New AI Platform Transforms Research Data Management for Scientists

Over 90% of scientific data remains lost or unused, but a new platform aims to change that. Frontiers, an open-science publisher, has launched its AI-powered FAIR2 Data Management system, designed to enhance the usability, verification, and citation of research data. This innovative approach supports scientists in responsibly sharing their findings, ultimately accelerating global scientific discovery.

For decades, much of the scientific data produced has been inaccessible, often stored on hard drives or hidden in lab notebooks. Data experts estimate that out of every 100 datasets created, approximately 80 never leave the laboratory, 20 are shared but seldom reused, and fewer than two meet the FAIR (Findable, Accessible, Interoperable, Reusable) standards. Typically, only one dataset leads to new discoveries.

A Revolutionary Solution with AI Integration

The FAIR2 Data Management system integrates curation, compliance checks, peer review, and interactive data visualization into a single streamlined process. As a result, it ensures that research data is preserved, recognized, and reused effectively.

“Ninety percent of science vanishes into the void,” stated Dr. Kamila Markram, co-founder and CEO of Frontiers. “With FAIR2, no dataset needs to be lost again — every contribution can fuel progress and earn the credit it deserves.”

At the heart of the FAIR2 platform is the AI Data Steward, powered by Senscience, a venture of Frontiers. This advanced system dramatically reduces the time required for data management tasks, completing in minutes what previously took months. Researchers using FAIR2 receive four key outputs: a certified Data Package, a peer-reviewed and citable Data Article, an Interactive Data Portal with visual tools and AI chat, and a FAIR2 Certificate ensuring quality and compliance.

Extending FAIR Principles into Practice

FAIR2 extends the principles of FAIR into an AI-compatible model, enhancing the responsible reuse of data by both humans and machines. This launch arrives at a crucial juncture as scientific output continues to grow and artificial intelligence increasingly reshapes the research landscape.

According to Dr. Sean Hill, CEO of Senscience, “Science invests billions in generating data, but most of it is lost — and researchers rarely get credit. FAIR2 changes that by making every dataset reusable, recognized, and rewarded.”

FAIR2 has already began pilot projects demonstrating its capabilities. Notable examples include:

– **SARS-CoV-2 Variant Properties**: A dataset featuring 3,800 spike protein variants that connects AlphaFold2 predictions with laboratory data, contributing to pandemic preparedness.
– **Preclinical Brain Injury MRI**: A collection of 343 harmonized MRI scans that support reproducible neuroscience research.
– **Environmental Pressure Indicators (1990–2050)**: A comprehensive dataset offering insights for climate policy across 43 countries.
– **Indo-Pacific Atoll Biodiversity**: Data from 280 atolls that integrates biodiversity, climate, and human-use metrics to inform conservation efforts.

Researchers involved in these pilot projects have noted that FAIR2 not only preserves data but also enhances understanding and collaboration among scientists. This initiative is poised to accelerate progress across various disciplines.

The launch of FAIR2 represents a significant step toward an open, intelligent, and interconnected scientific ecosystem. As Dr. Markram emphasizes, “This is how we unleash science,” ensuring that no discovery goes to waste and that every dataset provides value to the global research community.