In This Issue
Spring Bridge on the US Metals Industry: Looking Forward
March 29, 2024 Volume 54 Issue 1
In this issue of The Bridge, guest editors Greg Olson and Aziz Asphahani have assembled feature articles that demonstrate how computational materials science and engineering is leading the way in the deployment of metallic materials that meet increasingly advanced design specifications.

The Materials Genome Initiative and the Metals Industry

Monday, April 8, 2024

Author: Jim Warren

The Materials Genome Initiative highlights the promising role of government in materials innovations.

In June 2011, during a speech at Carnegie Mellon University, President Barack Obama announced the Materials Genome Initiative (MGI).[1] The MGI is a multiagency effort to create a new era of policy, resources, and infrastructure that support US institutions in the endeavor to discover, manufacture, and deploy advanced materials twice as fast and at a fraction of the cost. The MGI is now twelve years old, but its roots go back a very long way, and the connection to the metals industry is especially strong.

The MGI was founded to accelerate materials deployment, and the mechanism that enables this is the so-called Materials Innovation Infrastructure (MII), which lies at the heart of the MGI, as is graphically depicted in figure 1. This simple-looking diagram is an attempt to articulate the tight integration of computation, experimentation, and data management that is essential to achieving the goals of the MGI. All of these concepts are not particularly profound, but the creation of the MII seeks to lower the barrier to applying advanced approaches to materials design and deployment, which, if successful, would have an enormous impact on manufacturing and the US economy.

In a very real sense, the MGI, building off a large number of prior studies (Glotzer et al. 2009; Jou et al. 2004; NRC 2004; NRC 2008), was about raising the profile of computational modeling within the materials R&D community to the same status as experimental approaches, as it was becoming clear that computation was at a place where it could substantively accelerate materials R&D, but it was not broadly adopted by industry. This was a central argument for the creation of the MII, which could make these approaches far more accessible and integrate modern techniques in materials modeling into industrial workflows.

Warren Fig 1.gifThere are a large number of technical approaches that fall under the MGI, all of which are united by the tight fusion of modeling, experimental approaches, and data management to accelerate materials R&D. One of the inspirations for the MGI was the Materials Genome ­Project at MIT, which has now graciously changed its name to the Materials Project to avoid confusion.[2] This effort, supported by the US Department of Energy, generates and makes available a vast trove of density functional theory calculations (quantum mechanical calculations) of materials properties for use by the materials R&D community. Another prominent MGI approach is integrated computational materials engineering (ICME) (NRC 2008). ICME has several decades of demonstrated use by industry to concurrently design materials for insertion in products, vastly reducing the time and cost and showing a near-order-of-magnitude return on investment.

Perhaps one of the most remarkable stories of the successful use of ICME comes from the small company QuesTek Innovations. QuesTek has been using ICME techniques as its core technology to solve alloy development problems for a variety of industrial clients for ­several decades, and in 2012 QuesTek sold their technology to Apple while remaining in business for themselves. Apple subsequently used the QuesTek approaches to develop many of the metals found in their products, such as the Apple Watch and iPhone. In time, QuesTek-trained people migrated to SpaceX, where their approaches have allowed for massively accelerated alloy development to be deployed on their rockets. In support of the MGI, in 2016, the National Institute of Standards and ­Technology (NIST) supported several Quantitative Benchmark for Time to Market Framework case ­studies, one of which was the QuesTek alloy Ferrium M54,[3] which was developed in about six years from conception to deployment on a US Navy aircraft (far shorter than the usual fifteen to twenty years typical of alloy development and deployment times).

Another example of the power of these approaches can be found in what might seem an unlikely venue: money. In this case, NIST developed new coinage materials for the US Mint using MGI approaches and associated tools (Lass et al. 2018). Prototype alloys were designed in just eighteen months and validated by the US Mint in their 2022 Biennial Report to Congress (Alloy C99750T-M). The bill allowing the US Mint to use this alloy was re-introduced in the Senate in the spring of 2023.

Of course, beyond modeling it was broadly understood that materials data, whether from trusted models or experiments, was a poorly managed resource. In the case of industry, data will only be shared where it is in the interest of the sharing entity. All data costs money to produce and often would be of considerable value to the broader R&D community if made widely available. At a minimum, the data is valuable to the researchers that generated it. And yet, most workflows for curating these data remain extremely crude, with little metadata captured that would make understanding and reusing the data far more tractable. As is a topic largely ignored in traditional materials research environments, the MGI has focused considerable energy on exploring and supporting work to address these challenges.

NIST has framed its support for the MGI around data. NIST and its predecessor, the National Bureau of Standards, have a 120-plus-year history of providing high-quality data to the world (e.g., the US time standard[4] or the Chemistry WebBook[5]). For the MGI, NIST identified three principal thrusts: (i) enabling the exchange of materials data, (ii) ensuring the quality of materials data, and (iii) developing new methods and metrologies based on the broad availability of materials data (like data-driven materials R&D). More information on the NIST program and its numerous projects can be found on the NIST website.[6]

The MGI at the Outset

One of the areas of greatest success in the application of modeling to accelerate the design of new materials is in metallic systems. There are a number of reasons for this, including (i) the millennia-long history of metallurgy and several centuries of more quantitative understandings, and (ii) the relative simplicity (compared to polymers and ceramics) of the chemistry of metals, which allows for the application of rigorous thermodynamic principles to these systems. Starting in the 1970s, the technique known as CALPHAD (CALculation of PHAase Diagrams is the origin of the acronym, although the meaning is now broader) was developed to allow a database of measured properties to be assembled, and these databases became the basic inputs in predictive models of the processing-structure-property models that enable materials designers to create new alloys with targeted properties.

This predictive power is the essence of MGI ­approaches, allowing materials R&D efforts to feed experimental results into models, design new alloys, and iterate to converge rapidly to the desired combination of properties. However, making approaches like these more broadly available entails more than encouraging the use of certain software packages. It requires a host of associated additional considerations. It was in this light that the National Science and Technology Council (NSTC) subcommittee on the MGI rolled out a strategy in 2014. This strategy contained four goals:

  1. Leading a culture shift in materials-science research to encourage and facilitate an integrated team approach.
  2. Integrating experiment, computation, and theory and equipping the materials community with advanced tools and techniques.
  3. Making digital data accessible.
  4. Creating a world-class materials-science and engineering workforce that is trained for careers in academia or industry.

As can be seen from this list, the lead goal was not a technical issue but instead revolved around a culture shift in the conduct of materials research. This requirement was essential to achieving the goals of the MGI, which otherwise focuses on developing the MII, specific issues around data that had largely been left unexplored within the materials R&D community, and the ever-present need for training in state-of-the-art methods. As the MGI evolved, the culture shift began to take hold. And as the MGI approached its first decade, it was time for a fresh look at the materials research landscape, its requirements, and new opportunities.

The MGI Today

As the MGI was taking hold across academia, translating these ideas into industrial practice became ever more pressing, as the adoption of MGI approaches was uneven. The economic arguments for widespread adoption seem irrefutable. In 2018, NIST supported a prospective study including an economic analysis and interviews with more than 120 industry experts on their needs for new technological infrastructure supporting advanced materials innovation and the potential economic impacts of meeting those needs. It was concluded that an improved MII (and its adoption) would deliver between $123 billion and $270 billion in value annually.7

One of the areas of greatest success in the application of modeling to accelerate
the design of new materials
is in metallic systems.

The reasons for lack of adoption were, and remain, complex, but the tide is beginning to turn. Certainly, the metals industry was far ahead of other sectors in their adoption of MGI approaches, but even there, significant impediments remain. Most of these barriers will fall away as the costs of adoption are lowered and the ­success ­stories become more broadly appreciated. One of the most promising areas that has arisen is the application of artificial intelligence (AI) approaches to materials R&D.

When NIST rolled out its strategy to support the MGI, as discussed above, the third element was the development of new methods and metrologies enabled by the broad availability of materials data (like data-driven materials R&D). Data-driven materials R&D is, of course, an accurate characterization of AI, although all data-driven approaches are not AI. Thus, NIST was extremely well situated when the AI revolution (as far as the application to materials is concerned) began around 2016 and continues to accelerate today.

The current importance of AI to the MGI is hard to overstate.

While the need for high-quality data and careful experi­mentation informed by physical models is never going away, the current importance of AI to the MGI is hard to overstate. Indeed, there are substantial reasons to be cautious with the applications of these techniques, as they can deceive the practitioner if used improperly. However, the potential benefits are readily apparent. With the ability to detect patterns in a manner that used to be strictly the provenance of humans, the process of micro­structural characterization has become easier, faster, cheaper, and more accurate. Combining these approaches with ­robotics and mathematical inference models, autonomous (self-driving) materials R&D laboratories are becoming a reality, with groundbreaking demonstrations being reported ever more frequently. All of these developments are, of course, precisely in the wheelhouse of the MGI, as AI is a computational model and slots into the MGI paradigm with no need for modification. The success of these approaches, their relative low cost, and ease of deployment should effectively increase the adoption of MGI approaches.

In light of these developments, the subcommittee on the MGI realized it was time for a new strategy, and in 2021 they released an updated strategic plan.1 The plan consisted of three goals:

  1. Unify the Materials Innovation Infrastructure.
  2. Harness the power of materials data and accelerate materials R&D through the application of AI.
  3. Educate, train, and connect the materials R&D workforce.

The observant reader will note the absence of anything like a culture shift goal, as it is now clear to most of the MGI stakeholders that these approaches to ­materials R&D are clearly effective. The remaining goals are ­natural evolutions of the original goals. The MII is and always will be a work in progress. So, much of the effort must be on unifying otherwise disparate pieces to enable more effective R&D. The second goal has moved from a data-centric view of materials R&D to an AI-focused goal, with data as the enabling technology, while the third goal is an updated version of the education and training goal found in the original plan, taking into account the new approaches and technologies that have come to the fore since the original strategy was crafted. Indeed, the federal agencies supporting the MGI are all working aggressively to execute on this strategy in a rapidly evolving landscape. For example, generative AI models (e.g., Stable Diffusion and ChatGPT) have all risen to prominence in the less than two years since the strategy was released, and these models seem to have extraordinary potential for applications in the MGI space (as well as across much of science and other disciplines.)

It is worth drawing attention to one of the major objectives under the first goal: accelerate the adoption of the MII through the National Grand Challenges. This objective has the benefit of rallying the community around pressing national and worldwide ­challenges while simultaneously defining and accelerating the development of MGI approaches crucial for the long-term success of the MGI. The grand challenges could be any of the pressing environmental or social issues that occupy much of our attention, but the technical solutions to these challenges are undergirded by MGI approaches. In the metals arena, all of the areas articulated in the strategy are relevant, including issues around climate change, environmental degradation, energy storage, renewable power generation, critical materials substitution, advanced healthcare technologies that rely on new biocompatible materials, new manufacturing capabilities that address the lack of resilience in existing systems, and improved materials to rebuild an aging physical infrastructure.

The Metals Industry and the Way Forward

The metals industry and, more generally, the technologies that make the design of new metal alloys possible using the techniques and data made available through the MII have been touchstones for the MGI. The ­metals industry has led the charge towards demonstration of the high return on investment for MGI approaches while also exposing the various gaps in methods, data, and knowledge that must be filled for even more widespread adoption of these techniques. Indeed, there are some fascinating issues around the generalization of MGI approaches that use the phase diagrams that are successful for metallic systems for polymer systems. A great deal of the thinking around those ideas has been explored in the Polymer Properties Predictor Database (PPPDB) effort based at the Center for Hierarchical Materials Design in Chicago.[7] While such approaches are not inherently the best way to proceed for industrially relevant model of polymer systems, it is certainly the case that the more the successes in metals systems can be leveraged, the better it will serve the broader materials R&D community.

Ultimately, the MGI and the concepts it ­encompasses are the only way forward for materials R&D, and for the metals industry’s goal of designing new alloys with ­tailored properties. While there are many other issues, both social and technical, that complicate the insertion of MGI approaches into industrial practice, the ultimate realization of the goals of the MGI has and will continue to pave the way for a nimbler and more cost-­effective industrial base and increasingly impactful metallic systems.


Certain commercial entities are identified in this paper in order to clearly elucidate the landscape of materials R&D. Such identification does not imply recommendation or endorsement of any product or service by NIST, nor does it imply that the entities identified are necessarily the best available for the purpose.


Glotzer SC, Kim S, Cummings PT, Deshmukh A, Head-Gordon M, Karniadakis G, Petzold L, Sagui C, Shinozuka M. 2009. WTEC Panel Report on International Assessment of Research and Development in Simulation-Based Engineering and Science. World Technology Evaluation Center, Inc. Baltimore, Maryland.

Jou HJ, Voorhees P, Olson GB. 2004. Computer simulations for the prediction of microstructure/property variation in aeroturbine disks. In: Superalloys 2004, 877–886. Green KA, Pollock TM, Harada H, Howson TE, Reed RC, Schirra JJ, Walston S, eds. Warrendale, Pennsylvania: The Minerals, Metals & Materials Society.

Lass EA, Stoudt MR, Campbell CE. 2018. Systems design approach to low-cost coinage materials. Integrated Materials and Manufacturing Innovation 7:52–69.

National Research Council (NRC). 2004. Accelerating Technology Transition: Bridging the Valley of Death for Materials and Processes in Defense Systems. Washington, DC: The National Academies Press.

NRC. 2008. Integrated Computational Materials Engineering: A Transformational Discipline for Improved Competitiveness and National Security. Washington, DC: The National Academies Press.

The United States Mint (Department of the Treasury). 2022. 2022 Biennial Report to Congress as Required by the Coin Modernization, Oversight, and Continuity Act of 2010 (Public Law 111-302). Washington, DC. 04/2022-USM-Biennial-Report_P5_FINAL.pdf.



[3] innovation_case_study_questek_090616.pdf





About the Author:Jim Warren is director, the Materials Genome Program, the National Institute of Standards and Technology.