UTA opens AI-driven Smart Agriculture Research Center

USDA-backed program puts students on research teams forecasting bird flu and other ag threats

University of Texas at Arlington

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

UTA's Smart Agriculture Research Center grand opening.

view more 

Credit: UT Arlington

In the last year, highly pathogenic avian influenza (HPAI), commonly known as bird flu, swept through chicken coops across the nation, killing egg-laying hens and contributing to rising egg prices. The outbreaks underscored how vulnerable food systems can be to rapidly spreading biological threats and how urgently new predictive tools are needed to help producers respond.

That’s why the U.S. Department of Agriculture’s Agricultural Research Service (USDA-ARS) partnered with The University of Texas at Arlington to develop solutions through the Smart Agriculture Research Center (SARC), a new research hub using artificial intelligence and data science to tackle agriculture’s most pressing challenges.

“Agriculture is essential to society, yet it has historically seen less AI integration than other industries,” said Jianzhong Su, professor of mathematics and co-director of SARC. “UTA has tremendous strength in technology and data science, and that positions us to help modernize agriculture in Texas and beyond.”

Opened in August 2025, SARC serves the entire UTA campus through four core pillars: providing AI capacity and data discovery tools for agriculture research projects; serving as a resource hub for faculty pursuing agriculture-related research; securing major USDA and external training and center grants; and serving as UTA’s institutional gateway for external partners focused on sustainability and global environmental impact. Dr. Su and Co-Director Gautam Das, professor of computer science and engineering, work together with more than 20 faculty members in science and engineering.

On Feb. 9, the center hosted a grand opening that included UTA and USDA officials.

“The work done by SARC will turn interdisciplinary research into practical solutions that strengthen our region and drive progress worldwide,” said Kate Miller, vice president for research and innovation at UTA, at the grand opening event. “It is a testament to our 130-year legacy and our bold future.”

Backed by growing federal investment, the center brings together UTA faculty, students and USDA-ARS scientists to apply machine learning to real-world agricultural problems—from predicting plant disease and modeling soil health to forecasting outbreaks of HPAI.

“This center is UTA’s direct response to the national call for climate-smart agriculture and resilient food systems,” said Scott Miller, associate vice president for research and innovation at UTA, during the grand opening event. “We are here to ensure that the innovations born in Arlington scale to support the entire nation.”

Students at the core of the research

A major engine behind the center’s work is a USDA-supported summer research program that immerses UTA students in federal agricultural research projects.

Each year, 20 to 25 undergraduate and graduate students—primarily from mathematics, computer science, engineering and science—participate in an eight- to 10-week summer research experience. Students are grouped into small research teams, each paired with a UTA faculty mentor and a USDA-ARS scientist, to tackle real agricultural data challenges using artificial intelligence and machine learning.

Projects span a wide range of agricultural issues, including predicting plant diseases, modeling the effects of weather on crop resilience, assessing the environmental effects of fertilizers and pesticides, and developing data-driven tools for livestock and poultry health monitoring.

One of the newest research directions focuses on predicting HPAI outbreaks. UTA researchers and USDA scientists are developing models that automatically collect publicly reported outbreak data and generate short-term forecasts. These tools can help poultry producers take preventive measures—such as enhancing biosecurity, increasing sanitation or modifying facility management—to reduce the risk of virus spread.

Although students conduct their research on UTA’s campus, they collaborate remotely with USDA scientists located across the country and participate in site visits to observe agricultural research operations firsthand. The experience provides students with direct exposure to national research networks, which helps build a highly skilled workforce in AI-enabled agriculture.

Building national research capacity

Additional collaborative USDA projects total over $5.5 million in external research investment connected to UTA faculty and ARS partners.

By connecting North Texas talent with national agricultural research networks, the Smart Agriculture Research Center aims to train the next generation of AI-enabled agricultural scientists, strengthen food and environmental resilience, and help producers respond to emerging biological and climate threats.

 

AI and process integration: charting the future of polymer composite manufacturing

Higher Education Press

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

AI-aided composite process optimization integration.

view more 

Credit: HIGHER EDUCATON PRESS

Lightweight, high-strength polymer composites are essential to modern engineering. Yet their manufacturing remains complex, slow, and often reliant on manual adjustments. In a forward-looking analysis published on October 31, 2025, in Frontiers of Chemical Science and Engineering, an international team of researchers outlines a transformative roadmap, using artificial intelligence (AI) to create fully integrated, self-optimizing production systems.

The study synthesizes recent advances and real-world cases to demonstrate how AI can bridge gaps between design, process selection, and quality control. Unlike conventional methods that treat each manufacturing step in isolation, the proposed framework uses machine learning and digital twins to simulate and optimize the entire production chain as a single, cohesive process.

