Saucedo-Martínez et al. (2025) Development of a smart irrigation system integrating IoT and Tree-Based Machine Learning Techniques

Identification

• Journal: Revista de Tecnología Informática
• Year: 2025
• Date: 2025-12-30
• Authors: Alan Francisco Saucedo-Martínez, Jorge Antonio Rodríguez-Contreras
• DOI: 10.35429/jct.2025.9.21.7.1.5

Research Groups

• Universidad Tecnológica de Calvillo

Short Summary

This paper presents the development and validation of an AI-powered intelligent irrigation system that integrates IoT technologies, a cross-platform mobile application, and tree-based machine learning models to optimize water usage and improve operational efficiency in agriculture. The system achieved perfect classification metrics (1.00) in field validation for autonomously predicting irrigation requirements based on real-time sensor data and weather forecasts.

Objective

• To develop and validate an intelligent agricultural irrigation system integrating IoT technologies, a cross-platform mobile application, and artificial intelligence models to optimize water usage and improve operational efficiency in the agricultural sector.

Study Configuration

• Spatial Scale: Local field application, designed for scalability to diverse crops and agricultural environments.
• Temporal Scale: Real-time data collection and decision-making, integrating short-term weather forecasts.

Methodology and Data

• Models used: Decision Tree, Random Forest, XGBoost (supervised classification algorithms).
• Data sources: Real-time soil moisture sensors, ambient temperature sensors, weather API for real-time forecasts, historical environmental and climatic data (for model training).

Main Results

• All three implemented machine learning models (Decision Tree, Random Forest, XGBoost) achieved perfect classification performance with an Accuracy, Precision, Recall, and F1-Score of 1.00 on the test dataset for predicting irrigation needs.
• The system enables fully autonomous irrigation control, significantly reducing human intervention and optimizing water usage.
• Field validation confirmed that the models' recommendations perfectly aligned with optimal irrigation practices determined by agronomic experts.
• The integrated system provides predictive capabilities through climate forecasts and real-time remote control via a mobile application.

Contributions

• Implementation of an AI-based autonomous irrigation system achieving perfect metrics (1.00) in all evaluations.
• Integration of predictive climate analysis and real-time remote control through a mobile application.
• Modular and scalable design adaptable to different crops and agricultural environments.
• Establishment of a technological foundation for future extensions, including early pest detection via computer vision.

Funding

• This research did not receive external funding. It was supported by the Universidad Tecnológica de Calvillo as part of the institutional research activities of full-time professors.

Citation

@article{SaucedoMartínez2025Development,
  author = {Saucedo-Martínez, Alan Francisco and Rodríguez-Contreras, Jorge Antonio},
  title = {Development of a smart irrigation system integrating IoT and Tree-Based Machine Learning Techniques},
  journal = {Revista de Tecnología Informática},
  year = {2025},
  doi = {10.35429/jct.2025.9.21.7.1.5},
  url = {https://doi.org/10.35429/jct.2025.9.21.7.1.5}
}