Recalibrating the Defense Industrial Base for Systems Competition

In a 1944 memorandum to Adolf Hitler, German armaments minister Albert Speer described the Second World War as “a contest between two systems of organization.”¹ The conflict, he explained, pitted Germany’s “system of overbred organization against the art of improvisation on the opposing side.” Germany’s “outmoded, tradition-bound, and arthritic organizational system” manifested in exacting requirements and small numbers of high-end, technologically advanced systems.² The American system, on the other hand, had its roots in Henry Ford’s model of mass production. Guided by a combination of patriotic duty and government carrots and sticks, American manufacturers of products ranging from household appliances to automobiles converted their factories to assemble war materiel. Famously, Ford Motor Company built B-24 bombers on a mile-long assembly line at the Willow Run plant near Detroit. By the end of the war, the United States produced more than 300,000 military aircraft, 88,000 tanks, and 1,200 major combatant ships.³ American industrial might crushed the Axis. 

Today, the United States finds itself in the reverse position. By 2030, China’s share of global manufacturing is projected to rise to 45 percent, while the United States’ share drops to 11 percent.⁴ This asymmetry has grave implications for the defense industrial base—China’s position as the world’s sole manufacturing power grants it a superior industrial rate of replacement for munitions, drones, and ships. In a Taiwan contingency, the U.S. military is projected to burn through supplies of precision munitions in just eight days.⁵ The United States can produce an estimated 20,000 to 30,000 drones per year; the Ukrainian Ministry of Defense burned through roughly 10,000 drones per month and has announced the nation can now produce three million drones annually.⁶ In addition, China’s shipbuilding capacity is about 232 times greater than that of the United States.⁷ Taken together, Beijing’s advantages across these sectors grant it the inside track to prevail in a protracted war.

A political and industry consensus has formed that the United States must rebuild its defense industrial base to offset China’s scale-based manufacturing advantage. These calls to action, though critically important, often suffer from both a lack of definitional clarity and an unclear theory of victory. For example, policymakers refer to the defense industrial base as a single entity when, in reality, it consists of a patchwork of tens of thousands of companies and suppliers. Beyond simply reshoring production for key military products and critical materials, intermediate objectives for rebuilding defense production are unclear. A more precise set of goals is needed. In systems engineering, a system of systems (SoS) is any large-scale system made up of many independent, geographically dispersed components. Framing the defense industrial base as a SoS reveals three functions—control, networking, and automation—that policymakers must address to prepare the nation for protracted conflict. This paper explores each function in turn, drawing lessons from U.S. defense mobilization during World War II while acknowledging the fundamentally different nature of twenty-first century manufacturing. By examining defense industrial preparedness through this lens, policymakers can move beyond simplistic calls for reshoring and identify specific intervention points that maximize impact within constrained timeframes and budgets.

Systems Engineering and Defense Mobilization

The field of systems theory developed over the course of the twentieth century as a means of understanding and controlling the technologies and infrastructure on which modern life now depends. One leading theorist defines systems as “collections of different things that together produce results unachievable by the elements alone.”⁸ Systems can be natural or technological, from the process of photosynthesis to the operation of a Columbia-class space shuttle. All systems share a few characteristics: a central goal or purpose, adaptation to changing environments, a series of networked relationships between heterogeneous parts, and emergent (or unpredictable) behavior.⁹  

The U.S. industrial base is certainly an example of societal-scale infrastructure, but framing it as a single, monolithic system risks oversimplification. Producing a single military aircraft or Virginia-class submarine, for instance, requires a multi-tier network of thousands of manufacturers and suppliers. Missile batteries, aircraft, and communications networks that the U.S. military deploys in the field are even referred to as systems in their own right. How should the United States make sense of what appears to be a massive system made up of many smaller systems? Ultimately, the defense industrial base is an SoS. These sprawling forms of technology and infrastructure display unique qualities, including unexpected behaviors that cannot be predicted from individual components, multiple layers of operational autonomy, and complex interdependencies that create both resilience and vulnerability. Other examples of SoS include the internet, global value chains, and air traffic control systems.¹⁰

