Source: AGRICULTURE INTELLIGENCE, INC. submitted to NRP
AGROVIEW AND AGROSENSE FOR AI-ENHANCED PRECISION NUTRIENT MANAGEMENT PHASE II
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
ACTIVE
Funding Source
Reporting Frequency
Annual
Accession No.
1031177
Grant No.
2023-39410-40832
Cumulative Award Amt.
$650,000.00
Proposal No.
2023-03938
Multistate No.
(N/A)
Project Start Date
Sep 1, 2023
Project End Date
Aug 31, 2025
Grant Year
2023
Program Code
[8.13]- Plant Production and Protection-Engineering
Recipient Organization
AGRICULTURE INTELLIGENCE, INC.
747 SW 2ND AVE STE 310
GAINESVILLE,FL 326016282
Performing Department
(N/A)
Non Technical Summary
AgIntel, a spinoff company of the University of Florida (UF), has licensed two novel technologies from UF: Agroview and AgroSense. Agroview, a digital twin of an agricultural field, can deliver accurate zone-based fertility and prescription maps for precision nutrient applications. Agrosense can convert a traditional fertilizer applicator (sprayer or spreader) to a smart machine for precision nutrient application. With Agrosense, the smart machine can vary the amount of the applied chemicals based on each tree's size and leaf density.In Phase I of this SBIR project, we integrated these two novel AI technologies and developed a more accurate tree level based precision nutrient management, named Smart Nutrient. This system integrates Agrosense's real-time data (e.g., tree canopy volume and leaf density) with the provided prescription map (from Agroview) to adjust the application rate for each individual tree. A data fusion algorithm was developed on Agrosense for this purpose (aka, determining in real-time the tree canopy volume and leaf density and also identifying the desired application rate based on the management zone from the prescription map).The specific objectives of the SBIR Phase II project are to (i) Enhance and evaluate in commercial orchards the Agroview's fertility/prescription map and the Smart Nutrient feature (integration of Agroview and Agrosense), and (ii) Enhance, field-evaluate, and finalize the development of the Smart Sprayer and Smart Fertilizer Spreader. To the best of our knowledge, we are the first to develop this novel technology for tree crops.
Animal Health Component
60%
Research Effort Categories
Basic
10%
Applied
60%
Developmental
30%
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
4027210303050%
4045310202050%
Goals / Objectives
The main goal of this project is to enhance precision nutrient management in tree crops (case study on citrus) by developing a Smart Nutrient approach that integrates AI, UAV-based multispectral imaging, Agroview, and Agrosence. This novel technology will be able to utilize fertility/prescription maps (Agroview data) for zone-based and variable rate chemical applications, and at the same time, adjust the applied rate, in real-time, within a specific zone based on individual tree canopy volume and leaf density (Agrosense data). It will integrate AI-enhanced zone-based with AI-enhanced real-time sensing nutrient management. To the best of our knowledge, we are the first to develop this novel technology for tree crops. The specific objectives of this SBIR Phase II project are to develop:Enhance and evaluate in commercial orchards the Agroview's fertility/prescription map and the Smart Nutrient feature (integration of Agroview and Agrosense).Enhance, field-evaluate, and finalize the development of the Smart Sprayer and Smart Fertilizer Spreader.
Project Methods
The specific activities per objectives are:Objective #1: Enhance and evaluate in commercial orchards the Agroview's fertility/ prescription map and Smart Nutrient feature.Activity 1: Evaluate and improve the accuracy of the AI models to determine plant nutrient concentrations (from aerial multispectral imaging) and develop fertility mapsActivity 2: Investigate the integration of soil data into the Smart Nutrient feature. Extraction of soil spatial indicators and improvement based on super-resolution techniquesActivity 3: Evaluate and redesign Agroview and Agrosence UIs based on stakeholders' feedbackObjective #2: Enhance, field-evaluate, and finalize the development of the Smart Sprayer and Smart Fertilizer SpreaderActivity 4: Evaluate and redesign the Smart Sprayer and Smart Fertilizer SpreaderActivity 5: Automatic airflow control on the smart sprayer to minimize off-target lossActivity 6: Improve the precision application of multiple nutrientsOutreach activities (e.g., workshops and field days, production videos illustrating the technology) will be developed to demonstrate the developed AI-enhanced precision nutrient technology to growers and allied industries, and policymakers.

Progress 09/01/23 to 08/31/24

Outputs
Target Audience:tree crop growers and allied industry, crop insurance companies, extension personnel, and policymakers. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?Two research associates involved in this project attended and presented in two international conferences: Zhou C., Ampatzidis Y., Guan H., and Neto A.D.C., Kunwar S., Batuman O., 2024. Agrosense: AI-enabled sensing for precision management of tree crops (poster). 16th International Conference on Precision Agriculture (ICPA), Manhattan, Kansas, USA, July 21-24. Lacerda C., and Neto A.D.C., Ampatzidis Y., 2024. Agroview: enhance satellite imagery using super-resolution and generative AI for precision management in specialty crops. AgEng International Conference of EurAgEng, Agricultural Engineering Challenges in Existing and New Agrosystems, Athens, Greece, July 1-4, 2024. How have the results been disseminated to communities of interest?AgIntel CEO Matthew Donovan and other members of AgIntel presented results of this project in several Ag Expos and Ag events: Sep 2023: International Citrus & Beverage Conference (Tampa, FL) Oct 2023: UF, AI Days (Gainesville, FL) Nov 2023: AgriSompo Southeast User Group Conference (Lakeland, FL) Feb 2024: UF/Gainesville Tech Week (Gainesville, FL) Feb 2024: Future of Florida Summit (Gaineville, FL) Mar 2024: World AgriTech (San Francisco, FL) May 2024: AI in Ag (College Station, TX) Apr 2024: UF Foundation, AI Week (Gainesville, FL) Jun 2024: Florida Citrus Industry, Annual Meeting (Fort Myers,FL) Dr. Ampatzidis and his team presented the results of this project in several venues, including: Emerging enterprises and research commercialization (panel discussion). UF/IFAS Research Forum, Gainesville, FL, April 30, 2024. AI and remote sensing for precision crop management. Center for Remote Sensing, IFAS/UF, April 26, 2024. AI and robotics in agriculture. AI In-service-training (IST), Wimauma, FL, April 9, 2024. Robotics, automation, and AI for specialty crops. Florida AgTech and AI Expo, Punta Gorda, FL, December 14, 2023. AI-enabled technologies for precision management of specialty crops. USDA ARS-Northeast Area retreat, November 14, 2023. Drones and AI for Precision Crop Management. Florida Ag Expo, Gulf Coast Research and Education Center, Wimauma, FL, November 9, 2023. Artificial intelligence enhanced technology for precision citrus production. Cold Hardy Citrus Field Day, Quincy, FL, October 26, 2023. Commercializing your AI: Lessons and challenges from spinning off AI companies. UF AI Days, Gainesville, FL, October 18, 2023. What do you plan to do during the next reporting period to accomplish the goals?We will continue working on redesigning the user interfaces (UIs) of both Agroview and Agrosense based on feedback from stakeholders. We expect to release the new UIs for both systems by the end of this project. We are also working on developing a tool to integrate soil data into the Smart Nutrient feature. Furthermore, we are developing an automatic airflow control for the smart sprayer to minimize off-target losses. The designs of this system are ready, and we plan to build and field-test this system in the next six months. Finally, we will explore the development of a fuzzy logic controller for an optimal application of multiple nutrient inputs. This task will start in the next few months.

Impacts
What was accomplished under these goals? Objective #1: Enhance and evaluate in commercial orchards the Agroview's fertility/ prescription map and Smart Nutrient feature. We have refined Agroview's pixel segmentation algorithm, which identifies and isolates tree canopy pixels in aerial multispectral images. By achieving this, we can ensure that the average values extracted for each spectrum are more representative of the tree canopy (and not from the ground or weeds) and its nutrient concentrations. This leads to more accurate predictions by the AI models to generate fertility maps. We have collected multispectral images from several commercial citrus orchards. In some fields, data were collected monthly. Leaf samples were also collected and analyzed for nutrient estimations (ground truth) by our grower collaborators and Dr. Albrecht's lab (Co-PI). These data were used to improve and make more robust the AI models to deliver reliable nutrient concentration estimates and create accurate fertility maps. Furthermore, we are working on redesigning the user interfaces (UIs) of both Agroview and Agrosense based on feedback from stakeholders (aka., equipment operators, farm managers, growers, agricultural consultants, equipment manufacturers, and industry professionals). This process will help us fine-tune the UIs, ultimately leading to a more user-friendly and efficient experience for growers and other users. We expect to release the new UIs by the end of this project. Objective #2: Enhance, field-evaluate, and finalize the development of the Smart Sprayer and Smart Fertilizer Spreader We have developed a feature on Agroview to create application maps, based on user inputs, from the fertility maps. We have also redesigned the Agrosense sensing system and the control of the fertilizer to make it more robust and accurate. We are working on developing a code for Agrosense to read the application maps from Agroview. This will be ready and field-tested by the end of this project.

Publications

  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Lacerda C., and Neto A.D.C., Ampatzidis Y., 2024. Agroview: enhance satellite imagery using super-resolution and generative AI for precision management in specialty crops. AgEng International Conference of EurAgEng, Agricultural Engineering Challenges in Existing and New Agrosystems, Athens, Greece, July 1-4, 2024.
  • Type: Conference Papers and Presentations Status: Published Year Published: 2024 Citation: Zhou C., Ampatzidis Y., Guan H., and Neto A.D.C., Kunwar S., Batuman O., 2024. Agrosense: AI-enabled sensing for precision management of tree crops (poster). 16th International Conference on Precision Agriculture (ICPA), Manhattan, Kansas, USA, July 21-24.