Download PDF Fall Bridge on the Materials Genome Initiative September 29, 2025 Volume 55 Issue 3 The Fall 2025 issue explores the Materials Genome Initiative’s progress and future outlook, showcasing advances in autonomous experimentation, sustainable polymers, next-generation batteries, and the broader role of AI in engineering. Accelerating Materials Innovation: Lessons Learned and Opportunities Ahead Tuesday, September 30, 2025 Author: Lisa E. Friedersdorf and James A. Warren A decade of the Materials Genome Initiative: insights gained and the transformative potential of AI and automation. New materials are the foundation of each generation’s ability to enhance prosperity and ensure security. To remain competitive globally, materials research and development (R&D) must be cutting edge, moving the latest innovations from the laboratory to the factory floor—a process that is typically too slow. The Materials Genome Initiative (MGI) was launched more than a decade ago to deploy advanced materials twice as fast and at a fraction of the cost. Here, we review the progress made toward advancing the goals of the MGI and reflect on lessons learned that can inform future efforts. The MGI was established as an interagency initiative under the National Science and Technology Council (NSTC) of the White House Office of Science and Technology Policy. Initial participants included the Department of Commerce (National Institute of Standards and Technology [NIST]), Department of Defense (DOD), Department of Energy (DOE), National Aeronautics and Space Administration (NASA), and National Science Foundation (NSF) (NSTC 2014). As the initiative evolved, additional participation came from the US Patent and Trademark Office, Department of the Interior (US Geological Survey), National Institutes of Health, and US Department of Agriculture (NSTC 2021). The participating agencies have employed both intramural and extramural R&D mechanisms to advance the goals of the MGI. For example, NIST established the Materials Genome Program, a broad effort to support the MGI through its intramural research portfolio and targeted grants (NIST 2025b). NIST has a century-long tradition of disseminating high-quality data for industry use, so it was natural to focus its efforts on the aspects of the MGI where data played a crucial role. NIST defined three core areas in support of the MGI: Data and Model Dissemination Data and Model Quality Data Driven Materials R&D Although the second area was a short leap from NIST’s traditional role, the first—while seemingly technical and straightforward—proved remarkably challenging, touching on deep incentive structures embedded in the research enterprise. Data dissemination is not something researchers naturally do outside the framework of traditional archival publication. The third area foreshadowed the rise of AI approaches to materials R&D by only a few short years and has since become a central focus of the MGI. NSF’s flagship program to support the MGI, Designing Materials to Revolutionize and Engineer our Future (DMREF), began as a small initiative with partners across NSF and a few federal agencies and launching with 14 projects (NSF 2025a). DMREF has since grown to well over 200 active projects and now includes many NSF divisions as well as collaborative efforts with DOD (Air Force Research Laboratory, Office of Naval Research, Army Research Laboratory; Ground Vehicle Systems Center), DOE, and NIST. DMREF has evolved to address advanced materials needs up the technology readiness ladder and across the entire materials spectrum from biomaterials to quantum materials while incorporating interdisciplinary teams of scientists, mathematicians, software developers, and engineers. These teams share the common goal of advancing materials innovation, while leveraging multiple agency perspectives. The program has been instrumental in translating materials discovery into adoption by fostering the transition from fundamental science to applied research. NSF also developed and launched the Materials Innovation Platforms (MIP) to establish, at a larger scale, scientific ecosystems that include in-house research scientists, external users, and other contributors who share tools, codes, samples, data, and knowledge to strengthen collaborations and accelerate the discovery and development of new materials (NSF 2025b). These platforms focus on specific materials domains—currently semiconductors and biomaterials—with the goal of creating a self-sustaining engine of materials discovery and development that can operate nimbly and rapidly in these critical areas. DOD is strategically investing in materials and manufacturing research to ensure the effectiveness and safety of US service members and to advance national defense capabilities across all military domains. A central element of this collaborative effort, which unites the expertise of universities, industries, and the Service Laboratories, is the development of a data-centric materials and manufacturing digital pipeline, along with the workforce to support it. The aim is to enhance the agility of system design, enabling the rapid incorporation of emerging technologies to achieve peak performance against adversary threats while proactively identifying potential material supply risks throughout the acquisition lifecycle. Additionally, this will power affordable sustainment practices that maximize the operational readiness of military forces. Illustrating this effort, the Air Force Research Laboratory is building new research facility capabilities in several key areas. These include autonomous material characterization, fabrication, and synthesis to establish robust data repositories; development of continuously improving processing–structure–performance models using artificial intelligence (AI) and heterogeneous data fusion; robot-human teaming for high-mix, low-volume manufacturing and inspection; and new materials for ultralow-power edge-computing