ARGOS: Machine Learning Irrigation Reduces Water Usage and Improves Plant Growth

Overview

Approximately 70% of global freshwater is used for agricultural irrigation, even though 40% of agricultural freshwater is wasted annually, making water conservation a critical environmental challenge. The Autonomous Robotic Garden for Optimized Sustainability (ARGOS) was designed to reduce water waste while maintaining healthy plant growth through precise, data-driven irrigation. ARGOS uses a Raspberry Pi and Raspberry Pi Camera with a machine learning model trained on over 25,000 soil images to estimate soil moisture percentage in real time. Based on these predictions, the system delivers multiple small doses of water throughout the day to maintain an ideal moisture range, prevent runoff, and improve irrigation efficiency. Because the system is camera-based, it also has the potential to be scaled to aerial or large-scale agricultural monitoring systems. ARGOS achieved a mean absolute error of approximately 4% in soil moisture prediction, enabling accurate watering decisions. The system was tested over a seven-day period using six salad plants divided into an ARGOS group and a control group. Compared to the control group, ARGOS reduced total water usage by 58% and increased biomass by approximately 9%. These results demonstrate that ARGOS effectively meets its design goals by conserving water, improving plant health, and supporting sustainable agriculture through autonomous, intelligent control.

Competition history

• CSEF 2026 Environmental Engineering (Track 2) (Junior Division) · Entry J-12-15

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Environmental Engineering projects
• CSEF Environmental Engineering projects
• Every CSEF 2026 project