For all the jargon that systems engineering can introduce, applying it to the defense industrial base sheds light on the core functions that policymakers should seek to strengthen as they shore up defense supply chains. According to systems engineer Daniel DeLaurentis of Purdue University, an SoS can be classified based on three characteristics: degree of control (subsystems controlled by a central entity, and how effective that control is), connectivity (the degree to which subsystems are networked together), and system type (to what degree are parts of a system human versus technological).¹¹ The study of systems engineering encourages policymakers to design better incentives, denser networks, quicker feedback loops, and more resilient subsystems.¹²

These principles of systems engineering highlight a critical insight for defense planning: preparedness requires not just technology and capability, but structural readiness for rapid scaling. This means that scalable systems must be in place before a crisis. During World War II, the U.S. industrial base began to mobilize 18 to 24 months before the United States formally entered the combat—thanks to proactive measures such as the Lend-Lease Act, which committed Washington to provide bombers, fighter aircraft, and ships to support the United Kingdom during the Battle of Britain. Today, a potential Taiwan contingency could be decided in a matter of days or weeks. As former Deputy Secretary of Defense for Industrial Policy Jerry McGinn has written, “effective mobilization requires systems designed for and developed with a principal focus on producibility and scalability.”¹³

Today, however, the U.S. defense industrial base is optimized to produce small numbers of technologically sophisticated, high-end platforms—not to scale production in a protracted conflict. As a recent report argues, “The Pentagon’s ‘lowest price technically acceptable’ ethos, that is, spending not a penny more than is necessary to meet the most basic immediate requirements, has brought damaging secondary effects.”¹⁴ Instead of investing in surge capacity, contractors have long been incentivized to source components from a single producer or facility. Efforts to scale production for Ukraine aid packages have created backlogs for munitions like 155mm artillery shells and Stinger missiles. Plagued by bureaucracy and an inability to scale, the United States finds itself on the losing end of the systems contest which Speer described in 1944.¹⁵  

Control: Balancing Centralization and Market-based Incentives

The central challenge of defense mobilization is the need to conserve and direct scarce resources. This explains why the United States has tended to adopt more centralized planning and allocation of resources during wartime. Writing in 1943, the Truman Committee—a Senate special committee tasked with investigating waste and fraud in the defense program, chaired by then-Senator Harry S. Truman—made this observation in stark terms:

The function of government in peacetime is simply to state the rules under which business will be conducted and to insure that there is fair play between businessmen and a fair deal to the public. In a defense or war program the situation is very different. The Government…must determine what is to be produced, the quantities thereof, the materials to be used in connection therewith, and even the exact methods by which the articles are to be produced.¹⁶

Control of a system means more or less what it sounds like: the ability to tell a system what to do and ensure it responds as intended. In the case of the defense industrial base, today’s generation of policymakers has discovered that they have far less control over the system than previously thought. Sending military aid to Ukraine, for example, meant restarting shuttered production lines for Javelin anti-tank missiles and a two-year backlog to produce 155mm artillery shells. Yet central planning is not typically the answer to reestablishing control over an unwieldy SoS. Because the defense industrial base is made up of thousands of independent entities, attempting to directly manage each component through top-down directives would create inefficiencies, bottlenecks, and resistance. Instead, creating a strong, streamlined control mechanism that accounts for the most important variables, then accounting for other variables by setting market-based incentives, offers a more effective approach to guiding this complex system toward strategic objectives.