“The deep integration of AI with composite material design and manufacturing will drive the industry’s transition from experience-based to data-intelligent practices.” says Dr. Zijie Wu, a corresponding author from the Yaoshan Laboratory. “AI doesn't just automate tasks. AI understands the interplay between material behavior, process parameters, and final performance, enabling us to manufacture components that are lighter, stronger, and more reliable with far less waste.”

One highlighted innovation is the use of physics-informed neural networks to model the curing stage. By learning from historical sensor data, these AI models predict the optimal heating and pressure curve for each part, reducing cycle times by up to 30% and minimizing energy use. In another example, AI integrates hot pressing with injection molding in a single run, allowing a structural base and complex functional features to be formed together, which previously required multiple separate steps.

The approach addresses critical industry challenges: the high cost of trial-and-error prototyping, inconsistency in part quality, and the difficulty of scaling up new materials. Companies like Boeing and Airbus are already piloting similar AI tools to optimize autoclave processes and automated fiber placement, reporting notable gains in precision and throughput.

For sustainability goals, the implications are significant. Lightweight composites already lower emissions in transport; making their manufacturing smarter and less resource-intensive further amplifies their environmental benefit. The technology also supports the trend toward “smart composites” with embedded sensors or self-healing capabilities, enabled by more precise and adaptable production routes.

In conclusion, this research provides a clear and practical pathway to modernize composite manufacturing. Harnessing AI for integrated process control not only elevates product quality and production agility but also reinforces the role of advanced materials in building a more innovative and sustainable industrial future.

---

Journal

Frontiers of Chemical Science and Engineering

DOI

10.1007/s11705-026-2637-7 

Method of Research

Experimental study

Subject of Research

Not applicable

Article Title

Optimization and integration of polymer composites manufacturing powered by artificial intelligence

 

Course correction needed quickly to avoid pathway to ‘hothouse Earth’ scenario, scientists say

Oregon State University

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

Allan Hills, Antarctica

view more 

Credit: Photo by Austin Carter, COLDEX.

CORVALLIS, Ore. – Scientists say multiple Earth system components appear closer to destabilization than previously believed, putting the planet in increased danger of following a “hothouse” path driven by feedback loops that can amplify the consequences of global warming.

Published today in the journal One Earth, “The risk of a hothouse Earth trajectory” is an analysis by an international collaboration led by Oregon State University’s William Ripple that synthesizes scientific findings on climate feedback loops and 16 tipping elements – Earth subsystems that may undergo loss of stability if critical temperature thresholds are passed.

Those sharp changes could likely result in a cascade of subsystem interactions that would steer the planet toward a path to extreme warming and sea level rise – conditions that could be difficult to reverse on human timescales, even with deep emissions cuts.

“After a million years of oscillating between ice ages separated by warmer periods, the Earth’s climate stabilized more than 11,000 years ago, enabling agriculture and complex societies,” said Ripple, distinguished professor of ecology in the OSU College of Forestry. “We’re now moving away from that stability and could be entering a period of unprecedented climate change.”

Tipping elements include ice sheets in Antarctica and Greenland, mountain glaciers, sea ice, boreal forests and permafrost, the Amazon rainforest and the Atlantic Meridional Overturning Circulation or AMOC, a system of ocean currents that’s a key influencer of global climate.

The researchers note that nearly 10 years after the Paris Agreement, which sought to limit long-term average warming to 1.5 degrees Celsius above preindustrial levels, global temperature increases exceeded 1.5 degrees Celsius for 12 consecutive months – a period that also included extreme, deadly and costly wildfires, floods and other climate-related natural disasters.

“Temperature limit exceedance is usually evaluated using 20-year averages, but climate model simulations suggest the recent 12-month breach indicates the long-term average temperature increase is at or near 1.5 degrees,” said study co-author Christopher Wolf, a former OSU postdoctoral researcher who is now a scientist with Corvallis-based Terrestrial Ecosystems Research Associates, known as TERA. “It’s likely that global temperatures are as warm as, or warmer than, at any point in the last 125,000 years and that climate change is advancing faster than many scientists predicted.”

It’s also likely that carbon dioxide levels are the highest they’ve been in at least 2 million years, the scientists say. At more than 420 parts per million, the atmospheric CO2 concentration is about 50% higher than it was prior to the Industrial Revolution.

When the climate changes, the researchers note, responses can be triggered that circle back to affect the climate itself, amplifying or dampening the original change. These processes are known as climate feedback loops.

“Amplifying feedbacks increase the risks of accelerated warming,” Ripple said. “For example, melting ice and snow, permafrost thaw, forest dieback and soil-carbon loss can all magnify warming – and in turn affect the climate system’s sensitivity to greenhouse gases.”

Ripple, Wolf and their collaborators – Wolf’s TERA colleague Jillian Gregg and top climate scientists in Germany, Denmark and Austria – say current data coupled with the inherent uncertainties of climate forecasting should be taken as a signal that urgent climate mitigation and adaptation strategies are needed.

