How AI is Transforming Irrigation Management for Farmers

CIWA+: AI-Driven Irrigation Management Platform

Faculty

Athanasios Aris Panagopoulos, Department of Computer Science, College of Science and Mathematics, Fresno State

Student Researchers

• Undergraduate Students (Fresno State): Brian Kurzeja, Samuel Christopherson, Laura Mas Serra, Praise Okoli, Dalton Cowin
• Volunteers (Fresno State Undergraduates): Sarida Ngo
• External Collaborators:
  
  • Ioannis Skouroliakos, Diploma Student, Technical University of Crete
  • Janaki Panneerselvam, PhD, Auburn University
  • Minas Pantelidakis, PhD Researcher
  • Amani Mankouri, Visiting PhD Researcher
  • Arvin Agarwal, High School Student, Fresno Unified School District

Project Description

The CIWA+ project is an AI-based irrigation management system that helps growers make efficient water use decisions. It uses RGB and thermal images captured with a compact FLIR camera and processed with machine learning models such as Convolutional Neural Networks and Vision Transformers. The models identify sunlit leaves and calculate the Crop Water Stress Index, which indicates when irrigation is needed. The project also investigates multi-crop support, thermal channel integration, drone-based image collection, and mobile application development to make the tool practical and accessible for farmers.

Why Is This Research Important/Relevant

Agriculture in California and globally faces critical water shortages. CIWA+ provides farmers with a low-cost and easy-to-use technology that supports more precise irrigation, reduces water waste, and promotes sustainable farming practices. The project also trains undergraduate, graduate, and high school students, preparing future researchers and professionals in agricultural technology.

Research Outcomes/Results

• Development and annotation of almond, citrus, and grape datasets
• Creation of a segmentation pipeline and evaluation of Vision Transformer models
• Refinement of Crop Water Stress Index methodologies for improved accuracy
• Architectural documentation for CIWA+ and progress toward a mobile application
• Engagement with growers, outreach to students, and collaborations with universities and research centers
• Establishment of high-performance computing resources to support deep learning research