Project ProtoStar is a new initiative bringing together universities, research organizations, industry partners, and the open science community to make datasets, models, and code from publicly funded research easier to discover, reuse, and apply—especially for AI‑driven research.
Who is involved
Founding partners include Carnegie Mellon University, Purdue University, Georgia Institute of Technology, Arizona State University, and The George Washington University, participating through the Open Forum for AI (OFAI) as the university coalition partner; 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 via a memorandum of understanding supporting the release of suitable scientific datasets, models, and code in AI‑ready formats on open infrastructure. Organizations interested in joining can contact protostar@orcaopen.org.
Why the initiative was launched
The U.S. federal government invests approximately $190 billion annually in research and development, and federal policies increasingly expect research results to be shared publicly. Agencies such as the National Institutes of Health (NIH) and the National Science Foundation (NSF) have strengthened public‑access requirements, and researchers are required to plan for sharing underlying data. Hugging Face is one of the ten generalist repositories listed by NIH for sharing research data.
But public availability does not equal usability. Scientific datasets, models, and code are often dispersed across incompatible systems, stored in static or legacy formats, or released without documentation, metadata, and provenance that enable understanding and reuse. Those limitations make it harder for researchers—and for AI systems—to discover, interpret, and combine scientific resources across disciplines.
Project ProtoStar aims to narrow this gap by encouraging systematic sharing of publicly funded scientific artifacts in AI‑ready formats, supported by open infrastructure and practical publishing guidance. The initiative builds on prior 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 public‑access expectations while increasing the scientific value of released outputs.
Demand for reusable scientific resources
Demand already exists: 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. That number approaches 29,000 when medicine and mathematics are included, alongside a comparable number of scientific models.
Sharing is accelerating: researchers uploaded more than three times as many new science datasets to Hugging Face in the first nine months of 2026 as in the same period of 2025, and growth remained substantial even after excluding the largest individual uploader.
When organizations share research in reusable formats, the community can build on it at scale. The 18 public repositories maintained by NIH's National Center for Biotechnology Information (NCBI) on Hugging Face have received more than 11 million downloads in total, while NASA IMPACT's 27 repositories have received approximately 770,000 downloads, as of October 26, 2026. Documentation matters: 90% of dataset downloads on Hugging Face are from fully documented datasets.
Open genomic data has contributed to downstream scientific advances; Hugging Face researchers have built on openly available resources, including releases from HuggingFaceBio such as Carbon‑A and Carbon‑X. By releasing models, datasets, and tools openly, these efforts show how existing scientific resources can support new, shareable research.
Carnegie Mellon University has taken an early institutional step: its Open Source Programs Office is encouraging researchers to share suitable datasets, AI models, and other research outputs on Hugging Face, while recognizing the continuing importance of domain‑specific repositories.
Open science challenges: objectives
Making artifacts AI‑ready is the first step. ProtoStar also aims to demonstrate what researchers, developers, and the broader scientific community can build when those resources are reusable. The central question for ProtoStar is: can publicly funded scientific datasets, models, and code released in AI‑ready formats be readily discovered, reused, combined, and extended to accelerate discovery and public benefit?
To explore this, ProtoStar will organize a series of open challenges. Each challenge will be hosted by a participating research institution and built around scientific resources that the host has released openly. The challenges have two complementary goals:
- Demonstrate scientific reuse: produce concrete examples of researchers and developers building on publicly funded artifacts, including applications that combine resources across institutions or disciplines.
- Learn what makes reuse possible: identify technical barriers, documentation gaps, interoperability challenges, and successful practices to inform future research infrastructure, funding decisions, and public policy.
Projects and lessons from the challenges will create a practical evidence base for making publicly funded science more accessible and useful.
First challenge: Climate Risk Intelligence
- Host: Carnegie Mellon University
- Dates: November 2–13, 2026
The inaugural ProtoStar challenge asks participants to test whether AI can improve predictions of urban climate conditions using historical weather data from 210 U.S. cities. Based on Carnegie Mellon University research, teams will train on climate records through 2018 and predict actual conditions from 2019 to 2026, enabling evaluation against observed weather rather than hypothetical futures.
Participants will forecast four indicators relevant to urban planning: average annual temperature, total annual precipitation, the hottest day of the year, and cooling degree days (a measure of air‑conditioning demand). Predictions will be compared against simple statistical baselines, including historical averages and linear trends, to assess whether AI approaches offer meaningful accuracy improvements.
In addition to forecasting, teams will propose a short climate adaptation policy idea informed by their data work. This encourages connecting scientific analysis to 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 their outputs. Forecasting performance will be evaluated via a public leaderboard that measures prediction errors against established baselines; policy ideas will receive separate consideration. Final forecasting results will be published in early 2027, after complete 2026 weather records are available.
The CMU challenge is the first in the series. Other participating institutions will sequentially host their own challenges, each highlighting different publicly funded scientific resources and reuse opportunities. Future challenges will be announced on the Project ProtoStar page on Hugging Face.
How to join
Project ProtoStar invites academic researchers, students, independent developers, and companies building on open scientific infrastructure to participate. Universities, federal research agencies, national laboratories, nonprofits, and other research organizations interested in making their resources more accessible are also welcome.
To explore participation or collaboration, contact protostar@orcaopen.org.
The initiative aims to help publicly funded scientific work reach the people and tools that can build on it while producing practical evidence about the infrastructure and practices needed to support AI‑ready open science at scale.
