News and Announcements
Towards Large Reasoning Models for Agriculture
Hossein Zaremehrjerdi, Shreyan Ganguly, Ashlyn Rairdin, Elizabeth Tranel, Benjamin Feuer, Juan Ignacio Di Salvo, Srikanth Panthulugiri, Hernan Torres Pacin, Victoria Moser, Sarah Jones, Joscif G Raigne, Yanben Shen, Heidi M. Dornath, Aditya Balu, Adarsh Krishnamurthy, Asheesh K Singh, Arti Singh, Baskar Ganapathysubramanian, Chinmay Hegde, Soumik Sarkar
Agricultural decision-making involves complex, context-specific reasoning, where choices about crops, practices, and interventions depend heavily on geographic, climatic, and economic conditions. Traditional large language models (LLMs) often fall short in navigating this nuanced problem due to limited reasoning capacity. We hypothesize that recent advances in large reasoning models (LRMs) can better handle such structured, domain-specific inference. To investigate this, we introduce AgReason, the first expert-curated open-ended science benchmark with 100 questions for agricultural reasoning. Evaluations across thirteen open-source and proprietary models reveal that LRMs outperform conventional ones, though notable challenges persist, with the strongest Gemini-based baseline achieving 36% accuracy. We also present AgThoughts, a large-scale dataset of 44.6K question-answer pairs generated with human oversight and equipped with synthetically generated reasoning traces. Using AgThoughts, we develop AgThinker, a suite of small reasoning models that can be run on consumer-grade GPUs, and show that our dataset can be effective in unlocking agricultural reasoning abilities in LLMs. Our project page is here: https://baskargroup.github.io/Ag_reasoning/
Read MoreWeedNet: A Foundation Model-Based Global-to-Local AI Approach for Real-Time Weed Species Identification and Classification
Yanben Shen, Timilehin T. Ayanlade, Venkata Naresh Boddepalli, Mojdeh Saadati, Ashlyn Rairdin, Zi K. Deng, Muhammad Arbab Arshad, Aditya Balu, Daren Mueller, Asheesh K Singh, Wesley Everman, Nirav Merchant, Baskar Ganapathysubramanian, Meaghan Anderson, Soumik Sarkar, Arti Singh
Early identification of weeds is essential for effective management and control, and there is growing interest in automating the process using computer vision techniques coupled with AI methods. However, challenges associated with training AI-based weed identification models, such as limited expert-verified data and complexity and variability in morphological features, have hindered progress. To address these issues, we present WeedNet, the first global-scale weed identification model capable of recognizing an extensive set of weed species, including noxious and invasive plant species. WeedNet is an end-to-end real-time weed identification pipeline and uses self-supervised learning, fine-tuning, and enhanced trustworthiness strategies. WeedNet achieved 91.02% accuracy across 1,593 weed species, with 41% species achieving 100% accuracy. Using a fine-tuning strategy and a Global-to-Local approach, the local Iowa WeedNet model achieved an overall accuracy of 97.38% for 85 Iowa weeds, most classes exceeded a 90% mean accuracy per class. Testing across intra-species dissimilarity (developmental stages) and inter-species similarity (look-alike species) suggests that diversity in the images collected, spanning all the growth stages and distinguishable plant characteristics, is crucial in driving model performance. The generalizability and adaptability of the Global WeedNet model enable it to function as a foundational model, with the Global-to-Local strategy allowing fine-tuning for region-specific weed communities. Additional validation of drone- and ground-rover-based images highlights the potential of WeedNet for integration into robotic platforms. Furthermore, integration with AI for conversational use provides intelligent agricultural and ecological conservation consulting tools for farmers, agronomists, researchers, land managers, and government agencies across diverse landscapes.
Read MorePlant patch can detect stress signals in real time
Environmental conditions can cause damaging stress to plants, posing challenges for home gardeners and farmers. Therefore, early detection — before leaves visibly discolor, wilt or wither — is crucial. Now, researchers reporting in ACS Sensors have created a wearable patch for plants that quickly senses stress and relays the information to a grower. The electrochemical sensor attaches directly to live plant leaves and monitors hydrogen peroxide, a key distress signal.
Pests, drought, extreme temperatures and infections all cause stress in plants. In response, plants' normal biochemistry gets out of whack, and they produce hydrogen peroxide, which also acts as a signal between cells to activate their defense mechanisms. Early detection of this chemical clue could help people expertly tailor plant care and prevent further damage, thereby maximizing crop yields, even in difficult conditions. But most current methods for detecting hydrogen peroxide require removal of plant parts and multiple processing steps or external detectors that observe fluorescence changes, which can get muddled by chlorophyll. And researchers have previously investigated plant-wearable devices to monitor leaf water content as an indicator of plant health. So, Liang Dong and colleagues set out to design a stand-alone patch that quickly and accurately detects the hydrogen peroxide distress signals from living plants.
Read More
Other Versions of this Article:
Picture this: Massive photo collection could fuel AI innovations in crops
Artificial intelligence (AI) needs mountains of data to operate accurately and effectively, and scientists backed by USDA's National Institute of Food and Agriculture (NIFA) just dropped the Mt. Everest of all image collections to help further biodiversity research.
The massive new collection of images, called Arboretum, includes captions and detailed data of nearly 327,000 plant and animal species — includes 134.6 million expert-verified images. The next largest dataset contains just 10.2 million images.
By making this resource freely available to the public, the teams behind the project hope to stimulate the development of new AI tools that can assist with tasks like pest control, crop monitoring and environmental conservation.
Read More
NIFA-Funded Scientists Spur Research in Artificial Intelligence with Release of Massive Dataset
To spur research in artificial intelligence (AI) for biodiversity research, a team of scientists funded by the National Institute of Food and Agriculture (NIFA) has launched Arboretum, a massive new collection of images with rich text captions and metadata of nearly 327,000 plant and animal species.
The dataset, which is an order of magnitude beyond anything similar before it, includes 134.6 million expert-verified, captioned images curated from the iNaturalist citizen science platform. In comparison, the next largest dataset contains just 10.2 million images. By making this resource freely available to the public, the creators hope to stimulate the development of new AI tools that can assist with tasks like pest control, crop monitoring and environmental conservation.
This research is important because it can lead to the development of AI tools that help address critical global
challenges such as food security, ecosystem preservation and climate change mitigation,
said Chinmay Hedge,
associate professor at New York University Tandon. These directly impact average Americans in terms of better
food prices, public health and the overall quality of the environment.
Read More
New institute aims to unlock the secrets of corn using artificial intelligence
Iowa State University researchers are growing two kinds of corn plants.
If you drive past the many fields near the university's campus in Ames, you can see row after row of the first. But the second exists in a location that hasn't been completely explored yet: cyberspace.
The researchers, part of the AI Institute for Resilient Agriculture, are using photos, sensor data and artificial
intelligence to create digital twins
of corn plants that, through analysis, can lead to a better
understanding of their real-life counterparts. They hope the resulting software and techniques will lead to
better management, improved breeding, and ultimately, smarter crops.
We need to use lots of real-time, high-resolution data to make decisions,
Patrick Schnable, an
agronomy professor and director of Iowa State's Plant Sciences Institute, told Agri-Pulse. Just collecting
data for data's sake is not something that production ag wants. But data which is then linked to statistical
models or other kinds of mathematical models that advise farmers on what to do has a lot of value.
Artificial intelligence next step in the future of Ag decision making
Three universities in the Corn Belt are part of an initiative that uses artificial intelligence to ramp up crop yields, create new seed varieties and allow producers to be more profitable.
James Schnable, an associate professor with the
University of Nebraska, says the technology models crop conditions
under different circumstances that change over several decades. In both cases, the challenge we face is
that we cannot test every possible combination of different varieties of crops grown in different parts
of the state or different parts of the country with different growing practices.
The AI Institute for Resilient Agriculture also includes Iowa State and the University of Missouri as part of a $20 million grant from the National Science Foundation and USDA's National Institute of Food and Agriculture.
Read MoreUSDA-NIFA and NSF Invest $220M in Artificial Intelligence Research Institutes
The U.S. Department of Agriculture's National Institute of Food and Agriculture (USDA-NIFA) and the U.S. National Science Foundation (NSF) announced a $220 million investment in 11 new NSF-led Artificial Intelligence Research Institutes. USDA-NIFA and other agencies and organizations have partnered with NSF to pursue transformational advances in a range of economic sectors and science and engineering fields - from food system security to next-generation edge networks.
The new investment builds on the first round of seven Artificial Intelligence (AI) Research Institutes funded in 2020, totaling $140 million last year.
Read MoreNSF Partnerships Expand National AI Research Institutes to 40 States
The U.S. National Science Foundation announced the establishment of 11 new NSF National Artificial Intelligence Research Institutes, building on the first round of seven institutes funded in 2020. The combined investment of $220 million expands the reach of these institutes to include a total of 40 states and the District of Columbia.
The institutes are focused on AI-based technologies that will bring about a range of advances: helping older adults
lead more independent lives and improving the quality of their care; transforming AI into a more accessible
plug-and-play
technology; creating solutions to improve agriculture and food supply chains;
enhancing adult online learning by introducing AI as a foundational element; and supporting underrepresented
students in elementary to post-doctoral STEM education to improve equity and representation in AI research.
$20 Million Federal Grant Launches AI Institute for Better Crops, Agricultural Production
The latest artificial intelligence tools will allow researchers to develop digital twins of individual crop plants and entire farm fields, helping plant breeders improve crop varieties and farmers boost production.
The researchers behind a new artificial intelligence research institute say their work can accelerate the productivity and sustainability of agriculture at a time when the world’s population is increasing, cropland is decreasing and the climate is changing.
The National Science Foundation and the U.S. Department of Agriculture's
National Institute of Food and Agriculture are supporting the researchers'
idea with a five-year, $20 million grant to establish an AI Institute for Resilient Agriculture (AIIRA – eye-rah
)
based at Iowa State University. The institute is one of 11 AI institutes announced today.
Other Versions of this Article:
Husker Researcher Part of Multi-University Effort to Improve Ag Decision-Making
The use of digital twins,
virtual copies of physical objects and operations, is gaining steam across
a wide range of industries. Updated constantly with real-time data, these virtual mirrors allow engineers to
keep an eye on and predict traffic flow, retailers to optimize supply chains and railway operators to spot wear
and tear on tracks. Researchers are even working toward digital twins of the human heart, which would let doctors
diagnose, treat and monitor patients from afar.
Until now, the technology had not been widely employed in agriculture, even as the world races to secure a sustainable food supply for a population on track to reach nearly 10 billion by 2050. Today, the National Science Foundation and U.S. Department of Agriculture's National Institute of Food and Agriculture announced a five-year, $20 million grant to establish the AI Institute for Resilient Agriculture, or AIIRA. It's part of a $200 million federal effort to develop artificial intelligence hubs that address a variety of national needs.
Read MoreCyVerse Receives $1.3M to Provide Cyberinfrastructure and Training for New NSF Artificial Intelligence Institute
The University of Arizona will take part in a $20 million institute that aims to transform agriculture through artificial intelligence.
The Artificial Intelligence Institute for Resilient Agriculture, led by Iowa State University and funded by the U.S. Department of Agriculture National Institute of Food and Agriculture, will focus on innovative AI-driven methods for agriculture, promote the study of cyber-agricultural systems, and support education, workforce development and community engagement.
With $1.3 million from USDA-NIFA, CyVerse – headquartered at the University of Arizona BIO5 Institute - will provide the institute with expertise in cyberinfrastructure, along with education and engagement opportunities for Native Nations, farmers and community stakeholders to address how technological advances in AI can answer agricultural needs.
Read MoreOther Versions of this Article:
Mason to Participate in New Five-Year $20 Million Grant
The proliferation of technology, in particular emerging platforms and services that deploy sensors and artificial intelligence (AI), creates opportunities for improving society.
The National Science Foundation and the U.S. Department of Agriculture's
National Institute of Food and Agriculture announced a five-year,
$20 million grant to establish an AI Institute for Resilient Agriculture (AIIRA – eye-rah
) based at Iowa State University.
The institute is one of 11 new National Artificial Intelligence Research Institutes that NSF has established in 2021
with a total investment of $220 million. George Mason University professor
Aditya Johri will serve as a member of the
institute's education and outreach team.
AIIRA aims to transform agriculture by creating a new AI-driven framework for modeling plants at various agronomically relevant scales. The researchers will accomplish this by introducing AI-driven digital twins that fuse diverse data with siloed domain knowledge. They will deploy these twins across agricultural applications that directly impact the USDA Science Blueprint for crop improvement and production.
Read MoreTandon Researcher Joins Major Collaboration Aimed at Using AI Models to Improve Agriculture
Chinmay Hegde, professor of computer science and engineering and electrical and computer engineering at the NYU Tandon School of Engineering is part of a multi-institutional collaboration to pursue foundational advances in artificial intelligence (AI) to enhance the resiliency of the nation’s agricultural ecosystem.
The program, AIIRA: AI Institute for Resilient Agriculture, is a national Research Institute supported by a $20M, five-year grant from the U.S. National Science Foundation. It is led by Iowa State University, and includes researchers at Carnegie Mellon, the University of Arizona, George Mason University, and the University of Nebraska, along with several industry and federal partners.
Driving these advances is the concept of biophysics-aware AI-driven digital twins that are capable of assimilating sensor data along with agronomic knowledge to build accurate models of agricultural phenomena across scales, from individual plants to plots/fields.
Read MoreMU Plays Crucial Role in New NSF Artificial Intelligence Institute
The U.S. National Science Foundation (NSF) today announced the establishment of the AI Institute for Resilient Agriculture (AIIRA), one of 11 new NSF National Artificial Intelligence Research Institutes that will work to advance AI technologies and their associated benefits to society. The AIIRA will be led by Iowa State University (ISU) and feature collaboration across eight institutions, including the University of Missouri.
While the 11 AI Institutes will cover a wide range of AI technology and applications, the AIIRA
will focus on a simulation technology that can create digital twins
of real-world crops
and farms, an approach that would provide specific, detailed information for better-informed
crop management, technological development and agricultural policies. Using sensors to
send real-time data about weather, soil composition and other factors to the digital twin,
researchers plan to create a system that can accurately predict the outcome of a variety
of what-if
scenarios.