The U.S. Department of Defense Systems Engineering Handbook identifies four types of systems of systems.¹⁷ On one end of the spectrum, directed systems are centrally managed to achieve specific purposes. Virtual systems, on the other hand, have no central governing authority at all.¹⁸ The classic example of a virtual system is the internet, which has no central governance mechanism (aside from a set of standards-setting bodies).¹⁹ The defense industrial base sits between these extremes. Since the U.S. defense industry sits within a market economy, meaning firms are not forced to produce defense products but instead voluntarily sell into the Department of Defense, it is an acknowledged SoS. This partial centralization creates a unique set of challenges. In an acknowledged SoS, “constituent systems retain their independent ownership, objectives, funding, development, and sustainment approaches,” explains the Department of Defense handbook.²⁰ Yet these independent systems must somehow be aligned to the goals of Pentagon war planners.

Type	Definition	Example
Directed	Engineered for specific purposes with centrally managed operations throughout their lifecycle. Components can operate independently but are typically subordinated to central control and authority.	Military Command & Control Systems: Integrated air defense networks where radar, missiles, and command centers work under unified command.
Acknowledged	Have recognized shared objectives and designated managers with resources, but constituent systems retain independent ownership, funding, and development. Changes require cooperative agreements between systems.	Emergency Response Systems: Police, fire, medical, and government agencies coordinating during disasters while maintaining separate command structures.
Collaborative	Component systems interact voluntarily to fulfill agreed-upon central purposes through informal coordination mechanisms. Systems choose their level of participation in shared goals.	Supply Chain Networks: Manufacturers, suppliers, distributors, and retailers collaborating to deliver products.
Virtual	Lack central management authority and centrally agreed-upon purposes. Large-scale emergent behavior develops through self-organizing mechanisms, with systems operating autonomously.	The Internet: Interconnected networks operated by different organizations without central control.

Figure 1: Four Types of SoS²¹

This challenge of developing strong yet streamlined control mechanisms is illustrated by the successes and shortcomings of the World War II-era War Production Board (WPB). Throughout the early stages of the war effort, “the major bar to high production goals remained the lack of effective controls over the flow of materials,” explains historian Maury Klein.²² In late 1941, after a series of predecessor organizations failed to alleviate shortfalls in defense production, President Franklin D. Roosevelt established the WPB as an independent government agency with a sweeping set of authorities, including the ability to order U.S. manufacturers to halt civilian production or conserve scarce materials. Limitation orders restricted production of certain products; these were first used to convert the auto industry entirely to war production in early 1942. Material orders, meanwhile, were used to restrict the use of critical materials, such as copper or aluminum.  

In May 1942, American war production remained plagued by shortages in critical materials, despite Roosevelt’s invocation of emergency powers. WPB chair Donald Nelson set out to fix these continued shortages by mandating the use of a sprawling requirements system called Production Requirements Plan, or PRP. Sweeping in its ambition, PRP sought to track and control the flow of at least 90 percent of 36 materials critical to the war effort.²³ But PRP never achieved these lofty goals. The system’s complexity created mountains of paperwork for already strapped contractors, which then had to be shipped to Washington and processed by War Department bureaucrats. Both camps quickly fell behind their reporting requirements, but the more serious problem was that WPB’s directives were simply ignored by the military, which refused to limit its supply orders within the scope of what WPB informed them was available. Nelson ultimately pitched Roosevelt on a joint materials prioritization process between WPB and the Joint Chiefs informed by warfighting requirements. Yet the services, which were unwilling to cede control of their supply orders to the whims of civilian industrial planners, successfully resisted.²⁴ 

PRP’s failure holds a clear lesson: attempting to centrally manage and control every aspect of a complex system of systems is nearly impossible, even with extraordinary wartime powers. Nelson responded by reorganizing WPB in July 1942 and hiring a new materials czar, who proposed replacing PRP with a control system that was simple and realistic.²⁵ Rather than controlling everything, WPB would adopt a streamlined, total allocation system for three primary war materials: carbon and alloy steel, aluminum, and copper.²⁶ 

