Source: COLORADO STATE UNIVERSITY submitted to
HYPERAIXPERT PLANT PHENOTYPING EQUIPMENT FOR DISCOVERY IN INTEGRATIVE DIGITAL AGRICULTURE
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
ACTIVE
Funding Source
Reporting Frequency
Annual
Accession No.
1033263
Grant No.
2024-70410-43775
Cumulative Award Amt.
$500,000.00
Proposal No.
2024-06100
Multistate No.
(N/A)
Project Start Date
Sep 15, 2024
Project End Date
Sep 14, 2028
Grant Year
2024
Program Code
[EGP]- Equipment Grants Program
Project Director
Dao, P.
Recipient Organization
COLORADO STATE UNIVERSITY
(N/A)
FORT COLLINS,CO 80523
Performing Department
(N/A)
Non Technical Summary
We propose to purchase a HyperAIxpert plant phenotyping system for rapid real-time analysis of plant phenotypes impacted by pests, diseases, beneficial microbial communities, pesticides, and other biotic and abiotic stress. The HyperAIxpert system has an RGB sensor, a PAM fluorescence sensor, a 3D laser scanner, and two hyperspectral sensors (400-2500 nm) to measure plant responses as well as advanced software to acquire and interpret the images. Colorado State University currently has no comparable system on campus, so faculty at CSU are unable to simultaneously measure a suite of plant responses indicative of plant health. This equipment will be used for work in sustainable pest management, plant adaptation, and regenerative agriculture research. The HyperAIxpert system is equipped with the similar types of sensors that are used by the CSU Drone Center and Dr. Phuong Dao's (the project PI) group, which will allow researchers to go from phenotypes observed in the lab to large-scale applications in integrative precision agriculture. This equipment will support graduate and undergraduate research in multiple programs at CSU, including people in the departments of Agricultural Biology, Horticulture, Crop and Soil Sciences, and Biology, which together house approximately 130 graduate students and approximately 150 undergraduate researchers each year. In addition, the system will be critical to the training of about 200 students in our new Agricultural Data Science minor (led by PI Dao) each year. The equipment will be housed in Plant Sciences and be managed in partnership between the Department of Agricultural Biology and the Plant Growth Facilities.
Animal Health Component
30%
Research Effort Categories
Basic
20%
Applied
30%
Developmental
50%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
4027210106030%
2032499100015%
2062499102015%
1022499107010%
2152499116020%
2132499114010%
Goals / Objectives
Our objective is to purchase a HyperAIxpert plant phenotyping system for rapid real-time analysis of plant phenotypes impacted by pests, diseases, beneficial microbial communities, and pesticides. The HyperAIxpert system has multiple types of sensors to measure plant responses as well as advanced software to acquire and interpret the images. Colorado State University (CSU) and surrounding institutions in Rocky Mountain Regions currently have no comparable phenotyping system. This equipment allows our faculty, postdocs, and students to simultaneously measure a suite of plant responses indicative of plant health. Similar sensors are available on drones managed by the CSU Drone Center, so this will allow us to connect field and laboratory studies and more quickly make progress on many aspects of integrative digital agriculture.This equipment will be used for work in sustainable pest management, plant adaptation, and regenerative agriculture research, which fit the USDA strategic goals, including combating climate change and providing all Americans safe, nutritious food. The HyperAIxpert equipment will support graduate and undergraduate research in multiple programs at CSU, including Agricultural Biology, Horticulture, Crop and Soil Sciences, and Biology, which together train approximately 90 graduate students and 200 undergraduate researchers in plant sciences and allied sciences, such as entomology, each year. We also expect to support researchers and students in various departments across the campus.The equipment will be housed in the Plant Sciences Building and be managed by the CSU Plant Growth Facilities (PGF), which manages multiple PGFs on campus. Fee collection and a machine use schedule will be developed by the PGF, as is done for other plant science-related equipment, such as growth chambers and some lab space. Students will receive training on this equipment from the team managing the HyperAIxpert system. They will have the opportunity to learn how to take their discoveries from the lab to the field through training and collaborations with the CSU Drone Center, which has drones that carry similar sensors. We expect the HyperAIxpert system to be functional within the first year of this grant.
