIoT) for precision agriculture: applications in smart irrigation, nutrient and disease management},
  journal = {Smart Agricultural Technology},
  year = {2025},
  doi = {10.1016/j.atech.2025.101629},
  url = {https://doi.org/10.1016/j.atech.2025.101629}
}