This experience offers clear lessons for today. In a high-intensity conflict, control mechanisms would be needed to conserve and guide the flow of a variety of critical materials and products, from aluminum and rare earths to brushless motors and battery cells. Yet like the WPB’s updated controls, a streamlined system focused on the scarcest, most vital materials and components would be most likely to work as intended. Mark Maier, a systems engineer who first developed the concept of SoS, refers to this targeted approach to control mechanisms as “policy triage.” ²⁷ “Let the dying die. Ignore those who will recover on their own. And treat only those who would die without help,” Maier writes.²⁸

Systems theory also teaches that thoughtfully designed incentives are often more effective than direct control in coordinating independent components toward common goals.²⁹ Today’s defense industrial base suffers from an overreliance on centralized direction and control mechanisms, such as the Planning, Programming, Budgeting, and Execution (PPBE) process, the Pentagon’s annual cyclical process to determine Department funding requirements and to allocate resources. The PPBE process was modeled off the Soviet Union’s five-year plans during the 1960s.³⁰ While these rigid requirements aim to minimize waste, they inadvertently create barriers to innovation by preventing the Department of Defense from leveraging technologies emerging from America’s startup ecosystem. Despite years of reform efforts to drive the adoption of innovative technologies, spending on products from venture-backed startups remains a tiny portion of the U.S. defense budget. 

To revitalize the defense industrial base, policymakers must restructure incentives both within and outside the Pentagon. Internally, the Department of Defense must incentivize contract officers to embrace calculated risks on scalable, software-defined systems from nontraditional players.³¹ On April 9, 2025, President Donald Trump signed an Executive Order that appears to do just that.³² The order, titled “Modernizing Defense Acquisitions and Spurring Innovation in the Defense Industrial Base,” directs the Secretary of Defense to reform the Pentagon’s acquisition process by prioritizing speed, commercial solutions, and streamlined decision-making.³³ The order also requires focusing performance evaluations on how well employees use efficient procurement methods and take calculated risks to innovate. Externally, market signals through consistent contracting are vital to drive additional investment. These measures are an excellent start. However, performance is measured in contracts. As former government officials-turned defense investors Michael Brown and Pavneet Singh argue, “meaningful and consistent contracts are the capitalist signal to current and future suppliers that more production and more production capacity are necessary.”³⁴ 

Networking: Interoperability, Resilience, and the Arrival of Software-Defined Manufacturing 

At their core, systems of systems are interconnected networks of people, organizations, and technologies. How well these networks communicate determines whether the entire system can function effectively, especially during a crisis. In the case of the defense industrial base, the system hinges on effective wartime communication and interoperability between different players in the supply chain. “The administrative key to mass production,” writes analyst David Novick in a review of the U.S. World War II defense program, “is the painstaking planning, timing, and direction of the flow of materials and parts through the manufacturing process so that each item arrives at the final assembly line where and when it is needed.”³⁵ In the initial year after Pearl Harbor, poor networking proved a major bottleneck to the American defense buildup: 

Some shipyards hoarded an eighteen-month supply of steel while the building of escort vessels to defend against submarines was delayed for lack of steel. Critical machine tools went to the wrong firms, depriving those who needed them urgently. Some components were underordered, others overordered, causing bottlenecks. No agency tried to coordinate and reconcile the competing demands for materials, parts, tools, and other components.³⁶

The U.S. industrial base would not fare much better in a protracted conflict today. Policy experts have rightly highlighted the lack of planning mechanisms for defense mobilization, but the deeper obstacles are technological: much of the U.S. manufacturing sector runs on outdated, 1980s-era technology protocols and software which are incompatible with one another. The arrival of cloud computing and enterprise software transformed how large companies collect, manage, and use their data, but the manufacturing and defense sectors largely missed the boat. Many smaller manufacturers have not even connected their systems to the internet, meaning they operate in complete informational isolation and cannot share real-time production data, inventory levels, or capacity availability with potential defense contractors or with government agencies during a crisis.³⁷ Similar challenges also extend to the Department of Defense’s logistics enterprise. Currently, the Department of Defense has six separate enterprise resource planning (ERP) systems for defense logistics—one for each military service or department—but none of these systems can communicate with one another. To make matters worse, orders are placed using non-classified routers.³⁸ 

