Loading Events

« All Events

  • This event has passed.

AI-Ag TrAC-a-thon – Spring 2024

March 2 @ 10:00 am 5:00 pm

AI-Ag TrAC-a-thon 2024 is our very first Translational-AI hackathon

About

TrAC’s Hackathon with a theme of Smart Digital Agriculture

  • A bottom-up approach to innovation by bringing together farmers and student teams to develop AI tools
  • Showcase how AI can be scale-agnostic and beneficial to smaller-scale farmers to participate in markets
  • Revolutionize AI applications across farm operations intelligence
  • Immersive sessions to gain insights into ground-reality of farmers and find the meeting point of bottom-up community needs with top-down emerging technologies

Who can participate?

Students: Graduate and undergraduate students with science and engineering skills and agricultural interests or experience are encouraged to participate. You have an opportunity to explore what technologies can do for agriculture as well as a unique bottom-up approach to understanding beneficiaries and communities (Learn more at www.subsistencemarketplaces.org and www.marketplaceliteracy.org). 

Farmers: Farmers with an interest in newer technologies

Why should you participate?

Students: Join us to level up your AI and digital agriculture skillset and for the opportunity to win cash prizes (the first, second, and third places) sponsored by AIIRA, TrAC, Bayer, John Deere, Principal, and Corteva.

Farmers: Learn how farmers can have a voice in innovation. Work with ISU student teams to develop ag technology tools. All farmers will be compensated for their participation

Timeline

Feb 15, 2024 Feb 20, 2024
Feb 20 – 27, 2024

Workshops (problems, datasets, computing resources, and other training from organizing committee)

Feb 27, 2024

Initial solution submission

March 2, 2024

AI-Ag TrACathon event (presentation/demo, immersive activity)

Eligibility

Students enrolled at all US-based universities/colleges from all years of undergraduate/graduate school are eligible to participate; priorities given to students affiliated with TrAC/AIIRA; Students with a range of science and engineering skills and agricultural interest or experience are encouraged to apply.

The Challenges

Computer vision for plant diseases/pests detection: Plant diseases and pests are critically important factors determining the annual yield and the quality of plants. The detection of plant diseases and pests can be implemented by using digital image processing. However, in real complex natural environment, detecting plant diseases and pests efficiently still remains a challenging task due to small difference between the lesion area and the background, low contrast, large variations in the scale of the lesion area and various types, and a lot of noise in the lesion image. Therefore, developing more effective and efficient machine vision-based plant diseases/pests detection technologies is significantly vital for farming.

AI for Ag weather/crop yield prediction: A major challenge for food security worldwide is the large interannual variability of crop yield. Accurate foresights of crop yield in advance can have a considerable impact on decisions regarding crop selection, rotational location in crop rotations, agrotechnical methods employed, and short-term land management and long-term land use planning. The ability to promptly and reliably forecast crop yield is an important aspect of regional and global food security. Additionally, accurate crop yield prediction models can help farmers to decide on when to grow and what to grow. However, precisely identifying the relationship between crop yield and many other variables is still quite challenging. Agriculture weather is critical for securing the yield of most field crops. Especially when severe weather strikes unexpectedly, this might result in heavy losses. Hence, developing accurate ag weather prediction models to provide weather information is significantly beneficial for crop production.

Note: Each team is only allowed to select one challenge problem to work on: 1) computer vision for plant diseases detection; 2) computer vision for plant pests detection; 3) AI for Ag weather prediction (precipitation, soil moisture, drought, snow cover, etc.); 4) AI for crop yield prediction.

Frequently Asked Questions

Please let us know if you have any questions or comments on registration, participation, and event through email to zhjiang@iastate.edu.

Who will I be working with?

Please assemble your own team (list teammate’s names on the registration form). Teams will have between 3-4 members. You can also fly solo if you feel you can and would like to solve the problem by yourself.

What resources will be provided?