devices. Other efforts supporting the MGI include the Energy Materials Network, established in 2016 by DOE as a community of practice in state-of-the-art materials R&D specifically aimed at advancing critical energy technologies. The network comprises core consortia focused on different high-impact energy technologies, each leveraging world-class capabilities at DOE’s National Laboratories to better integrate all phases of materials R&D, from discovery to scale-up and qualification. It also provides streamlined access to these capabilities for industry and academia to accelerate the energy materials development cycle and enable US manufacturers to deliver innovative, made-in-America products for affordable energy. Outcomes from these efforts and progress toward the goals of the MGI have been detailed in several publications, including results from the NSF-sponsored workshop “Advancing and Accelerating Materials Innovation: New Frontiers for the Materials Genome Initiative” (de Pablo et al. 2019), “The Materials Genome Initiative and the Metals Industry” (Warren 2024), as well as other papers presented in this issue. Case studies in these publications highlight examples of accelerated deployment of new materials, such as new alloys developed in a fraction of the traditional time for use in a US Navy aircraft and in coins produced by the US Mint. Rather than enumerating additional success stories, this article focuses on identifying key lessons learned that can inform future efforts. Lessons Learned The general premise of the MGI is that to accelerate materials discovery, design, manufacture, and deployment—computation, data, and experiment must be brought together in a tightly integrated manner. To enable this integration, the MGI introduced the Materials Innovation Infrastructure (MII), consisting of physical and computational tools and data. Building out the MII allows materials R&D practitioners to design fit-for-purpose materials concurrently with product design, as illustrated above through the efforts of NIST, DOD, DOE, and NSF. The MGI paradigm was heavily informed by concepts such as Integrated Computational Materials Engineering (ICME). While not all MGI efforts are inherently ICME, it is an excellent example of an “MGI approach” that successfully translates basic scientific research into materials engineering. The concept of ICME has served for about two decades as a useful framework for computational materials design. Although the underlying ideas date back further, a report from the National Research Council crystallized the concept for many in the materials R&D community (NRC 2008). That report also provided practical examples and, perhaps most importantly, a detailed discussion of the return on investment that manufacturers could realize by employing ICME to design fit-for-purpose materials for integration into their products. MGI approaches can significantly accelerate the development of new materials, but there are some caveats. The greatest successes to date have come in areas where both the theories of the materials and the software to translate those theories into practical engineering decisions are most developed. For example, in metallic systems, the CALPHAD modeling approach (NIST 2025a) has benefited from 50 years of steady improvement and widespread adoption by industry. Another strategy for realizing success with the MGI approach is to start with a system that is already well understood. Thus, for example, if the desired new material is “close” in composition to an existing material, iterative (physics-informed) methods can usually converge with the new system with relative ease. The general premise of the MGI is that to accelerate materials discovery, design, manufacture, and deployment—computation, data, and experiment must be brought together in a tightly integrated manner. Substantial challenges remain, though, in realizing all the goals of the MGI. If efforts must stray far from current systems, it becomes a much heavier lift to engage in the materials design required for optimal manufacturing insertion. Similarly, if physics-informed models are unavailable or not mature enough for immediate engineering use, as is often the case in polymer systems, the application of ICME or other integrated modeling efforts is severely constrained. Even in metallic systems, where ICME has had the most success, reliance on existing CALPHAD databases—which offer only limited coverage of possible alloys—means that in many cases the necessary data simply are not available. Acquiring experimental data to fill these gaps remains the critical bottleneck to ICME success. Another issue impeding wider application of MGI approaches is the set of barriers to entry, including the extensive domain knowledge required, the need for in-house modeling capacity, and the costs involved. For these reasons, small enterprises with limited expertise and resources may find it impossible to undertake a significant ICME campaign. To address these challenges, a major focus of the MGI has been on developing the MII, but progress has been uneven, and much work remains to fully realize this vision. Finally, even under the best circumstances, the models employed in MGI-style materials design campaigns can be prohibitively slow, requiring clever approximations to achieve speedups. Of course, “clever” here implies that significant domain expertise is still essential for success. Nevertheless, a number of promising new avenues to address this issue are now becoming clear. Moving beyond the notion of a single investigator who “throws results over the wall” for others to use has been crucial to the success of the MGI. In addition to the successes and challenges discussed thus far, several other advancements have been precipitated by the MGI. Of particular note is the broad acceptance of more tightly integrated teams as an essential component of the materials R&D enterprise. Moving beyond the notion of a single