Scientific Machine Learning for Complex Systems | DE FOA 0002958

Opportunity Information: Apply for DE FOA 0002958

The Department of Energy (DOE) Office of Science, through the Advanced Scientific Computing Research (ASCR) program, is soliciting research proposals under the grant opportunity "Scientific Machine Learning for Complex Systems" (Funding Opportunity Number DE-FOA-0002958). The core aim is to fund potentially high-impact research that advances scientific machine learning (SciML) and artificial intelligence (AI) methods specifically for predictive modeling, simulation, and analysis of complex physical systems and processes. The opportunity is framed around the idea that modern science increasingly depends on high-performance computing (HPC), sophisticated numerical models and algorithms, large experimental and observational datasets, and automation to speed up discovery and innovation. In practical terms, DOE is looking for work that helps integrate AI/ML with the kinds of computational workflows used across DOE-relevant science, rather than generic ML that is disconnected from domain physics or simulation realities.

The scientific motivation comes from DOE-wide planning and community reports that identify AI and ML as key tools for science, energy, and national security missions. The announcement explicitly points to application areas spanning materials science, environmental and life sciences, and major physics domains such as high-energy physics, nuclear physics, and plasma physics, along with cross-cutting DOE initiatives like the Energy Earthshots. While the FOA references broad SciML foundations (including domain awareness, interpretability, robustness, and scalability), the primary research emphasis here is narrower: it targets the capability direction focused on Machine Learning-Enhanced Modeling and Simulation (identified as Priority Research Direction #5 in a 2018 DOE Basic Research Needs report) along with uncertainty quantification. That focus signals an interest in methods that do more than analyze data; DOE wants approaches that improve or accelerate simulation, couple ML models with established numerical solvers, and provide credible measures of uncertainty so predictions can be trusted and used in decision-making.

On eligibility, the opportunity is broadly open ("Eligible Applicants: Unrestricted") with one major statutory exception: 501(c)(4) nonprofit organizations that engaged in lobbying activities after December 31, 1995 are not eligible. National laboratories and federally funded research and development centers (FFRDCs) are allowed to participate, either as lead applicants or as team members/subrecipients, but they must follow special DOE administrative pathways. DOE/NNSA national labs, if funded, receive support via the DOE Field-Work Proposal System and conduct the work under their existing DOE contracts, rather than under the standard administrative terms that apply to typical grant recipients. Non-DOE/NNSA FFRDCs are similarly eligible, but awards flow through interagency agreements to the FFRDCs sponsoring federal agencies. By contrast, other federal agencies are explicitly not eligible to apply and cannot be included as subrecipients on another organization’s application.

DOE also highlights a policy priority around broadening participation and investigator diversity. The FOA notes DOE’s continued use of program policy factors in award selection and encourages proposals that are led by, or include partners from, EPSCoR states that are underrepresented in the ASCR portfolio. It also encourages applications led by individuals from groups historically underrepresented in STEM. This is presented as a deliberate effort to diversify both the institutions and people supported by ASCR in this research area.

Key administrative details from the source listing include: the funding instrument is a discretionary grant under CFDA 81.049; the opportunity was created January 24, 2023; and the original closing date was April 19, 2023. The listed award ceiling is $1,200,000. The summary information does not specify the exact number of expected awards in the excerpt provided, but the ceiling indicates DOE anticipates funding projects at a scale consistent with multi-year research efforts or substantial single projects, especially those involving integrated simulation, ML, and uncertainty quantification.

Overall, this FOA is best read as a targeted call for research that pushes the state of the art in SciML methods that work hand-in-hand with computational science: physics- and domain-aware learning, ML that enhances modeling and simulation pipelines, and rigorous uncertainty quantification so AI-enabled predictions for complex systems are robust, scalable on advanced computing platforms, and scientifically defensible across DOE mission-relevant disciplines.

• The Office of Science in the science and technology and other research and development sector is offering a public funding opportunity titled "Scientific Machine Learning for Complex Systems" 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 2023-01-24.
• Applicants must submit their applications by 2023-04-19. (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 $1,200,000.00 in funding.
• Eligible applicants include: Unrestricted.

Apply for DE FOA 0002958

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