The application of twenty-first century information and communications technologies to manufacturing offers a path to fix the Pentagon’s broken networking function. Advanced manufacturing technologies, such as additive manufacturing and AI-enabled 3D computer-aided design (CAD) systems, enable rapid prototyping and faster design and production cycles. Most importantly, applying modern software engineering practices to industrial production offers an opportunity to stitch together fragmented defense supply chains. As Palantir Chief Technology Officer Shyam Sankar posits:

What if companies could create a superstructure of software over the entire, fragmented production process…software would absorb and analyze data from the company’s countless suppliers, components, machines, and workers to create a complete model of the production process. And it would control physical machines, like industrial robots and machine tools, allowing refinements to the production process on the fly.³⁹ 

The demonstrated success of software-enabled production networks at the company level suggests that the Department of Defense could theoretically launch a Manhattan Project-style initiative to create a comprehensive digital map of the entire defense industrial base. Nearly all the technology needed to digitalize and map defense supply chains exists today.⁴⁰ Companies like Govini have developed AI-powered platforms that can map connections across defense supply chains to identify foreign ownership risks, analyze financial health of suppliers, and assess production capacity vulnerabilities across thousands of defense contractors. Similarly, manufacturing software startup Sustainment uses its software platform to connects users with U.S. manufacturing suppliers to increase throughput and streamline procurement workflows, digitizing market research processes to strengthen the defense industrial base. The primary hurdle to creating a holistic picture of the defense industrial base is data availability: the National Defense Industrial Association, the leading industry group representing defense firms, acknowledged last year that no one inside or outside of the Department of Defense has accurate data on the true size or number of firms which make up the defense industrial base.⁴¹ Gathering this data is a vital step towards mobilization.

China also appears to be building such a system. The project, called the Industrial Internet, aims to track the flow of goods through the supply chain in real-time.⁴²  It has received funding via a series of national and provincial level implementation plans and includes a series of pilot projects for 5G-enabled factories.⁴³ Given the deep ties cultivated between China’s manufacturing sector and the People’s Liberation Army (PLA) under Beijing’s Military-Civil Fusion strategy—which aims to harness China’s manufacturing sector to support its defense enterprise—the Industrial Internet could offer Beijing a high-tech system to monitor and control its own defense mobilization.⁴⁴ 

Once an AI-enabled map of defense supply chains is built, the next step would be to use it to fill gaps in production capacity during a conflict. However, one basic obstacle to this goal is the lack of interoperable design files for parts. If a machine shop producing metal parts for a defense contractor attempted to send a digital file of a metal part to another machine shop, the second supplier would most likely be unable to read the file and produce the part. These simple CAD software incompatibility issues could prevent the Department of Defense from mapping and allocating resources between firms during a protracted conflict, creating a major obstacle to rapidly scaling production. To enable dynamic shifting of production capacity during conflicts, the Department of Defense must foster compatibility of CAD and computer-aided manufacturing (CAM) systems across multiple tiers of the supply chain. The STITCHES (System-of-systems Technology Integration Tool Chain for Heterogeneous Electronic Systems) program, a recent Department of Defense effort that solved crippling incompatibility issues in the Pentagon’s command and control system, offers a potential model. STITCHES created an innovative middleware solution, essentially a software translation layer, that enables thousands of different sensors, weapons systems, and platforms to communicate with each other without requiring modifications to their original code.⁴⁵ 

