Scientific Machine Learning and Artificial Intelligence: Uncertainty Quantification | DE FOA 0002122

Opportunity Information: Apply for DE FOA 0002122

The Department of Energy (DOE) Office of Science, through its Artificial Intelligence (AI) Program and the Advanced Scientific Computing Research (ASCR) program, offered this funding opportunity (DE-FOA-0002122) to push forward scientific machine learning (SciML) and AI in ways that directly support DOE's mission and the broader public interest. The call is framed around the federal priority of maintaining U.S. leadership in AI, but the practical goal is very specific: improve how scientists can trust, validate, and use AI-driven predictions in real scientific workflows that rely on experimental data from DOE user facilities and large-scale scientific simulations.

The central theme of the opportunity is Uncertainty Quantification (UQ) for AI validation and prediction. In plain terms, the DOE is looking for research that makes AI models not just accurate, but reliably honest about what they know and do not know. A major open problem in SciML is how to produce predictions that remain dependable when models are used outside the narrow conditions they were trained on (for example, extrapolating to new physical regimes or experimental setups). The FOA emphasizes that better predictive performance and credibility can come from explicitly accounting for uncertainties in inputs (such as noisy measurements, uncertain boundary conditions, or imperfect simulation parameters) and from understanding model discrepancy (the gap between what a model assumes and what the real system does).

This opportunity highlights the need for foundational advances that strengthen the mathematical and statistical basis of AI validation. The DOE is signaling interest in methods that connect machine learning with established scientific and engineering principles, rather than treating AI as a black box. That includes developing UQ approaches that are grounded in mathematics and statistics, can handle scientific constraints, and can generate uncertainty estimates that are meaningful for decision-making in research settings. The FOA also notes that there is a large and mature body of statistical literature that can be leveraged to improve validation practices, suggesting that strong proposals could bridge modern ML with rigorous statistical model checking, calibration, error decomposition, and principled uncertainty estimation.

From a funding and administrative standpoint, this was a discretionary grant opportunity in the Science and Technology / R&D category (CFDA 81.049). Eligibility was listed as unrestricted, meaning a wide range of applicant types could apply, subject to any additional eligibility clarifications in the full announcement text. The opportunity was created on April 16, 2019, with an original closing date of May 31, 2019. DOE anticipated making around 6 awards, with an award ceiling of $150,000.

Overall, the FOA is essentially a targeted research push to make scientific AI more dependable by advancing the theory and practice of uncertainty quantification, especially for validation and for predictions that go beyond the training data. The expected payoff is stronger mathematical and scientific computing foundations that allow SciML and AI to accelerate discovery while providing uncertainty information that scientists can actually rely on when interpreting results, comparing to experiments, and guiding future studies.

• The Department of Energy - Office of Science, Office of Science in the science and technology and other research and development sector is offering a public funding opportunity titled "Scientific Machine Learning and Artificial Intelligence: Uncertainty Quantification" and is now available to receive applicants.
• Interested and eligible applicants and submit their applications by referencing the CFDA number(s): 81.049.
• This funding opportunity was created on Apr 16, 2019.
• Applicants must submit their applications by May 31, 2019. (Agency may still review applications by suitable applicants for the remaining/unused allocated funding in 2026.)
• Each selected applicant is eligible to receive up to $150,000.00 in funding.
• The number of recipients for this funding is limited to 6 candidate(s).
• Eligible applicants include: Unrestricted (i.e., open to any type of entity above), subject to any clarification in text field entitled Additional Information on Eligibility.

Apply for DE FOA 0002122

[Watch] Creating a grant proposal using the step-by-step wizard inside the applicant portal: