Project ProtoStar Aims to Make Publicly Funded Science AI-Ready

Project ProtoStar Aims to Make Publicly Funded Science AI-Ready

Published October 29, 2026

Project ProtoStar is an initiative designed to make publicly funded scientific research easier to discover, reuse, and extend. It brings together universities, research organizations, industry, and the open science community to encourage the release of scientific datasets, models, and code in AI-ready, machine-actionable formats.

The initiative will also demonstrate how these resources can be used through a series of open challenges. The first challenge, hosted by Carnegie Mellon University, opens on November 2, 2026.

Founding partners include Carnegie Mellon University, Purdue University, Georgia Institute of Technology, Arizona State University, and The George Washington University. These universities participate through the Open Forum for AI (OFAI), the university coalition partner. Other founding partners are the Astera Institute, Hugging Face as the infrastructure partner, and the Open Research Community Accelerator (ORCA), through its Fostering Innovation Network, as the community partner.

Participating organizations formalize their involvement through a memorandum of understanding supporting the release of suitable scientific datasets, models, and code in AI-ready formats on open infrastructure.

Making Publicly Funded Science More Usable

The U.S. federal government invests approximately $190 billion annually in research and development, while federal policy increasingly expects the results of that work to be shared publicly. Agencies such as the National Institutes of Health (NIH) and National Science Foundation (NSF) have strengthened public-access requirements for funded research publications and require researchers to plan for sharing the underlying data.

Hugging Face is among the ten generalist repositories listed by NIH for sharing research data.

However, public availability does not automatically make research usable. Scientific datasets, models, and code are often distributed across incompatible systems, stored in static or legacy formats, or released without the documentation, metadata, and provenance needed for interpretation and reuse. These problems make it more difficult for researchers to build on previous work and for AI systems to discover, interpret, and combine resources across disciplines.

Project ProtoStar aims to address this gap by encouraging systematic sharing of publicly funded scientific artifacts in AI-ready formats. The initiative combines open infrastructure with practical publishing guidance and builds on efforts such as the National Cancer Institute's AI Data-Readiness Challenge and Hugging Face's guide to preparing public-sector data for AI.

The goal is to help research organizations meet existing public-access expectations while increasing the scientific value of the work they release.

Demand for reusable scientific resources is already visible. Hugging Face hosts more than 17,000 public datasets across core scientific domains, including genomics, protein design, chemistry, weather and climate, Earth observation, and astronomy. The number approaches 29,000 when medicine and mathematics are included, alongside a comparable number of scientific models.

Sharing is also accelerating. Researchers shared more than three times as many new science datasets on Hugging Face in the first nine months of 2026 as during the same period in 2025, with substantial growth remaining after excluding the largest individual uploader.

When organizations publish research in reusable formats, other researchers can build on it at scale. The 18 public repositories maintained by NIH's National Center for Biotechnology Information (NCBI) on Hugging Face have collectively received more than 11 million downloads. NASA IMPACT's 27 repositories have received approximately 770,000 downloads, as of October 26, 2026.

Documentation is also significant. 90% of dataset downloads on Hugging Face are from fully documented datasets, indicating the role that clear documentation plays in enabling reuse.

Public genomic data has contributed to research in areas such as biology. Researchers at Hugging Face have built on openly available scientific resources, including releases from HuggingFaceBio such as Carbon-A and Carbon-X. Releasing the resulting models, datasets, and tools openly provides examples of how existing scientific resources can support further research.

Project ProtoStar seeks to encourage this type of activity through institutional policies adopted by participating organizations. As an early example, Carnegie Mellon University's Open Source Programs Office is encouraging researchers to share suitable datasets, AI models, and other research outputs on Hugging Face, while continuing to recognize the importance of domain-specific repositories.

Open Science Challenges

Making scientific artifacts AI-ready is an important first step. ProtoStar will also examine what researchers, developers, and the wider scientific community can accomplish when those resources are available for reuse.

A central question is whether publicly funded scientific datasets, models, and code released in AI-ready formats can be readily discovered, reused, combined, and extended to accelerate scientific discovery and public benefit.

ProtoStar will organize a series of open challenges. Each challenge will be hosted by a participating research institution and will focus on scientific resources that the institution has released openly.

The challenges have two main goals:

• Demonstrate scientific reuse: Produce concrete examples of researchers and developers building on publicly funded scientific artifacts, including applications that combine resources from different institutions or disciplines.
• Identify conditions for reuse: Examine technical barriers, documentation gaps, interoperability problems, and successful practices that could inform future research infrastructure, funding decisions, and public policy.

The projects and lessons from these challenges will contribute to a practical evidence base for making publicly funded science more accessible and useful.

The First Challenge: Climate Risk Intelligence

Hosted by Carnegie Mellon University | November 2-13, 2026

The first ProtoStar challenge will test whether AI can improve predictions of urban climate conditions using historical weather data from 210 U.S. cities. Based on research by Carnegie Mellon University, teams will use climate records through 2018 to predict actual conditions from 2019 through 2026. This allows the models to be evaluated against observed weather rather than hypothetical future scenarios.

Participants will forecast four indicators relevant to urban planning:

• Average annual temperature
• Total annual precipitation
• The hottest day of the year
• Cooling degree days, a measure of air-conditioning demand

The predictions will be compared with simple statistical baselines, including historical averages and linear trends, to determine whether AI approaches provide meaningful improvements in accuracy.

Teams will also propose a short climate adaptation policy idea based on what they learn from the data. This part of the challenge connects scientific analysis with practical public benefits, such as preparing cities for extreme heat, improving energy efficiency, or managing changing precipitation patterns.

Teams will submit a Hugging Face dataset containing these outputs. Forecasting performance will be measured through a public leaderboard that compares prediction errors with the established baselines. Policy ideas will be considered separately.

Final forecasting results will be published in early 2027, after complete 2026 weather records become available.

Explore the CMU Climate Risk Intelligence Challenge

The Carnegie Mellon challenge is the first in the series. Other participating institutions will host challenges sequentially, each focused on different publicly funded scientific resources and reuse opportunities. Future challenges will be announced through the Project ProtoStar page on Hugging Face.

Participating in Project ProtoStar

Project ProtoStar is intended for academic researchers, students, independent developers, and companies building on open scientific infrastructure. Universities, federal research agencies, national laboratories, nonprofits, and other research organizations can also participate by making scientific resources more accessible and reusable.

Organizations interested in participation or collaboration can contact protostar@orcaopen.org.

The initiative is intended to help publicly funded scientific work reach the people and tools that can build on it, while generating practical evidence about the infrastructure and practices required to support open science at scale.