Lastly, systems theory highlights the importance of resilience—the ability to absorb shocks and maintain critical functions during disruptions. Applying software to the defense industrial base heightens the urgency of hardening the defense industrial base against cyber attacks. Manufacturing has already been the number one most targeted industry for attackers in recent years.⁴⁶ Adversaries would almost certainly target manufacturing systems early in a conflict: a recent assessment found that by launching “just 25 well-constructed attacks, using any of a variety of means, an adversarial military planner could cripple much of America’s manufacturing apparatus for producing advanced weapons.”⁴⁷ PLA military theorists frequently describe systems destruction warfare as the “basic operational mode” for future conflict.⁴⁸ While much of the literature surrounding this concept focuses on adversary attempts to disable U.S. command and control systems, widespread PLA intrusions into allied industrial systems through attacks such as Volt Typhoon, a recent series of sweeping PLA hacks into U.S. critical infrastructure, suggest that defense production will also be targeted.⁴⁹

Autonomy: Know-How, Automation, and the Skilled Worker Shortage

The third element of an SoS is the system type, or the balance between machine and human elements in a system. A core question in systems engineering is whether and how to automate key subsystems. Today, the defense industrial base relies on mostly human labor, and this is a limiting factor: the United States faces a massive shortage of workers across its manufacturing sector. By 2030, this gap is projected to widen to about 2.1 million workers.⁵⁰ A lack of skilled labor is a major factor behind persistent delays in U.S. shipbuilding, including for Columbia– and Virginia-class submarines. Recent advances in robotics offer the potential to address workforce shortages, but automation must be balanced with manufacturing process innovation—an underrated wartime necessity which cannot be replicated by machines.    

Skilled labor became a significant bottleneck to U.S. war production in the year after Pearl Harbor. President Franklin Roosevelt highlighted the issue during an October 1942 fireside chat, likening workforce shortfalls to shortages of critical materials: “The problem,” he explained, “is to have the right numbers of the right people in the right places at the right time. We are learning to ration materials; and we must now learn to ration manpower.”⁵¹

Roosevelt’s remarks underscore a key difference between twentieth and twenty-first century manufacturing: training millions of workers during World War II was mostly a numbers game. Most manufacturing-related skills could be picked up on the job. Today, however, the problem is not simply the number of workers available, but the specialized know-how they need to possess. This is a major factor which underpins China’s scale-based manufacturing advantage over the United States. As Apple CEO Tim Cook opined in 2017: “In the U.S., you could have a meeting of tooling engineers and I’m not sure we could fill the room. In China, you could fill multiple football fields.”⁵² Fixing this issue could take years and might require overhauling the U.S. education system to emphasize apprenticeships and hands-on learning.

Instead of tackling shortages of highly skilled engineers head on, some observers have proposed aggressively automating U.S. manufacturing. One approach would design the human out of the loop entirely by creating so-called lights-out factories. These facilities would require little human intervention and could be controlled remotely.⁵³ A growing number of scalable, technologically advanced “factories of the future” already exist in various stages of construction today.⁵⁴ Defense startup Anduril announced in August 2024 that it would build Arsenal-1, a highly automated five million square foot production facility “designed to produce tens of thousands of autonomous military systems annually.”⁵⁵ Manufacturing startup Divergent3D has also developed a highly automated advanced manufacturing system which combines additive manufacturing and AI to produce unmanned systems for a range of defense primes.⁵⁶

Another approach would involve strategically deploying millions of robots to address critical manufacturing gaps. Tech luminaries like Elon Musk and Jeff Bezos have called for the widespread deployment of humanoid robots, or general-purpose robots built to resemble human form-factor, but the United States also trails in deploying more basic industrial robots. U.S. factories maintain only one robot for every 66 workers, significantly lagging behind global competitors.⁵⁷ The case for robotic automation is compelling: due to persistent skilled labor shortages, most U.S. factories operate at just 30 percent utilization, resulting in millions of lost production hours annually.⁵⁸ Given current workforce constraints and the exponential production demands that would emerge during national security crises, rapidly scaling robotic capabilities would not only maximize production efficiency and lower costs but also ensure America’s defense manufacturing base can meet surge requirements. 