In order to help participating teams develop solutions, we will first host a few workshops to provide guidance for students from Feb 15 – Feb 27. In these workshops, we will help the students with problem and dataset selection and explain technical details about how to access some available computing resources. Additionally, the “roboflow” platform specifically for computer-vision-based solutions will be accessible if necessary. Teams are welcome to use their own resources to develop fast and simple prototype solutions. While in case some teams prefer developing large models, these computing resources may help. Please contact baditya@iastate.edu for additional technical support on the computing resources.

Where is this happening?

This is an on-campus event for students to work and present, have snacks and meet with local farmers and domain experts from the industry. The location is the Student Innovation Center at Iowa State University.

What background do I need to have?

Any student with some programming experience and a willingness to learn in teams is welcome to participate. Past experience with any of the programming languages (like Python, R, C/C++, etc.) and/or the use of AI/ML packages (like TensorFlow/PyTorch/Scikit-learn) will help.

What will happen during TrAC-a-thon?

Please note that the AI-Ag TrACathon is not a traditional hackathon that only gives 24-48 hours to solve challenges. Instead, participants can have sufficient time to work on the challenge problems. Starting morning on Saturday, March 2nd, with a welcoming overview, the immersive activities will take place from 9:00 am – 12:00 pm with showing simulations and interviewing with farmers. The final review round of solutions will start from 2:00 pm – 5:00 pm. Participants will present their solutions to a judge team consisting of local farmers and domain experts. Solutions can always be updated before the presentation. An awards ceremony will be held on Saturday evening (more detail will be announced)!

What is the review criteria?

Each solution will be reviewed and evaluated in detail by a judge team consisting of local farmers and domain experts. Specifically, local farmers will evaluate the solutions based on the feasibility, practicality, efficiency, cost and how many training efforts required for farmers to use. While domain experts will assess the solutions in terms of technical soundness on workflow, dataflow, model/algorithm development, front/backend design, etc., if applicable. Please note that each solution may not necessarily comprise all components mentioned and that they will be reviewed case by case. Showcasing format can be presentation and/or real demo. The potential of easy deployment of the solution to edging devices is significantly valued.

What is to be expected in the solution?

For example, if the team called “Epsilon” selected the problem “computer vision for plant disease detection” to solve, then the team could develop a computer vision-based solution using some shallow or deep learning models to detect plant diseases. The relevant accuracy metrics, such as the confusion matrix, can be used to show the capability of the model. Baseline comparison is a plus to the solution. To demo the results, the team can use the terminal, Streamlit, etc, whatever works for them well. If the team is willing to develop an API to showcase, that would be a plus, too. The entire App development is not required here to avoid the additional UI/UX design and testing. However, some work combining the front and back ends could make the prototype solution complete. Please note that a more comprehensive and thorough design for the solution may receive a higher evaluation score. E.g., the solution with API design could typically outweigh the solution without API design. If the team focuses more on the backend, then a baseline comparison is typically needed. If the team prefers the more balanced front/backend design, then baseline comparison may not be needed.

Who owns the IP?

The TrAC and AIIRA do not hold any rights to the IP. Please consult with your university for individual university policies and for any related IP questions.

Tentative Event Schedule

Date: March 2, 2024

Location: Student Innovation Center at Iowa State University

Time: 9:00 AM – 5:00 PM

Timeline:

9 AM – 10 AM

Welcome and Introduction

10 AM – 12 Noon

Immersive activities (interview with
farmers and marketing simulation)

12 Noon – 2 PM

Lunch and Networking

12 Noon – 2 PM

Solution Presentations and Review

2 PM – 5 PM

Awards Ceremony and Concluding Remarks


Registration

TrACathon Registration Form

Member 1 Name
Captain’s name
Member 2 Name
Member 3 Name
Member 4 Name
Member 5 Name
Only put the email address of the team captain

Faculty Organizers

Faculty members helping us in organizing the TrAC-a-thon.

Priyanka Jayashakar headshot

Priyanka Jayashankar

Madhu Viswanathan

Ronald Duncan