“Existing climate mitigation approaches, including scaling up renewable energy and protecting carbon-storing ecosystems, are critical to limit the increase in global temperatures,” said Ripple.

Strategies that embed climate resilience into governmental policy frameworks should be a priority as well, the authors say, along with a socially just phaseout of fossil fuels. The scientists also discuss the need for novel approaches, including coordinated global tipping-point monitoring and better plans for managing risk.

“Uncertain tipping thresholds underscore the importance of precaution – crossing even some of those thresholds could commit the planet to a hothouse trajectory with long-lasting and possibly irreversible consequences,” Wolf said. “Policymakers and the public remain largely unaware of the risks posed by what would effectively be a point-of-no-return transition. And while averting the hothouse trajectory won’t be easy, it’s much more achievable than trying to backtrack once we’re on it.”

Tipping may already be happening with the Greenland and West Antarctic ice sheets, the scientists say, and boreal permafrost, mountain glaciers and the Amazon rainforest appear on the verge of tipping.

In the Earth’s tightly coupled climate system, destabilization in one region can reverberate across oceans and continents as melting ice accelerates warming by reducing albedo and altering the Atlantic Meridional Overturning Circulation, resulting in changes to tropical rain belts.

For example, as the Greenland ice sheet melts, it could further weaken the AMOC, which in turn could trigger parts of the Amazon to tip from rainforest to savanna.

“The AMOC is already showing signs of weakening, and this could increase the risk of Amazon dieback, with major negative impacts on carbon storage and biodiversity,” Ripple said. “Carbon released by an Amazon dieback would further amplify global warming and interact with other feedback loops. We need to act quickly on our rapidly dwindling opportunities to prevent dangerous and unmanageable climate outcomes.”

Working with Ripple, Wolf and Gregg on the analysis were Johan Rockström and Nico Wunderling of the Potsdam Institute for Climate Impact Research; Katherine Richardson of the University of Copenhagen; Thomas Westerhold of University Bremen; and Hans Joachim Schellnhuber of the International Institute for Applied Systems Analysis.

---

Journal

One Earth

Method of Research

Data/statistical analysis

Subject of Research

Not applicable

Article Title

The risk of a hothouse 1 Earth trajectory

Article Publication Date

11-Feb-2026

 

Not all gigs are equal: Informal self-employment linked to lower pay, poorer health and instability

University of Michigan

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

Image of the researchers

 

Not all self-employment guarantees financial security, with informal arrangements posing the greatest risks to well-being for many workers. 

 

Using machine learning to classify self-employment, a new University of Michigan study analyzed narrative job descriptions from the 2003-2019 Panel Study of Income Dynamics, a longitudinal dataset with approximately 10,000 U.S. families. 

 

They divided self-employment into three categories: business owners, formal self-employed and informal self-employed, such as gig workers and those performing odd jobs.

 

The study, published in the journal ILR Review and funded by the National Science Foundation, found large differences in financial outcomes, health and stability among these groups. It also discovered a significant shift in the labor market. Informal self-employment has increased in recent years, while formal self-employment has declined.

 

Informal self-employed workers make up the largest segment of the self-employed workforce, but fare the worst. They earn less, report poorer physical and mental health, have lower life satisfaction and face a higher risk of unemployment.

 

The study found that informal self-employment is concentrated among lower-income workers, women, Hispanic people and Black people—patterns shaped by language barriers, discrimination and limited access to opportunities.

 

"This group is worse off across multiple dimensions," said Joelle Abramowitz, associate research scientist at U-M's Institute for Social Research. "It often lacks benefits, stability and a path to growth. These workers are also far more likely to cycle in and out of unemployment. And it becomes a revolving door—when health falters, people fall out of work, turn to informal jobs to get by, and then slip back into unemployment again."

 

On the opposite side, business owners earn the highest incomes and report the highest levels of life satisfaction and physical health, the study found. They own more business assets than the other groups. Formal self-employed workers, such as consultants, rank second on all measures. 

 

"If we lump all self-employed people together, we mask critical economic trends and realities," Abramowitz said. "And more: There's a real concern that informal self-employment locks people into work that doesn't pay much and doesn't lead to long-term growth. At the same time, what's striking is how fluid self-employment can be—some people move from informal gigs into more stable or even entrepreneurial roles, and that's something we're actively studying."

 

Abramowitz and co-author Andrew Joung, a U-M graduate student, recommend policy changes to address the workforce shift, underscoring the need for targeted support. 

 

These include extending unemployment benefits to self-employed workers, increasing access to health insurance and expanding job training and entrepreneurship programs to those in informal work. Support during the transition phase may help more workers move into better opportunities and improve overall economic security.

 

Study: Opening the Black Box of Self-Employment: Identifying Alternative Work Arrangements in the United States (DOI: 10.1177/00197939251411)

 

---

Journal

ILR Review

DOI

10.1177/00197939251411263