Project Methods
Research:We propose to purchase a HyperAIxpert plant phenotyping system for rapid real-time analysis of plant phenotypes impacted by pests, diseases, beneficial microbial communities, pesticides, and other biotic and abiotic stress.This equipment will be used for work in sustainable pest management, plant adaptation, and regenerative agriculture research.With a consistent measurement environment and high-precision sensors in the HyperAIxpert system, we will be able to precisely observe the spectral, trait, and genetic variation in plants in response to various biotic and abiotic stresses under controlled environments (e.g., growth chambers and research greenhouses). The measurements collected in this system will then be scaled up to drone data in field environments. With these combined analyses, we will be able to test observations made in microscopy and genetic studies as well as remove spurious correlations in drone data that are caused by the effects of environmental confounding factors. This allows us to obtain accurate and actual information of plant characteristics and health and soil properties at the landscape scale. With that unprecedented capacity, this integrated system provides large-scale phenotype data for the large-scale selection of crop genes and traits that are important to agriculture in crop breeding programs.In addition to two hyperspectral sensors, the HyperAIxpert system is also equipped with PAM fluorescence sensor, a 3D laser scanner. These integrated sensors provide valuable information about plant properties (e.g., height, growth rate, number of leaves and petioles, photosynthetic activity, light use efficiency, leaf temperature, leaf constituents, root chemicals and structure, seed characteristics, fruit quality) and soil properties (soil nutrient status, moisture, texture, and soil contamination). Discoveries made with the HyperAIxpert system willthen be integrated with Dao Lab's drone and ground-based hyperspectral systems and drone thermal, multispectral, and LiDAR systems at the CSU Drone Center to make the best use of the system and to foster large-scale applications in integrative precision agriculture in the Department of Agricultural Biology, College of Agricultural Sciences, and across CSU.The HyperAIxpert system will be housed in the Plant Sciences Building, which also houses faculty from Agricultural Biology, Soil and Crop Sciences, and Horticulture and Landscape Architecture. It will be primarily managed by Dr. Phuong Dao, with assistance from the Plant Growth Facilities staff. A fee-based structure will be used to provide funds for maintenance of the equipment. Data generated from HyperAIxpert will be stored in Dao Lab' two storage servers that can store 300TB of data with an application server and a backup server. Additional storage is also available in CSU RStor server. These big hyperspectral images will be processed with Dao's Lab computer server with 32 cores and a 256GB RAM and the CSU Alpine high-performance computing (HPC) system. The system is also equipped with comprehensive AI-based analytical software with a standard analysis pipeline for each sensor that allows users to set up experiments, analyze images, and visualize information.With this advanced phenotyping system, we can establish a strong interdisciplinary research cluster working toward advancement in image analysis, remote sensing, machine learning, and computer vision in integrated pest management, plant pathology, plant breeding, synthetic plant biology, and plant adaptation to environmental stress. This system also fosters the development of advanced machine learning and computer vision methods for precision agriculture applications in collaboration with the Department of Computer Science and College of Engineering. This system will bolster CAS College's research and educational investments in regenerative and climate-smart agriculture to meet the college strategic goals.Student training:This equipment will support graduate and undergraduate research in multiple programs at CSU, including people in the departments of Agricultural Biology, Horticulture, Crop and Soil Sciences, and Biology, which together house approximately 130 graduate students and approximately 150 undergraduate researchers each year. In addition, the system will be critical to the training of about 200 students in our new Agricultural Data Science minor (led by PI Dao) each year. We will incorporate the system and its data into teaching of our plant biology, ecology, plant pathology, and agricultural data science courses and the capstone courses and internship programs. We will also prioritize data collection for research projects of graduate students and postdocs.Extension and outreach:This system measures and analyzes plant health and traits, providing a powerful tool for outreach and extension efforts. We will work with CSU Agricultural Experiment Stations to develop solid outreach and extension programs in which we will use the system and its data to showcase best practices, compare varieties, and predict crop performance under various scenarios. We will organize tours and training workshops to make this advanced system and techniques more accessible to a wider audience. This hands-on approach not only enhances learning but also fosters more informed decision-making, ultimately leading to improved agricultural productivity and sustainability.Farmers and stakeholders frequently send us their plant samples across the state and ask us to analyze the condition or diseases. We can also use the system to analyze and predict plant health conditions and diseases for these samples in a timely manner, providing them with valuable information for timely action on plant protection.Furthermore, we can also organize tours and workshops to introduce the system to high-school students to raise their aware of the importance of applying advanced technologies in agriculture and may encourage them to pursue a career in digital agriculture in the future.