AI Unlocks Vast Reservoirs of Underutilized Research Data

Research institutions often find themselves sitting on vast troves of data that remain underutilized, with estimates indicating that around 80 out of every 100 datasets generated never leave the laboratory. In an effort to address this issue, the open-science publisher Frontiers has launched an innovative solution: the FAIR2 Data Management system. This AI-driven approach aims to ensure that valuable scientific data becomes reusable, verifiable, and citable.

FAIR2 combines multiple essential processes—curation, compliance checks, AI-ready formatting, peer review, and interactive visualizations—into a single platform. This comprehensive framework is designed to facilitate responsible data sharing among researchers while also enhancing their visibility and recognition within the scientific community.

Enhancing Data Accessibility

The FAIR2 initiative builds on the established FAIR principles, which stand for Findable, Accessible, Interoperable, and Reusable. These principles were first articulated in a March 2016 paper published in the journal Scientific Data, and received endorsement from G20 leaders at the G20 Hangzhou summit. By expanding on these principles, FAIR2 guarantees that every dataset is not only compatible with artificial intelligence but also ethically reusable by both humans and machines.

The FAIR2 Data Management system represents a significant advancement, particularly at a time when the volume of research output is increasing rapidly. It transforms high-level principles into actionable infrastructure, thus contributing to measurable impacts in fields such as health, sustainability, and technology.

Research efforts are often hampered by slow progress in areas like cancer treatment, insufficient evidence for climate models, and the challenges of replicating studies. FAIR2 aims to mitigate these obstacles by automating processes that previously required extensive manual labor. Tasks such as organizing and verifying datasets, generating metadata, and creating publishable outputs can now be accomplished in mere minutes with the help of the AI Data Steward, developed by Senscience, the Frontiers venture behind FAIR2.

Researchers who utilize this system receive four integrated outputs. These include a certified Data Package, a peer-reviewed and citable Data Article, an Interactive Data Portal featuring visualizations and AI chat, and a FAIR2 Certificate. Each component is equipped with quality controls and clear summaries, making the data more comprehensible for a broader audience and ensuring cross-discipline compatibility.

Real-World Impact and Reception

One notable example of data harnessed through this system is the “Environmental Pressure Indicators (1990-2050)” dataset. This comprehensive compilation tracks emissions, waste, population, and GDP across 43 countries over six decades, providing a critical foundation for sustainability benchmarking and evidence-based climate policy planning.

The FAIR2 system is designed not only to enhance visibility and accessibility but also to support responsible data reuse by various stakeholders, including scientists, policymakers, and communities. This collaborative approach aims to maximize the societal value derived from scientific investments.

In terms of its reception within the global scientific community, Dr. Ángel Borja, Principal Researcher at AZTI and a member of the Basque Research and Technology Alliance (BRTA), commended the initiative, stating, “I highly recommend using this kind of data curation and publication of articles, because you can generate information very quickly and it’s useful formatting for any end users.”

By facilitating the effective management and dissemination of scientific data, FAIR2 is poised to revolutionize how researchers share their findings and contribute to the advancement of knowledge across various disciplines.