investigator who “throws results over the wall” for others to use has been crucial to the success of the MGI. Modelers and experimentalists working hand-in-glove to accelerate materials design is now considered best practice under the MGI. Beyond this significant culture change, the MGI has also focused on many of the issues surrounding software and data publication to build out the MII. Alongside the technical challenges of such publication models, there remain significant cultural and incentive-related impediments. At present, there is little academic or industrial reward for publishing data and software, despite broad recognition of the value of data sharing in principle. The MGI has maintained close ties with other federal initiatives, including the National Nanotechnology Initiative (NNI). A key example of this intersection was highlighted on the cover of the NNI Supplement to the President’s Budget for fiscal year 2018, illustrating the close collaboration of experimentation and computation in nanomaterials (Lin et al. 2017; NSTC 2018). One lesson from the NNI is the importance of research infrastructure. Investigations at the nanoscale were enabled by tools that allowed manipulation of materials at the atomic level, such as scanning tunneling and atomic force microscopy. These breakthroughs have since led to an entire suite of scanning probe techniques. In the early days of nanoscience, only a few laboratories with these advanced tools were able to conduct research. A key area of NNI investment was the development of user facilities that provided access to the specialized equipment required for nanotechnology R&D. DOE developed user facilities at the National Labs, including the Nanoscale Science Research Centers, and NSF supported a series of networks based in universities, most recently the National Nanotechnology Coordinated Infrastructure (NNCI), which has 16 primary sites and 13 partner organizations providing researchers access to 71 distinct facilities and over 2,200 tools (NNCI 2025). These user facilities were instrumental in the development of the US nanotechnology community. They are often said to have “democratized” nanoscience, as researchers from smaller colleges and universities, as well as from small businesses, were able to conduct research that would not have been possible at their home institutions. The facilities played an important role across the entire ecosystem, from early-stage research through commercialization, and were pivotal in education and workforce training, with many sites hosting undergraduate students for summer research and other programs. Materials Innovation Infrastructure The fabrication and characterization tools available through the NNI user facilities are an important element of the infrastructure required to advance the MGI, but the MII also includes computational tools (models), data, and increasingly, robotics and autonomous systems. To better understand current capabilities, the MGI’s interagency working group on Autonomous Materials Innovation Infrastructure (AMII) hosted a June 2024 workshop bringing together experts from government, industry, and academia. This two-day workshop included a deep dive into existing resources for several materials classes and a discussion of key gaps. Participants identified hundreds of resources and needs, such as the development of automation in experimental hardware for materials synthesis, characterization, testing, and sample exchange; new AI decision methods; standardized data structures and representations; and improved sharing and reproducibility of data and results. The importance of strong industry–university–government collaboration was also emphasized, highlighting models such as public-private partnerships or consortia as potential vehicles to advance the AMII in the United States. More details are available in the effort’s workshop report (MGI 2024). The biannual MGI Principal Investigator (PI) meeting, which brought together hundreds of researchers funded by several MGI agencies, was another opportunity to build awareness of existing capabilities and identify areas where aligning the MII could help address key challenges. The MGI also hosted industry-focused events embedded in larger gatherings, including the “AI-Accelerated Materials Design and Deployment” town hall (TechConnect 2024) and “The Materials Genome Initiative & Microelectronics: Designing the Next Generation of Materials” workshop at the Electronics Resurgence Initiative (ERI) 2.0 Summit (DARPA 2023). These events, along with a 2025 MGI request for information (Federal Register 2025), have provided valuable insight into the current AMII, key challenges and examples of industry interest in and adoption of MGI approaches. Opportunities Ahead For decades, accelerating materials design has depended on our ability to accurately model material behavior. While physics-based models offer powerful insights, their computational cost often limits practical application, especially in fast-paced manufacturing environments. As a result, even the most advanced practitioners rely on approximations and selective use of these models to make informed decisions. AI is now poised to upend this paradigm. AI can generate predictive models where none previously existed and create “surrogate models” that replace physics-informed simulations with AI-driven approximations. While not exact replicas, these models can operate at speeds many orders of magnitude faster than traditional simulations, making real-time materials design and analysis feasible in ways never before possible. As with all modeling efforts, verification, validation, and uncertainty quantification remain essential to ensure a trustworthy, predictive materials innovation infrastructure. The integration of AI doesn’t stop at modeling. Automation of laboratory processes is enabling high-throughput materials synthesis, which can then be followed by AI-powered characterization techniques