Executive Summary

Mid-scale manufacturing, for the purpose of this issue paper, is defined in terms of both the maturity level (Category A), which is between fundamental university research and industrial for-profit production, and the production volume (Category B), which is between single items and mass production that typically exceeds tens of thousands of pieces of one kind. Both categories are urgently needed for applications used by the U.S. Army, the U.S. Navy, and the U.S. Air Force (hereafter referred to as “the three services”) as the former is for translating scientific and technological innovations to practices providing warfighters with advantageous performance, and the latter is necessary for satisfying logistic and supply chain needs at any time to ensure an agile response to any rapid conflict development.

In January 2025 and March 2025, the National Academies of Sciences, Engineering, and Medicine hosted a two-part workshop event on mid-scale manufacturing and characterization capacity challenges, during which almost 30 percent of speakers were from industry. Demand for mid-scale manufacturing has been increasing, and the United States has a solid and wide base for innovation, assessment, and testing of new materials at university and national laboratories. Nevertheless, the following factors have hindered existing mid-scale manufacturing:

1. From companies’ viewpoints, operational investments are too risky (Category A) and market-driven economics favor mass production and large manufacturers (Category B). This results in companies, particularly small and medium-sized manufacturers (SMMs), disappearing from the mid-scale manufacturing supply chain for the three services, and this consolidation reduces capability.
2. Research innovation is highly regarded, whereas manufacturing innovation has received less attention and investment (Category A and Category B).
3. The three services have very specific material needs (e.g., energetics) that do not have dual-use operations, which creates challenges in producing enough material for needed analysis and tool development (Category A).
4. The overall U.S. productivity decline due to the lost workforce, the lack of investment in digitalization in SMMs, and more broadly the lack of long-term investment vision in SMMs puts additional stress on mid-scale manufacturing (Category B).

Opportunities or actions to address the above limiting factors to mid-scale manufacturing include the following four categories. Note that each category can address several of the above-mentioned barriers.

1. Physical Testing Development and Planning Facilities
   1. Provide better access to national laboratories equipped with flexible, agile, near-production-scale processing and test equipment with skilled staffing.
   2. Invest in and encourage universities to develop shared facilities for both material discovery and manufacturing innovation.
2. Digitalization Through Data, Information, and Modeling
   1. Establish a large-scale industry-wide “digital interstate” to provide the infrastructure that

1. can aggregate mid-scale production requirements into scalable mass production (i.e., use mid-scale manufacturing for distributed or modular manufacturing in supply chain);
2. can facilitate ingesting data and models;
3. can assist, build, and share artificial intelligence (AI) tools for material discovery with integrated synthesis, characterization, and modeling to assess toxicity and manufacturability; and
4. can assess models for planning and projecting access, investment, and operations.
2. Develop physics-based and AI-accelerated rapid simulation tools to assist the development of mid-scale manufacturing facilities and process knowledge.
3. Implement smart manufacturing to control, manage, and optimize at scale across factories and supply chains.
3. Workforce and Staffing
   1. Train and retain skilled staff members to operate the shared facilities mentioned in Category A as well as the models and tools developed as part of Category B.
   2. Provide short leave or joint appointments between manufacturers and national laboratories and academia to better understand their needs and capabilities.
4. Economics and Organization
   1. Coordinate national investment effort in manufacturing innovation related to material research and development (R&D) innovations, such as considering continuous processing as an alternative to batch processing.
   2. Encourage dual-use thoughts and markets.
   3. Develop a data-driven framework or tool (intellectual property data, market data, etc.) with mid-scale manufacturing as a central focus to help investment decisions.
   4. Organize investment pathways, Small Business Innovation Research programs, tax credits, and venture funding for mid-scale manufacturing to invest in needed facilities and staff.
   5. Streamline administrative and acquisition processes for the three services and remove access barriers to equipment and infrastructure.

Introduction

In January 2025 and March 2025, the National Academies hosted a two-part workshop event on mid-scale manufacturing and characterization capacity with a focus on the three services and their critical materials supply challenges. The presentations and conversations during this workshop event, which are summarized in two Proceedings of a Workshop—in Brief documents (NASEM 2025a,b), revealed key challenges related to mid-scale manufacturing and potential paths forward that we explore further in the current paper. We offer our understanding of the concept of mid-scale manufacturing, in general and in relation to operations by the three services; discuss the technical and operational, economic, and policy challenges of mid-scale manufacturing and possible strategies to address them; present relevant case studies; and describe key opportunities for the three services. With enhanced facilities and capabilities, data collection and

management initiatives, financing strategies, workforce development, and dual-use approaches, the three services could more successfully leverage mid-scale manufacturing.

Mid-Scale Manufacturing

Advanced materials and processing capabilities, including those composed of strategic and critical materials, enable the United States to mature existing approaches as well as develop and sustain emerging technologies to improve warfighter capability. Although these materials and their processing technologies support military and civilian industry, many scarce, strategic, and high-performance materials are often not found or produced in the United States in quantities of the right volume to meet the needs of the three services (NRC 2008a; USGS 2025). Maturing, accelerating, scaling, and transitioning technologies to strengthen domestic material production and use of materials for military and dual-use applications are essential actions to ensure reliable access, enhance national defense, and promote economic prosperity.

Mid-scale manufacturing is one solution bridging the efforts between university research and mass production. Mid-scale manufacturing typically involves a relatively low volume of materials—larger than for prototyping but smaller than for mass production (i.e., Category B). How individual companies specifically define “low volume” and “mid-scale,” however, may vary depending on the needs of each customer, the desired material, and whether the material is in the stage of development or production. But mid-scale manufacturing should not only be seen as a level of maturity (i.e., Category A) where a material, process, or product is between development and a final production stage of maturity. After maturity, mid-scale manufacturing can also be seen as a rate of production falling between mass production and single items. It is worth noting that while “small and medium-sized manufacturing” is a formal category within “small and medium-sized enterprises” with clear size, reach, capacity, and resources, “mid-scale manufacturing” is more of a practical, industry-relevant term lacking a formal definition.

Producing sufficient strategic quantities of emerging materials (e.g., alloys and energetic chemicals for military components and systems) enables one to assess their engineering utility or establish a digital footprint with quantified uncertainty. User facilities are essential; these facilities could conduct validation and verification as well as performance assessments of processes and products at a rate that is not limiting (e.g., high-throughput analysis of fracture, fatigue, and thermal shock) and that will close the loop on processing–structure–performance correlations using acquired data and appropriate modeling and simulation tools. Mid-scale manufacturing endeavors could include, for example, accessing novel alloys in quantities appropriate for conducting process and performance assessments; leveraging single crystals for initial wafer, electronic, or device evaluation; accessing sufficient volumes of specialty polymers and resins to produce and test full-scale components, composites, films, or coatings; and creating and analyzing realistic volumes of inks and feedstocks for additive manufacturing, electronic packaging materials, electronic materials and sensors, and novel energetics and propellants.

Despite excitement about the potential to develop new materials, if these materials cannot be produced at scale, they will not have the desired impact. Determining whether and when to mature a new material and begin producing and processing at scale, however, depends on the availability of connections and tools. Mid-scale and full-scale equipment for producing reasonable quantities in a realistic environment are necessary to guide scaling decisions and validate the opportunity or capability that a material promises.

Capital investments in facilities, infrastructure, tools, and people—as well as the ability to retool quickly, a commitment to and mechanism for data sharing, and sustainment models—are critical to achieve these objectives. Although federal investments have emerged, several challenges and gaps remain in mid-scale manufacturing, including the following:

1. One of these gaps is in the availability of, access to, and economic sustainment of facilities that provide sufficient quantities of the emerging materials the three services need so that manufacturability can be evaluated and prototype-engineered parts can be produced and tested (NASEM 2025a,b); facilities that support characterization in extreme environments across various areas of interest to the three services are particularly important. Readily accessible facilities for rapid characterization for each class of materials would provide critically needed data to assess performance and manufacturing quality, validate models, fine-tune modeling and simulation tools, and provide insight into processing–structure–property correlations for a specific environment or manufacturing process.
2. A second key gap is in the availability of common infrastructure, standards, computing hardware, simulation tools, and tools to ingest data and models—from scale-up and performance assessment to integration with discovery and deployment—to accelerate future development of alternative materials and processes (see, e.g., GAO 2025a; NASEM 2025a,b).
3. A third gap arises from the lack of dedicated engineers and other staff needed to operate all of these facilities and tools (Miller 2025).

Overall, U.S. industrial challenges reflect, to some degree, the cumulative effect of long-standing practices and policies that have affected mid-scale producers. Market dynamics are less favorable of mid-scale investments, which require significant capital and deliver moderate returns, leaving companies in the zone between large industries and small innovators and universities. At the same time, small and medium-sized industries face widening gaps in technology adoption, workforce skills, and production capabilities, limiting their competitiveness in advanced supply chains. Taken together, these factors can make domestic materials production disproportionately difficult and underscore the urgency of pursuing new opportunities—for example, through targeted investment models, workforce development, and innovation support—that can overcome barriers and strengthen mid-scale competitiveness.

The three services could benefit from a better understanding of how existing effective models for operations and access could begin to address their specific needs and close these gaps. Key strategic questions emerge about which technologies should move forward and how, especially given the difficulty of crossing the “valley of death” (i.e., the gap between research and commercialization, typically technology readiness

levels [TRLs] 3–7) and because the volume associated with mid-scale manufacturing can create economic challenges (i.e., mid-scale manufacturing does not lend itself to the high profit margins that result from mass production efforts). Traditional approaches to proceed from discovery to deployment can be too slow or too costly to meet the growing materials needs of the three services (GAO 2025b). The United States could benefit from developing novel means to manufacture and test small- and medium-batch materials in an economically and environmentally viable way as well as from pursuing potential pathways toward “dual-use” technologies, which could benefit both the three services and the commercial sector and address some of the challenges associated with the acquisition process and timeline within the three services.

Challenges of Mid-Scale Manufacturing

Despite the three services’ reliance on mid-scale manufacturing, many obstacles persist, including technical and operational, economic, and policy challenges. Developing specific strategies to address and overcome these challenges—which in turn surface crosscutting challenges for technology transition, supply chain resilience, domestic manufacturing capability, workforce development, strategic roadmapping, and acquisition processes in particular—is key to the three services’ ability to better leverage and benefit from mid-scale manufacturing.

Technical and Operational Challenges

Warfighters depend on a robust, agile, and adaptable industrial base. Continued innovation in mid-scale manufacturing is critical to supporting this need; however, several technical and operational challenges would need to be addressed and overcome to move forward. These challenges, described below, arise during the transition from the research scale to the mid-scale and as a result of vulnerable supply chains, limited access to information, decreases in both medium-sized manufacturers and manufacturing productivity, and concerns about the implementation of advanced tools and technologies.

A key technical challenge arises when making the transition from the laboratory to mid-scale production. For example, certain materials that may weld well at the small scale (e.g., button size) do not retain those properties after the scale-up process (e.g., rolling), as processing capabilities alter the microstructure of materials. Thus, the transition from the small scale to the mid-scale reveals key insights to help determine whether to continue developing a certain technology. The three services would benefit from a new mechanism to accelerate this transition and to reveal critical insights earlier in the process (a process that currently can take as long as a decade). For example, the three services could support mid-scale or pilot facilities to better enhance and ensure the viability of mid-scale capabilities. The three services could also develop facilities with flexible infrastructure that can be reconfigured quickly, enhanced with multiple embedded sensors in tools and the environment to help develop and debug scale-up challenges. Because data from small-scale manufacturing have not yet been gathered in significant quantities to support the use of data-driven methods, the three services could also develop physics-based rapid simulation tools to better understand these challenges. Focusing on enhancing processes at the mid-scale will likely be less expensive and less risky than moving directly to large-scale production, although this decision also depends

on allowable timelines and dual-use considerations. Employing graphical representations (e.g., similar to an Ashby plot) could be useful in the early stages of the manufacturing process to visualize the relationship among product flexibility, capital cost per unit, risk tolerance, and production volume; developing a better understanding of this relationship could help in prioritizing and identifying potential problems for mid-scale manufacturing.

A key operational challenge emerges when supply chains are not robust and resilient. For example, supply chains for critical materials are especially fragile. Without a resilient supply chain, material could be in short supply, surge capacity could be diminished, the supply chain could become more vulnerable, and the United States could lag in its strategic defense positioning (Monroe 2025). For energetics in particular, if a single component in the supply chain fails, or if an adversary contributes counterfeit components, national security could be compromised. Furthermore, energetic materials are single use, affecting both supply and demand and the ability to stockpile. Prioritizing the production of new munitions and platforms is thus critical, as is reshoring feedstocks and explosives. Use of distributed manufacturing and modular manufacturing could also be valuable approaches to addressing some of these supply chain issues (NASEM 2025a).

Further supply chain challenges result because many small businesses in a given supply chain have limited access to information about starting materials and processing histories. These supply chain members need to be better connected to enhance information sharing and enable the use of any advanced technologies. Government and academia could encourage better coordination along the supply chain for some technologies (Davis et al. 2022). Additionally, increased use of supply chain analytics could capture production- and demand-related data more efficiently, aiding demand forecasts, enhancing partnerships, identifying and protecting intellectual property (IP), and supporting dual-use defense contracting and demand signal management (NASEM 2025a).

A related operational challenge is that some SMMs are disappearing entirely from the supply chain. If this trend continues, a single company failure in the future could break an entire supply chain, especially if no large manufacturers are interested in stepping in due to the small scale of the effort or the restricted demand. This issue is especially critical because SMMs often have the best capabilities to produce materials that the three services need at appropriate scales to ensure the stability of U.S. security. Without these small and medium-sized companies, mid-scale capabilities, too, are at risk of disappearing. This is alarming because larger facilities are tuned to be financially viable when producing very large-scale volumes for consumer products and automotives.

Another example of the complicated challenges related to the supply chain for mid-scale manufacturing emerges from the steel industry. In 2024, six companies held more than 60 percent of U.S. casting capacity, whereas in 2003, these six companies held only 39 percent of the steel casting capacity in North America (Webler 2025). In part, this relates to broader economic challenges, as lower-cost cast steel is available in other parts of the world, and China now holds a dominant position in the production of steel as well as critical materials more broadly (Monroe 2025). Furthermore, U.S. manufacturing in general declined in 2020 to 16 percent of the world manufacturing output while that of China rose to 31 percent (NASEM 2025b),

perhaps in part because for decades the United States focused on R&D-led innovation, while other countries developed manufacturing-led innovation. If the United States could instead focus on both types of innovation and invest more in both people and technologies, these historical deficiencies could begin to improve (NASEM 2025b). Additionally, increased efforts to de-risk investments could entice more suppliers into the supply chain.

U.S. manufacturing productivity specifically, which has been stagnant or very weak for more than 15 years, presents yet another challenge for mid-scale manufacturing. Between 2000 and 2010, the United States lost one-third of its manufacturing workforce, and more than 60,000 manufacturing plants were closed. U.S. investment in manufacturing capital plants and equipment has also declined. In 2024, the United States had a $1.2 trillion trade deficit in manufactured goods. Small and medium-sized U.S. manufacturers in particular, who comprise approximately 46 percent of U.S. manufacturing output, have a significant productivity problem (NASEM 2025b). Process planning is difficult for these manufacturers, who would benefit from improved software and increased capital for scale-up.

Once the process of manufacturing is initiated, more technical challenges may arise related to manufacturing operations, materials discovery, process optimization, facilities, and smart manufacturing (which by design is based on AI, machine learning [ML], and digital twins [DTs]). AI, ML, and DTs could be transformative approaches to addressing such challenges (NASEM 2025a,b), as these tools could improve productivity, efficiency, and sustainability in the workforce, factories, and supply chains; accelerate manufacturing innovation; and increase U.S. economic competitiveness and resilience (NASEM 2024). However, effective data integration and system interoperability remain significant hurdles to the widespread implementation of these tools.

For example, DTs are only as powerful as the data that inform them. Yet, to collect these data, experiments would need to be run at scale, which can be both technically and economically challenging. The creation of a “digital interstate” in which data can be shared at various levels could be a first step to enhancing DTs with better data (NASEM 2024). DTs could then be used to de-risk uncertainties for mid-scale manufacturing. In additive manufacturing, for instance, DTs have the potential to encode knowledge about processes and mechanical behavior and thus could be used to model components and support scale-up and technology transfer (NASEM 2025b). However, ongoing work is needed to increase both trust in and use of DTs. Regarding the potential use of other tools, research is under way at the University of Toronto, for example, to determine whether AI and robotics could enable faster discovery, a more reproducible process at the small scale, and development at the mid-scale (NASEM 2025a). AI and ML tools are already being used at the National Energy Technology Laboratory (NETL), for example, to design and to learn how to control the microstructures of high-strength, low-alloy steels (NASEM 2025b). It is also worth noting that a digital interstate is a cross-company, cross-industry, and cross-agency infrastructure requiring coordinated national investment in manufacturing innovation. The point is that the issue is not just data; the issue is qualified, processed, aggregated, categorized, and engineered data accessible at scale, ideally as data sets and if possible as validated models.

Although advanced tools could indeed be transformative, fundamental technology development and technical know-how remain essential, and facilities cannot be operated successfully without the right people with the right skill sets. Furthermore, the implementation of advanced tools and technologies does not come without risk. An additional challenge arises regarding cybersecurity.

The data generated in laboratories and other facilities need to be protected, as do the information technology and operational technology systems within these environments. That said, data collected from facilities can be used to co-develop early-stage physics and ML models. These can then, in turn, be used to build a simultaneous understanding of material–machine phenomena and control and operations, which are both useful for “DT scale-up” planning. This early-stage modeling takes advantage of smart manufacturing’s emphasis on consistent data collection, contextualization, processing, and engineering for scaling and model building.

Economic Challenges

Many of the previously discussed technical and operational challenges are closely intertwined with economic challenges. The economic challenges of scaling and commercialization in prior generations still exist today. These challenges, described below, emerge from reduced market potential, the lower-volume production of mid-scale manufacturing, difficult decision making to balance risk and return, the need to secure appropriate and adequate investments, and a lack of critical incentives.

The decision to commercialize a particular technology is usually driven by economic opportunity. The profit/cost ratio and return on investment are central to decision making for both start-ups and established firms; without a return on investment or national interest, little motivation exists to invest in mid-scale manufacturing (NASEM 2025b). Decision makers in technology accelerators and public–private partnerships (PPPs) in particular need to pay close attention to market perspectives, and assessing customer needs is a first step toward understanding potential economic opportunities and commercial viability.

As mentioned previously, significant profits can be realized when a company makes large volumes of a particular item. However, even though many companies want to produce for customers that place large orders, at the same time, customers want assurance that a technology will be successful before they place a large order. These issues can be a force pushing small companies into merging or being acquired by larger entities to achieve economies of scale—many strategies exist for creating value with acquisitions (Goedhart et al. 2017). However, these larger companies may not be interested in mid-scale manufacturing if they are seeking billions of dollars in earnings, which could effectively eliminate mid-scale capabilities.

Because mid-scale manufacturing by nature does not typically involve high-volume production, learning about mass customization and new production capabilities could be leveraged to improve economic prospects. The demand for mid-scale, near-production-scale processing and test equipment is increasing, not only for technical capabilities but also for economic benefits. For example, in its Advanced Alloys Signature Center, NETL experiments on this larger scale (DOE n.d.; NASEM 2025b), which is especially important because when scaling to production, the quality of materials that were pure in the laboratory will typically decrease.

Implementing state-of-the-art mid-scale and hybrid R&D pilot facilities that include scaling would be beneficial. These facilities, run by highly skilled technical associates, could be used for initial production to de-bug processes and de-risk industry deployment, which could motivate increased investments and enhance U.S. competitiveness.

Roadmapping with long-term vision (i.e., more than 5 years) and that includes consideration for products, processes, markets, and talent is another important strategy to anticipate and accommodate economic roadblocks, plan investments, and manage risk; however, identifying best practices for balancing roadmap consensus and manufacturing innovation is a critical first step. By developing a decision-making framework with scale-up as a central focus, one can understand and make commercial investment decisions, which are often based on gross revenue projection (NASEM 2025a), and formulate a strategic direction.

Within such a framework, critical decision making is essential to advance key materials and technologies in an economically viable way. One would ensure that policies and guidance for technology accelerators and PPPs follow the minimum viable product model and that projects that do not meet milestones are culled; planning for these technology off-ramps is critical (NASEM 2025b). One would then think critically about the potentially permanent impact of early decisions on technology maturation and the market and be prepared for possible redirection early in the process to save money. Creating dual-use technologies may lower costs in many instances, but some energetics materials that are essential for the three services do not have a market in the commercial sector and an alternative strategy is needed.

As companies grapple with these challenging decisions about economically viable pathways, they will have to decide whether to pursue Small Business Innovation Research (SBIR) program funding (SBA n.d.) or venture capital funding, which address different needs and target different stages of the innovation process (NRC 2008b). Data tools that help “predict the future” could be useful. For example, Crunchbase¹ is an example of a tool that could be used to better understand investment activity and to consider potential opportunities for mid-scale manufacturing (NASEM 2025a).

Incentivizing investments from industry is also challenging. Small businesses need to invest to procure the right equipment, as successful production operations rely on having the latest technologies, while large companies might not be interested in the limited scale of a mid-sized project. These businesses also need to invest to retain the right workforce with the know-how to use these technologies. DTs, which were discussed previously as potential tools to address technical challenges, could also be leveraged to overcome the economic challenges of securing this needed hardware and the related investments—especially because hardware becomes obsolete so quickly and needs to be upgraded continually.

Standards and economic incentives (e.g., in the form of tax deductions and subsidies) would also help to increase these needed investments and better support mid-scale manufacturing (NASEM 2025a). Some companies are focusing on doing small- to mid-scale manufacturing abroad before moving the manufacturing of their technology, with the support of incentives, to one or more locations in the United

__________________

States. For instance, TS Conductor moved its technology from China to California and is expanding into South Carolina.² This example, though on a larger manufacturing scale, demonstrates that manufacturing capability, and consequently a supply chain, can be imported successfully into the United States. However, incentives would need to be tailored specifically to mid-scale manufacturing to address many of the aforementioned economic challenges that are unique to mid-scale manufacturing and to enable such a move.

Policy Challenges

Even if these technical, operational, and economic challenges are addressed, policy challenges create another potential barrier to success in mid-scale manufacturing. The federal government’s efforts overall could be better coordinated, as described below, particularly in terms of its approaches to contracting and regulations, R&D ecosystems, workforce development, and best practices related to IP and data sharing.

One specific policy challenge for mid-scale manufacturing relates to the difficulty of working within time-consuming and complex government acquisition processes. Many companies, particularly small and medium-sized businesses, hesitate to partner with the government for this reason (NASEM 2025b). Additional issues also arise over the complexity of export control and International Traffic in Arms Regulations. Furthermore, regulatory complexity, especially in energetics and chemicals, often creates barriers to manufacturing innovation and scale-up (NASEM 2025a). The Defense Advanced Research Projects Agency offers one potential model of success for accelerated processes. More versatile and coordinated regulatory frameworks as well as more streamlined and flexible contracting approaches (including accelerated payments to contractors) throughout government agencies are critical to further mid-scale manufacturing in the United States, especially in terms of enabling small and medium-sized businesses to better work with the federal government.

Policies that prevent access to equipment and infrastructure in some government laboratories can also be problematic. Increased coordination and opportunities to increase access—including the funding to do so—could help expand mid-scale manufacturing. Policies that encourage dual-use technologies, when appropriate, rather than prohibit them, could also expand the three services’ access to and use of advanced technologies at the right scale to best support the warfighter.

PPPs in particular foster an ecosystem for students, faculty, researchers, members of industry, government agencies, and entrepreneurs to innovate and collaborate. Better understanding the unique roles of these stakeholders (e.g., the government can contribute knowledge, facilities, and investments; universities can contribute ideas and workforce; and companies can develop a business model for production) could lead to more effective partnerships (NASEM 2025a). Important opportunities for R&D in the United States include federally funded R&D centers, such as the Lincoln Laboratory;³ university-affiliated research centers (UARCs), such as the Johns Hopkins University (JHU) Applied Physics Laboratory (APL);⁴ and the Manufacturing USA

__________________

institutes.⁵ The Manufacturing USA institutes in particular aim to revitalize open collaboration (e.g., by increasing access to capabilities) and to accelerate innovation (NASEM 2025a,b). However, work is needed to determine whether these partnerships and institutes are functioning at their best capacity or if any barriers to progress remain for mid-scale manufacturing efforts; government policies and investments could play an important role.

PPPs can take several different forms: one is the R&D Consortia and Innovation Institutes style, where public agencies (i.e., the Department of Energy and the National Institute of Standards and Technology [NIST]) co-fund industry-led consortia focused on advancing new materials, processes, or manufacturing technologies. Universities often play a central role in these partnerships. The United States already has a strong track record (e.g., Manufacturing USA institutes, National Science Foundation [NSF]-supported consortia), whereby government funding lowers risk, industry co-invests, and academia provides R&D capacity. Countries like Germany (Fraunhofer Institutes) and Japan (METI consortia) also excel here, but the United States tends to rely more heavily on competitive federal funding and industry leadership than centralized public institutes.

A second type of PPP comprises shared testbeds and demonstration facilities. These are facilities where firms can access advanced equipment, pilot lines, or process–scale-up infrastructure without bearing full capital costs. These PPPs fill a unique gap for mid-scale manufacturing by offering neutral, shared platforms that some firms would not acquire individually. U.S. examples include national laboratories and some Manufacturing USA pilot lines, while Europe and East Asia often embed testbeds within government-backed applied research centers.

A third type of PPP is the workforce development partnership. This is a collaborative program between community colleges, universities, workforce boards, and manufacturers to align curricula, apprenticeships, and training with industry needs. U.S. suitability for this type of partnership is very strong. The decentralized U.S. education system makes PPPs essential to coordinate local training pipelines, and industry co-design of programs ensures relevance. In contrast, Germany’s apprenticeship model is highly centralized and government directed, whereas the United States relies more on regional partnerships that reflect local industry needs.

A fourth type of partnership is the cluster development and regional ecosystem partnership. This partnership combines public investment in infrastructure (e.g., transport, broadband, research centers) with private investment in facilities, hiring, and supply chains, often organized through regional consortia. This partnership can be effective when linked to regional innovation funding (e.g., CHIPS Act regional hubs, Economic Development Administration tech hubs) and could be a strong fit for advanced materials clusters. Here, the United States leans on competitive grant models to seed clusters, whereas countries such as South Korea and Singapore use more centralized industrial planning.

The fifth partnership of relevance is the concession and infrastructure PPP. Such a partnership leverages long-term agreements wherein private firms build and operate major infrastructure (e.g., ports, industrial

__________________

parks). This type of partnership is less common in manufacturing: the regulatory environment in the United States makes long-term concessions less attractive compared to in Europe or Asia, where governments play stronger roles in industrial land development and infrastructure leasing.

A key component of accelerating innovation in these partnerships in the United States includes enhancing workforce development to fill skills gaps. A shortage of skilled labor—from factory floor technicians to PhD-level scientists—is ever present; we continue to create jobs that we cannot fill. Possible solutions include either training a new workforce with the same skills as the current and rapidly disappearing workforce or determining new ways to engage the current workforce. Policies to enable training programs that emphasize transferable skills and hands-on experience are important to address the technical, operational, and economic challenges of mid-scale manufacturing. If the skilled labor force could be grown, its applicability would extend across many mid-scale manufacturing companies and ensure long-term stability for both labor and the industry.

A key component of increasing collaboration in these partnerships relates to the ownership, management, protection, and sharing of IP and data more broadly, especially for dual-use and defense-related technologies (Wiseman 2020). Policies and other best practices that support secure, anonymized, and shareable data are needed to overcome hurdles in smart manufacturing and AI integration, for example (NASEM 2025a). Useful examples of how to best protect, curate, and use data from PPPs are readily available. For example, Germany’s Plattform Industrie 4.0 PPP harmonizes data structures, with particular attention toward “sovereign data spaces” (i.e., interoperability frameworks) and €156 million annual investment from the German government (Federal Ministry for Economic Affairs and Climate Action and Federal Ministry of Education and Research n.d.). Furthermore, an international advisory board includes several countries working together to create a mission and governance for data sharing, and the Open Data Alliance is implementing interoperability standards to enable data sharing. The U.S. government would benefit from joining all of these conversations (NASEM 2025b), as policy changes at the government level could enable a trusted data environment for various areas of mid-scale manufacturing.

Case Studies with Unique Considerations for the Three Services

The following case studies present the experiences of select PPPs, a UARC, private companies, and government laboratories, including their approaches to addressing some of the technical and operational, economic, and policy challenges associated with mid-scale manufacturing (NASEM 2025a,b). These case studies illuminate strategies to ensure agility and resilience, which could help leadership within the three services better prioritize government investments in the future.

Public–Private Partnerships

PPPs enable effective collaboration among the government, academia, and industry, which is critical to the success of mid-scale manufacturing. The shared infrastructure and goals of a PPP can accelerate all types of innovation.

For example, the New York Center for Research, Economic Advancement, Technology, Engineering and Science (NY CREATES) is a PPP within the semiconductor industry that was established in 2001 by IBM and the State of New York and is run by the State University of New York at Albany and the state’s research foundation. Multiple partners leverage shared facilities via the Albany NanoTech Complex in an effort to bridge the valley of death by working on advanced logic process technologies, memory technology neuromorphic computing, nano-biodevices, packaging technology, heterogeneous integration, integrated photonics, quantum technologies, and power electronics (NY CREATES n.d.). Essentially, the entire supply chain of the chip industry can work together via this single location and decrease the cost per transistor at a significant rate. NY CREATES has a sustainable, large-scale business model that centers on using a hub-and-spoke model to manage IP: its industry partners retain IP while the state focuses on economic development. Engagement with NY CREATES can take several pathways, from direct contact to participation at its organized events such as the Corporate Venture Exchange, which uniquely matches corporate investment groups with curated early-stage companies. The first technology developed in this facility that went through the Valley of Death was the 22-nm complementary metal-oxide-semiconductor technology. Over the years, NY CREATES has advanced through seven nodes: 22, 20, 14, 10, 7, 5, 3, and 2. In 2021, NY CREATES invented and produced devices for the first time at 2 nm.

In the near future, a high–numerical aperture extreme ultraviolet center in the complex, which will be supported in part by federal funding, will allow exploration below 1 nm (NASEM 2025a).

Another example of a successful PPP is NextFlex, a Manufacturing USA institute. In the realm of hybrid electronics manufacturing, NextFlex advances technology and manufacturing capabilities, establishes and grows the U.S. manufacturing industrial base, and secures human capital for manufacturing capacity. NextFlex has a technology hub that functions as a pilot facility, with 10,000 square feet of laboratory space. NextFlex also serves as a consortium, with more than 200 members and partners in materials, equipment, manufacturing, and products (NextFlex n.d.). NextFlex has 11 technology working groups with 5-year roadmaps that validate market needs; nationwide workforce development programs; workshops and webinars; and project calls for miniaturization, integration, ruggedization, and automation to accelerate hybrid electronics innovation. Some of the hybrid electronics–enabled systems that NextFlex has prototyped for handoff to manufacturers include solutions for security access and control, condition-based maintenance, personal physiological monitoring, and in-theater platoon situational awareness. Furthermore, all NextFlex equipment sets can exchange data, demonstrating the value of a complete DT in enhancing both modeling and the speed of decision making (NASEM 2025a).

Technology Transitions

Accelerating technology transitions is a key component of successful mid-scale manufacturing. For example, as a UARC, JHU APL focuses on developing and transitioning nano-engineered thermoelectric materials for the three services and the National Aeronautics and Space Administration (NASA). JHU APL moved this one-platform technology from basic research to applied research to technology transition over the past 10 years. For instance, in 2014, JHU APL implemented metal-organic chemical vapor deposition equipment and molecular beam epitaxy to grow thin nano-engineered thermoelectric films. In 2017, JHU APL created portable power units and soon after invested in device fabrication. In 2022, JHU APL invested in spark plasma synthesis to create a one-step manufacturing process for NASA missions, with devices running continuously at 1000°C for 3 years. JHU APL has learned that appropriate facilities, material validation, government funding, and multi-organizational collaboration are critical to enable these higher TRL demonstrations (NASEM 2025b). JHU APL’s Technology Transfer Office then works with scientists and inventors to transfer knowledge and technology innovations to the industry responsibly, quickly, and transparently (JHU APL n.d.). While JHU APL’s R&D tasks are challenging, they serve as part of validation and reduction to practice processes, which have been critical to its efforts to support technological advances for the three services (NASEM 2025b).

Opportunities in Entrepreneurship

Entrepreneurial endeavors also offer several opportunities to enhance the future of mid-scale manufacturing. Peak Nano’s journey showcases the effort required to transition projects to products. When making this type of transition, one has to think about a problem differently and change a model or process to succeed. Peak Nano succeeded by using an AI system for optics design: instead of changing how optics work, it changed how and how fast optics are implemented. As a result, the commercial version of the NanoPlex™ product emerged (Peak Nano 2024). The company consolidated the lenses of an optic system using a layered gradient refractive index. The results were a reduced focal point, fewer parts and less complexity, reduced size and weight, and increased clarity and alignment. Optimizing this technology addressed a key problem: the neck strain and injuries warfighters dealt with from the weight of the night vision gear in their helmets. Customers “bought” reduced neck torque, not better optic performance. Translating technological capabilities to warfighter needs such as optimized situational awareness, increased operational tempo, faster identification of targets, increased lethality, improved survivability, and reduced soldier burden is critical (NASEM 2025b).

Peak Nano’s success exemplifies how to manage venture capitalist relationships, partnerships, and other opportunities. Venture capitalists tend to invest in the same opportunities as their peers, and they want to know about the market demand, whether a company can scale, and if the supply chain is well understood. Peak Nano’s “playbook” for successful entrepreneurship includes thinking about a project as a product from the beginning, “testing before jumping” and talking to as many people as possible, engaging in competitive analysis (i.e., why your roadmap is best), understanding transformation alignment (i.e., fitting your technology into a company’s roadmap), having go-to-market champions, and doing phased scaling. Using a “test matrix” that aligns markets with champions, transition potential, regulatory issues, competition,

patents, disruption, price, channels, manufacturing, and supply chain is an effective strategy to eliminate ideas that will not be profitable (NASEM 2025b).

Other Models of Success

Several other models of success that could be leveraged to improve mid-scale manufacturing exist both in private companies and government-sponsored facilities, with the ability to explore, create, and scale new materials, processes, and technologies.

Synthio Chemicals, Inc.,⁶ founded in 2017, is an important case study for mid-scale manufacturing. Synthio works with exothermic reactions and hazardous feedstocks at the scale of tens of tons per year of production, which allows the company to offer both a safe work environment and a competitive advantage. To address the scaling challenges that arise in traditional batch manufacturing, especially in terms of aligning technology and manufacturing readiness levels, Synthio implemented continuous flow manufacturing for production. Continuous systems⁷ have a smaller footprint at scale with smaller systems, a smaller hazard profile, reduced labor costs owing to automation, and better quality; these systems are also easier to scale, and they help decrease risk going into production. Synthio is also involved in a Small Business Technology Transfer (STTR) program with Purdue University to create a modular system for the agile manufacturing of energetic materials; reconfigurable unit operations with a common interface could enable the creation of energetic materials without the need to develop new systems (NASEM 2025a).

Also founded in 2017, SkyWater Technology⁸ manufactures products to support both commercial and defense programs. SkyWater employs a technology-as-a-service (TaaS) model in which it collaborates with customers to co-create differentiated technologies and then ramps them up for manufacturing—all in one facility. With more efficient R&D, accelerated time to market, and volume manufacturing, this TaaS model streamlines the path from concept to production. SkyWater has been successful in integrating both technology and manufacturing in mixed-signal complementary metal-oxide semiconductors, radiation-hardened electronics, microelectromechanical systems, superconductors, photonics, and heterogeneous integration. As a “technology foundry” with both advanced technology services and wafer services, SkyWater epitomizes mid-scale manufacturing by supporting laboratory-scale volumes and manufacturing higher volumes within the same footprint. TaaS breaks high-volume processes into specialized modules for new applications, reinforcing SkyWater’s ability to adapt and scale differentiated technologies rapidly (NASEM 2025a).

In addition to the applicable progress of these private companies, lessons can also be learned from governmental entities that engage in innovative approaches. For instance, the National Renewable Energy Laboratory (NREL) oversees several current and future approaches that could be applied to address the challenges of mid-scale manufacturing. For example, the Bio-Optimized Technologies to keep

__________________

Thermoplastics out of Landfills and the Environment (BOTTLE) Consortium⁹ explores inputs and outputs of advanced polymer recycling and upcycling and develops new recycling technologies, with more than 100 scientists and engineers from national laboratories and universities working in deconstruction technology, creating building blocks, redesigning new materials, and helping companies solve their existing problems. Furthermore, NREL’s Energy Materials and Processing at Scale facility, expected to be completed by 2027, could help address the challenge of moving from the laboratory to piloting, in this case for new renewable energy and recycling technologies (NASEM 2025b).

Furthermore, NSF’s Materials Innovation Platforms focus specifically on the needs of mid-scale manufacturers by “support[ing] mid-scale infrastructure to advance materials discovery and development through integrated materials synthesis, characterization, manufacturability, scaling from batch to continuous processing, toxicity assessment and modeling while promoting collaboration, knowledge sharing, and access to state-of-the-art tools” (NSF n.d.). Each platform is described as a “scientific ecosystem” of scientists and external users, across disciplines, who work together as well as share cutting-edge tools, data, and know-how. This initiative offers great promise; however, it is currently accessible only to those selected from a competitive application process. More open-access initiatives would have significant impact on mid-scale manufacturing. For example, Oak Ridge National Laboratory (ORNL) offers different tiers of access, some of which include general users, to its beamline facilities (ORNL n.d.).

Opportunities for the Three Services

Given the many challenges identified for the future of mid-scale manufacturing, as well as the numerous examples of strategies to address them, the landscape points to additional new opportunities. Otherwise, without continued adoption of new technologies and approaches, the United States could struggle to compete. The three services could thus benefit from the following six considerations to help guide leadership in decision making and in prioritizing investments that enhance mid-scale manufacturing in the United States (NASEM 2025b):

1. Ensure that speedy and flexible funding is available. The first key to successful mid-scale manufacturing is adequate and readily available funding that supports the need for rapid iterations. To scale to appropriate levels of financing, tax breaks and other financial incentives are critical. The second key to successful mid-scale manufacturing relates to leveraging partnerships appropriately (e.g., the Manufacturing USA institutes). Instead of building new facilities, the facilities within these partnerships can be used for scale-up, and their technologies can be networked (NASEM 2025a,b), thus offsetting some expense. However, analyzing the value proposition for each partnership first is critical.
2. Expand pilot facilities. Although facilities already exist to test and produce large quantities of material, the development of new and continued use of pilot facilities (e.g., NETL, LIFT, national laboratories within the three services, ORNL’s accessible beamlines, and large three-dimensional printers) would

__________________

be especially beneficial for testing and production at the mid-scale. For example, facilities are needed to cast, roll, forge, and produce certain alloys. Some electrochemical systems used to manufacture batteries are also limited, as are reel-to-reel processing facilities; without the ability to test how well materials can be produced, quality could decrease and risk could increase. In some cases collective mid-scale capabilities could be coordinated to increase production volumes during development phases. These could be early-phase assessment and testing and later-phase scale-up assessment and production planning. There is also the production itself, when materials are manufactured across the mid-scale base in coordinated ways (i.e., modular, distributed).

3. Enhance workforce development. The skilled workforce is also limited for the processing described above; both an upfront investment and a longer-term commitment could help alleviate this problem. The three services could learn lessons from the Catapult Network in the United Kingdom,¹⁰ for example. With a dedicated portion of its funding coming from the government, it can incentivize and retain the appropriate individuals for such critical work. Generally, supporting mid-scale manufacturing requires a blend of technical expertise and operational planning. Materials scientists and process engineers are also essential to move innovations from the laboratory scale to reliable mid-scale production. Yet companies doing mid-scale manufacturing often struggle to recruit for their workforce, as they compete with firms doing large-scale manufacturing for scarce technical talent. Furthermore, they sometimes lack the resources to sustain in-house training pipelines, facing geographic and wage pressures that make it more difficult to attract and retain highly skilled staff. With manufacturing moving quickly toward digitalization and automation, another risk exists that firms doing mid-scale manufacturing are operating with legacy systems. Therefore, enhanced workforce education in general is needed to bolster mid-scale manufacturing in the United States (NASEM 2025b).
4. Inventory and provide access to current tools and capabilities. Enhancing mid-scale manufacturing includes creating and providing access to defense-specific tools and capabilities. For example, comprehensive inventories of relevant facilities could be developed so that people working within and beyond the government become aware of the numerous innovative facilities that exist and that could be leveraged to further mid-scale manufacturing efforts. Such an inventory could also readily illuminate gaps in defense technologies, capabilities, and facilities—as well as opportunities for dual-use approaches—which would help to prioritize both funding and next steps. Furthermore, to ensure access to both the knowledge generated by these tools and capabilities and the physical structures that house them, reasonable cost and support personnel are critical. Overall, increased organization and coordination throughout the three services is critical to achieving such goals. Additionally, involvement by mid-scale manufacturers with existing facilities, consortia, and cooperatives can help if scale-up to larger quantities is rapidly needed for any reason.
5. Support technology transition. The development of portfolios of both current and new manufacturing technologies is key to successful mid-scale manufacturing (NASEM 2025b). Funding to support this initiative could be sought from venture capitalists, SBIR or STTR programs, or existing authorities (e.g., via the Defense Production Act, Title III). Furthermore, assistance with technology transfer planning and implementation is needed for SMMs. The three services could request translation

__________________

plans in initial funding proposals for 6.2 and 6.3 research (e.g., describing an intent to engage the Manufacturing USA institutes or national laboratories to reach higher TRLs). Developing the right metrics to help evaluate the success of a technology transition is a key part of roadmapping the transition process, although such metrics may vary by technology and market. Consulting services could be engaged to assist SMMs; this goal could be realized either via existing facilities (e.g., Manufacturing USA institutes or NIST’s Manufacturing Extension Partnership program) or new entities. Existing Manufacturing USA institutes could also enable translation. Better informatics are key to successful technology transition. With the use of advanced analytics tools and increased analysis of public data (including on patents and via papers and trade information), the performance of individual companies could be assessed more actively, which could empower the three services to make the best investments at a lower level of risk.

6. Enhance digital representation capabilities. Several steps are needed to realize this opportunity for more effective mid-scale manufacturing. In the near term, the three services could provide resources to assist with the development of digital representations—for example, digital representations of facilities could be valuable for maintaining equipment and facilities properly. The three services could also use digital representations to build models and tools, with initial specifications for existing equipment and capabilities for validation, calibration, and uncertainty quantification to reduce risk. A “digital passport,” equivalent to a next-generation technical data package, could feed into such a digital representation.

A main point of this paper has been to provide the three services with an assessment of some of the issues and opportunities in the area of developing critical materials. The ultimate goal is to achieve the objectives while minimizing the cost. Many of these opportunities rely on making better use of existing entities and resources instead of creating entirely new infrastructure and processes. With enhanced facilities and capabilities, improved data collection and management initiatives, more attractive financing strategies, increased workforce development, and innovative dual-use approaches, the three services could more successfully leverage mid-scale manufacturing and better support the U.S. warfighter.