Automation is not a panacea for fixing U.S. defense production. It does carry clear downsides. Most importantly, automating a process actually risks stifling what manufacturing experts refer to as process innovation, or tweaks to manufacturing process flow or product design which can unlock efficiency gains or novel functions. Elon Musk has shared a five-step framework for refining manufacturing processes; only during the last step is automation applied.⁵⁹ Most gains in manufacturing are made because of creative observations on how to improve process flow—observations only a human can make. Tesla’s use of a giant aluminum stamping press, or Giga Press, provides one tangible example. Replacing front and rear assemblies with giant, single piece die castings allowed the company to phase out 370 parts. In an ironic twist, Tesla replaced about 300 robots thanks to an innovative idea pursued by humans.⁶⁰

Defense programs are fundamentally about creating lethality at scale—job creation, while beneficial, remains a secondary consideration. Ultimately, however a range of worker augmentation technologies already exist which lower barriers to entry for new hires. Policymakers may not need to choose between full automation or a lengthy overhaul of defense workforce development. Industrial AI assistants embedded in manufacturing equipment can dynamically adapt instructions based on a worker’s experience level, effectively creating personalized on-the-job training programs that accelerate competency. Collaborative robots designed for safe human interaction can handle precise, repetitive tasks while humans focus on higher-order process improvements and quality control. Perhaps most promising are augmented reality interfaces that overlay digital instructions onto physical workspaces, allowing novice workers to perform complex operations by following visual cues. These technologies could also allow rapid workforce expansion with minimal training overhead, addressing one of the most persistent bottlenecks to scalability.

Conclusion

U.S. policymakers generally agree that the nation must rebuild its defense industrial base, but the ways and means to do so have been less clear. Framing defense production as a system of systems provides a useful conceptual lens that identifies the highest-impact areas for policy intervention. This approach reveals that rather than attempting comprehensive control over every aspect of defense manufacturing, policymakers should focus on strategic “policy triage”—controlling only the most critical materials and processes while using market incentives to guide the rest of the system. Creating interoperable software networks across defense manufacturers would enable both rapid production scaling and dynamic reallocation of manufacturing capacity during conflicts. Meanwhile, a balanced approach to automation and human labor—leveraging worker augmentation technologies rather than pursuing either full automation or traditional workforce development alone—would help address critical skill shortages. 

Historical parallels to World War II mobilization efforts are instructive but insufficient. Unlike the 1940s, today’s manufacturing challenges are defined less by raw production capacity and more by technological complexity, specialized knowledge requirements, and digital integration. China’s manufacturing dominance represents a fundamental asymmetry that cannot be overcome through incremental improvements or isolated policy interventions. A comprehensive systems-based approach is necessary to create a defense industrial base capable of deterring conflict and sustaining production during protracted war. 

Time is not on America’s side. Experts estimate that the window to meaningfully transform the defense industrial base may be just three to five years—well within potential timelines for heightened tensions over Taiwan. Policymakers should prioritize immediate investments in digital infrastructure for supply chain mapping, worker augmentation technologies to quickly scale skilled workforces, and pilot programs for software-defined manufacturing. Meanwhile, longer-term initiatives to rebalance defense procurement incentives and foster a new generation of manufacturing talent must begin in parallel. Creating a twenty-first century industrial base optimized for scalability means reimagining how defense systems are designed, produced, and networked. By addressing the core systems functions of control, networking, and autonomy, policymakers can transform America’s defense industrial base from a vulnerability into a strategic advantage.

Brady Helwig is a geopolitical analyst based in Washington, DC. His research focuses on advanced manufacturing, semiconductors, and U.S.-China decoupling issues. He graduated from Hillsdale College with a bachelor’s degree in politics. Brady is an alumnus of SSS China.

---

Image: Looking up one of the assembly lines at Ford’s big Willow Run plant, where B-24E (Liberator) bombers are being made… – NARA – 196389, February 1943, from U.S. National Archives and Records Administration. Retrieved from: https://commons.wikimedia.org/wiki/File:Looking_up_one_of_the_assembly_lines_at_Ford%27s_big_Willow_Run_plant,_where_B-24E_(Liberator)_bombers_are_being_made…_-_NARA_-_196389.jpg, used under Wikimedia Commons.