that rapidly assess whether a newly developed material possesses the desired properties. If not, AI-driven optimization algorithms can refine the design, automatically specifying the next experimental iteration. This feedback loop is at the heart of autonomous experimentation (AE, or self-driving labs), a transformative technology set to redefine materials research (Stach et al. 2021). Beyond accelerating innovation, AE addresses a fundamental challenge in materials science: the scarcity of high-quality materials data. Unlike fields such as biology, weather forecasting, and finance, where vast datasets fuel machine learning models, materials R&D has historically struggled with limited and fragmented data. AE can break this bottleneck by generating vast and reliable datasets, laying the foundation for next-generation AI models. Critically, this revolution extends beyond research labs. AI-driven surrogate models, fueled by AE-generated data, can create “materials digital twins”—fast, accurate computational representations of materials and their behaviors across processing and performance conditions. These digital twins will operate at manufacturing-relevant timescales, enabling seamless integration into the digital thread of production. The result is a direct and unprecedented translation of fundamental materials knowledge from early-stage research to industrial applications, bridging long-standing gaps that have traditionally slowed innovation. AI-powered materials R&D will propel the field into an era of speed, efficiency, and predictive power unlike anything seen before. To realize the opportunities ahead, it will be important to leverage the strong community established under the auspices of the MGI to bring together researchers and developers from academia, industry, and government, across disciplines and stages of the technology development continuum. Likewise, it is imperative to build on the foundational tools and methods developed under the MGI—the Materials Innovation Infrastructure—while embracing and adopting new technologies such as AI and AE. AI-powered materials R&D will propel the field into an era of speed, efficiency, and predictive power unlike anything seen before. The convergence of AI, automation, and materials science is not merely an incremental improvement; it is a revolution poised to redefine how we discover, design, develop, and deploy the materials of the future. References DARPA [Defense Advanced Research Projects Agency]. 2023. Electronics Resurgence Initiative 2.0 Summit, August 22–24, 2023. Seattle. Online at eri-summit.darpa.mil. de Pablo JJ, Jackson NE, Webb MA, and 17 others. 2019. New frontiers for the materials genome initiative. npj Computational Materials 5:41. Online at https://www.nature.com/articles/s41524-019-0173-4. Federal Register. 2025. Notice of request for information (RFI) on autonomous experimentation platforms from Material Genome Initiative. Online at www.federalregister.gov/documents/2025/01/17/2025-01161/ notice-of-request-for-information-rfi-on-autonomous- experime ntation-platforms-from-material-genome. Lin H, Lee S, Sun L, Spellings M, Engel M, Glotzer S, Mirkin CA. 2017. Clathrate colloidal crystals. Science 355:931–35. Online at www.science.org/doi/10.1126/science.aal3919. MGI [Materials Genome Initiative]. 2024. Accelerated Materials Experimentation Enabled by the Autonomous Materials Innovation Infrastructure (AMII). A Workshop Report. Online at www.mgi.gov/sites/mgi/files/MGI_Autonomous_Materials_ Innovat ion_Infrastructure_Workshop_Report.pdf. NIST [National Institute of Standards and Technology]. 2025a. CALPHAD data informatics. Online at https://www.nist.gov/programs-projects/calphad-data- informat ics. NIST. 2025b. Materials Genome Initiative. Online at www.nist.gov/mgi. NNCI [National Nanotechnology Coordinated Infrastructure]. 2025. Tools. Online at nnci.net. NRC [National Research Council]. 2008. Integrated Computational Materials Engineering: A Transformational Discipline for Improved Competitiveness and National Security. Washington: National Academies Press. Online at https://nap.nationalacademies.org/catalog/12199. NSF [National Science Foundation]. 2025a. Designing Materials to Revolutionize and Engineer Our Future (DMREF). Online at www.nsf.gov/funding/opportunities/dmref-designing-materials- revolutionize-engineer-our-future. NSF. 2025b. Materials Innovation Plans (MIP). Online at www.nsf.gov/funding/opportunities/mip-materials-innovation- platforms. NSTC [National Science and Technology Council]. 2014. Materials Genome Initiative Strategic Plan. Online at www.mgi.gov/sites/mgi/files/mgi_strategic_plan_-_dec_2014. pdf. NSTC. 2018. The National Nanotechnology Initiative Supplement to the President’s 2018 Budget. Online at www.nano.gov/sites/default/files/pub_resource/NNI-FY18- Budget-Supplement.pdf. NSTC. 2021. Materials Genome Initiative Strategic Plan. Online at www.mgi.gov/sites/mgi/files/MGI-2021-Strategic-Plan.pdf. Stach E, DeCost B, Kusne AG, Hattrick-Simpers J, Brown KA, Reyes KG, Schrier J, Billinge S, Buonassisi T, Foster I, and 11 others. 2021. Autonomous experimentation systems for materials development: A community perspective. Matter 4(9):2702–26. Online at www.cell.com/matter/fulltext/S2590-2385(21)00306-4. TechConnect. 2024. Town hall: AI-accelerated materials design and deployment. TechConnect World Innovation Conference & Expo. Online at www.techconnectworld.com/World2024/sym/TCW_Town_Hall.html. Warren J. 2024. The Materials Genome Initiative and the metals industry. The Bridge 54(910):25–29. Online at www.nae.edu/313309/The-Materials-Genome-Initiative-and- the-Metals-Industry. About the Author:Lisa E. Friedersdorf is principal assistant director for physical sciences and engineering and executive director, National Science and Technology Council, White House Office of Science and Technology Policy. James A. Warren is director, Materials Genome Program, National Institute of Standards and Technology.