Smart & Green: An Internet-of-Things Framework for Smart Irrigation

Area of Science:

• Agricultural Engineering
• Data Science
• Environmental Science

Background:

• Irrigation is a major global water consumer, with increasing demand.
• Smart agriculture systems collect data but often contain outliers, impacting irrigation management precision.
• Data synchronization and preprocessing are crucial for effective smart irrigation.

Purpose of the Study:

• To introduce the Smart&Green framework for smart irrigation services.
• To enhance irrigation management through data preprocessing, fusion, and soil moisture prediction.
• To improve water use efficiency in agriculture.

Main Methods:

• Developed the Smart&Green framework for data monitoring, preprocessing, fusion, and synchronization.
• Implemented outlier removal techniques (Z-score, MZ-score, Chauvenet).
• Utilized a prediction model for soil matric potential using weather, crop, and irrigation data, applied to the Van Genuchten model.

Main Results:

• Outlier removal techniques significantly improve irrigation management precision.
• The framework enables accurate soil moisture prediction for fields lacking sensors.
• Average irrigation water savings ranged from 56.4% to 90% using the proposed methods.

Conclusions:

• The Smart&Green framework provides a robust solution for smart irrigation management.
• Accurate data preprocessing and soil moisture prediction are key to water conservation in agriculture.
• The proposed methods offer substantial water savings, contributing to sustainable agricultural practices.