COMMERCIALIZATION ENGINE

Data Challenges

Real problems from growers, named by industry, solved with shared data and compute — and piloted in the field.

ACTIVE

Orchard Stress

In Progress


COMPLETED

Frost Risk

November 2025


CURRENT CHALLENGE

ORCHARD STRESS MAPPING

Participants use free satellite imagery to detect and map underperforming zones in Central Valley orchards. The four-week sprint kicks off shortly — registration is open now, with workshops and judging seats also available.


TIMELINE

4 week sprint

COST

Free

CASH PRIZE

$1,500+

GET INVOLVED

Choose your role in the challenge.

There are three ways to take part. Pick the one that matches what you bring.

FOR TEAMS & INDIVIDUALS

Compete in the sprint.

Undergrad teams, graduate researchers, faculty, and independent builders. Use real Central Valley data and the AI Supercomputing Center's compute to develop a working solution over four weeks.


ALSO FOR PARTICAPANTS

FOR EXPERTS

Review submissions.

Industry leaders, technical experts, and academic researchers help evaluate team submissions on accuracy, creativity, and real-world applicability for Central Valley growers.


FOR GROWERS & COMPANIES

Solve your problem.

If your operation has a recurring problem that better data or AI could solve, propose it for the next sprint. F3i scopes the most promising challenges into structured data projects.


HOW THE DATA CHALLENGE WORKS

Three stages, one cycle, built for deployment.

Each challenge follows the same arc — from a problem named by industry, through model development on shared compute, to validated pilots in the field.

01 / IDENTIFY

A real problem, named by industry.

Undergrad teams, graduate researchers, faculty, and independent builders. Use real Central Valley data and the AI Supercomputing Center's compute to develop a working solution over four weeks.

02 / BUILD

Teams develop solutions on shared compute.

Interdisciplinary teams — undergrads, graduate students, faculty, independent researchers — build models using real-world data and the Supercomputing Center's infrastructure.

03 / DEPLOY

Top solutions move into the field.

Validated submissions are piloted with growers in the Central Valley. Leading teams may be considered for commercialization cohorts and the Innovation Grant program.

WHY WE BUILT THE SUPERCOMPUTING CENTER

Most agricultural problems aren't solved by code alone.

A small ag-tech startup can't afford a GPU cluster. A graduate student can't license the datasets. A grower doesn't have the bandwidth to run a research project. The Supercomputing Center exists so a problem named on Monday by a Madera County orchardist can be modeled on Tuesday by a CSU Fresno team — on infrastructure that would otherwise cost six figures to access.

INFRASTRUCTURE

San Diego Supercomputer Center at UC San Diego provides cyberinfrastructure and AI systems architecture.


HOST

California State University, Fresno hosts and operates the physical compute environment.


LEADERSHIP

F3 Innovate leads the center, its programs, and workforce development partnerships.


PRIORITY ACCESS

California-based startups, researchers, and educators in agriculture and biotech — with priority for Central Valley users.

ARCHIVE

01 — FROST RISK · 2025

Previous Challenges

NOVEMBER 2025

Forecasting Frost Before It Hits

Frost is one of California agriculture's most damaging weather risks. University teams developed predictive models using localized microclimate data — supported by the San Diego Supercomputing Center and the National Data Platform.


WINNING TEAMS

1st

FrostByte

$1,500

San Jose State University


2nd

$750

Tyler Aziz

UC Davis + Cal Poly SLO


3rd

$400

AgriFrost AI

Independent Contributor

BUILT WITH

A multi-institution partnership spanning advanced computing, biological modeling, and workforce development across California.

San Diego Supercomputer Center

UC SAN DIEGO

CSU Fresno

HOST INSTITUTION

National Data Platform

DATA INFRASTRUCTURE