The University of Hawaiʻi at Mānoa has secured $2 million from the National Science Foundation to lead a four-year project developing AI tools that can spot environmental threats to food production before they escalate into full-blown crises. The work, which began September 1, is part of a larger $4 million NSF Established Program to Stimulate Competitive Research Track II Focused EPSCoR Collaboration award conducted jointly with the University of Nebraska-Lincoln.
Current monitoring methods typically catch problems only after disease outbreaks have started or contaminants already reached harmful concentrations. This project takes a different approach, training AI models to continuously analyze biological and chemical data streams for earlier warning signs that precede system failures.
How the detection system will work
The research team will combine AI with environmental sampling, genetic analysis, and high-resolution chemical analysis to monitor risks across aquaculture, livestock, and agricultural systems. The AI models will process data from metagenomics - the study of genetic material recovered directly from environmental samples - and high-resolution mass spectrometry, which identifies chemical compounds at precise levels.
Tao Yan, director of UH Mānoa's Water Resources Research Center and a professor in the Department of Civil, Environmental and Construction Engineering, leads the project. The research team also includes scientists from the College of Engineering, UH Cancer Center, School of Ocean and Earth Science and Technology, and College of Tropical Agriculture and Human Resilience.
Real-world testing in production environments
Rather than remaining in a lab, the technology will face real production conditions. Researchers plan to test the AI-enabled tools in aquaculture operations and beef cattle production systems, where early detection of pathogens or chemical shifts could prevent economic losses and food supply disruptions.
The collaboration extends beyond the two universities. The project will create research opportunities for junior faculty, graduate and undergraduate students, and K-12 participants. UH said outreach efforts will include industry partners, regulators, and local communities, with the broader goal of building research capacity and sustained collaboration between Hawaiʻi and Nebraska.
Why this matters for science and research professionals
For researchers working at the intersection of environmental science and machine learning, this project signals where NSF funding is heading: toward predictive systems that integrate multiple data types rather than single-method monitoring. The combination of metagenomics and mass spectrometry with continuous AI analysis represents a methodological shift from reactive sampling to ongoing surveillance. Scientists developing similar approaches for water quality, soil health, or food safety should watch how the team handles the data integration challenges - particularly the computational demands of running high-resolution chemical analysis alongside genetic sequencing in near real-time. The project's dual focus on aquaculture and beef cattle also provides a useful comparison framework for how the same AI architecture performs across very different biological systems.
Your membership also unlocks:
