F3i's Frost Risk Data Challenge

Natasha Biasell
Jan 13
2 min read

Updated: Jan 20

F3 Innovate (F3i.org) announced the results of its Frost Risk Data Challenge. The competition invited university teams from various disciplines to develop innovative data analytics models to predict frost events. The challenge used a regional dataset of microclimate weather data from California, with methods that can be applied to other regions.

Frost is among the most damaging weather risks to California agriculture and the U.S., where economic losses from frost exceed those from all other weather-related phenomena.

Traditionally, frost protection has relied on manual readings and growers’ local knowledge. Machine learning and spatial modeling offer opportunities for earlier and more reliable frost warnings.

The challenge gave participants an opportunity to develop models that can predict frost events using real-world datasets and supercomputing resources to develop, train, and test their models. Teams were based on accuracy, creativity, and solution effectiveness.

Congratulations to the top three teams:

🥇 1st Place | FrostByte

Participants: Meera Bhaskbarbhai Vyas, Devarsh Shroff, Rishil Patel

San José State University

🥈 2nd Place | Tyler & Aziz

Participants: Tyler Roberts, Aziz Saries

California Polytechnic State University-San Luis Obispo

& University of California, Davis

🥉 3rd Place | AgriFrost AI

Participant: Zhengkun Li

Independent Contributor; Alumni, University of Florida

“We're excited to host this inaugural data challenge,” said Ryan Dinubilo, F3i Director of Innovation. "It provides a platform for students to showcase their talents, while fostering collaboration and innovation within the data analytics community.”

The challenge was made possible through the support of the University of California San Diego Supercomputer Center and the National Data Platform.

Judging Panel for the Frost Risk Data Challenge:

Ryan Dinubilo | Director of Innovation, F3 Innovate

Leads F3i’s innovation strategy and program development, including data challenges that advance agtech, commercialization, and applied research.

Dr. Neil White | Queensland Department of Primary Industries

Brings over 20 years of experience in climate research, crop modeling, and data science, with a focus on supply chain and shelf-life prediction.

Chase Barclay | Statistician & Data Scientist

Specializing in agricultural risk modeling and applied data science, with experience in crop insurance analytics, actuarial methods, and large-scale statistical systems.