LoRLoReAIS: Local Grower Knowledge + Remote Satellite Data for AI-Enabled Irrigation Decision Support

Irrigation decisions directly affect crop productivity, farm profitability, and water use. This project will develop the first functional prototype of LoReAIS (Local-Remote AI Synthesis), an interactive AI-enabled irrigation decision-support system that combines remote satellite information with local grower knowledge to support field-specific irrigation decisions. LoReAIS will integrate multiple sources of agricultural information, including satellite-based evapotranspiration (ET), weather data, soil and crop information, irrigation records, soil-moisture measurements when available, and grower observations and operational constraints. The system will allow growers to interact with the platform, provide local knowledge, review irrigation recommendations, and record the irrigation decisions they ultimately make.

https://et-irrigate.com/

Irrigation decisions influence economic profitability and scarce water resource consumption. This project aims to improve agricultural productivity, water use efficiency, and grower satisfaction by engineering and deploying and piloting the Local-Remote AI Synthesis (LoReAIS) system.

The student team will play a central role in engineering and implementing the LoReAIS decision-support platform. Working closely with faculty researchers, agricultural scientists, and growers, students will develop a web-based system that integrates heterogeneous agricultural data, including historical irrigation records, soil-moisture measurements, environmental and weather data, satellite-based ET estimates, and grower-provided knowledge, into a unified data and AI platform that can be incorporated into the existing AgWater.org infrastructure.

The student team will be responsible for several key components of the project:

1. Data integration and database development: Design and implement a scalable database and data pipeline for collecting, cleaning, storing, and integrating historical and real-time data from sensors, satellites, weather sources, and grower input. The system should be designed so that additional data sources can be incorporated in the future.
2. Interactive web-based decision-support platform: Develop an intuitive web portal through which growers and researchers can select and define fields, visualize field conditions, explore historical irrigation and environmental data, provide local knowledge/observations and operational constraints, and receive field-specific irrigation recommendations, and record the irrigation decision they actually made.
3. AI, data fusion, and adaptive learning: Develop and test methods for combining remote information with local grower knowledge to support irrigation recommendations. Students will explore advanced reinforcement learning (RL), data fusion, and other adaptive learning techniques to learn irrigation policies from historical data and environmental conditions, and how remote sensing information and grower knowledge can be incorporated into RL-based irrigation decisions that can improve over time.
4. Data visualization and analytics: Develop interactive tools for visualizing soil moisture, ET, weather, irrigation events, crop conditions, and AI-generated recommendations, enabling users to understand the factors underlying irrigation decisions and allow them to provide feedback.

basic programming skill

database, web programming, some machine learning knowledge