[1] Albert Speer and Eugene Davidson, Inside the Third Reich: Memoirs by Albert Speer, trans. Richard and Clara Winston (New York: MacMillan Company, 1970), 216.

[2] Speer and Davidson, Inside the Third Reich, 216. 

[3] Maury Klein, A Call to Arms: Mobilizing America for World War II (New York: Bloomsbury Press, 2013), 516;Mark C. Faram, “Midway, Momentum, and Manpower – The Navy’s Bureau of Personnel in World War II,” U.S. Navy, June 3, 2021, https://www.navy.mil/Press-Office/News-Stories/Article/2645303/midway-momentum-and-manpower-the-navys-bureau-of-personnel-in-world-war-ii/. 

[4] United Nations Industrial Development Organization, The Future of Industrialization: Building Future-Ready Industries to Turn Challenges into Sustainable Solutions, MIPF 2024 Conference Paper, November 22, 2024, 17, https://www.unido.org/sites/default/files/unido-publications/2024-11/The%20Future%20of%20Industrialization%20-%20Building%20Future-ready%20Industries%20to%20Turn%20Challenges%20into%20Sustainable%20Solutions.pdf. 

[5] Seth G. Jones, Empty Bins in a Wartime Environment: The Challenge to the U.S. Defense Industrial Base (Washington, DC: Center for Strategic and International Studies, 2023), 11, https://csis-website-prod.s3.amazonaws.com/s3fs-public/2023-01/230119_Jones_Empty_Bins.pdf?VersionId=y_iEwCalRVFiVedETHwrcuwDaenf7zez. 

[6] Bing West, “Drones Reduce US Military Capabilities,” Strategika, March 15, 2024, https://www.hoover.org/research/drones-reduce-us-military-capabilities; Oleksandra Molloy, Drones in Modern Warfare: Lessons Learnt from the War in Ukraine, Australian Army Occasional Paper No. 29 (Canberra: Australian Army Research Centre, 2024), 28, https://researchcentre.army.gov.au/library/occasional-papers/drones-modern-warfare. 

[7] Jeffrey L. Seavy, “The United States Must Improve Its Shipbuilding Capacity,” Proceedings 150, no. 2/1, 452 (February 2024), https://www.usni.org/magazines/proceedings/2024/february/united-states-must-improve-its-shipbuilding-capacity. 

[8] Mark W. Maier and Eberhardt Rechtin, The Art of Systems Architecting, 3rd ed. (Boca Raton, FL: CRC Press, 2009), 8.

[9] John Boardman and Brian Sauser, “Systems of Systems — the Meaning of of,” Proceedings of the 2006 IEEE/SMC Conference on Systems of Systems Engineering (Los Angeles, CA: IEEE, 2006), 118-123, https://ieeexplore.ieee.org/document/1652284. 

[10] Daniel A. DeLaurentis et al., Systems of Systems Modeling and Analysis, (Boca Raton, FL: CRC Press, 2023), 3-37.

[11] Daniel A. DeLaurentis et al., “Taxonomy to Guide Systems-of-Systems Decision-Making in Air Transportation Problems,” Journal of Aircraft 48, no. 3 (May-June 2011), 760-770.

[12] Mark W. Maier, “Architecting Principles for Systems-of-Systems,” Systems Engineering 1, no. 4 (1998), 267-284, https://incose.onlinelibrary.wiley.com/doi/abs/10.1002/(SICI)1520-6858(1998)1:4%3C267::AID-SYS3%3E3.0.CO;2-D.

[13] Jerry McGinn, Before the Balloon Goes Up: Mobilizing the Defense Industrial Base Now to Prepare for Future Conflict, Report No. 10 (Fairfax, VA: Greg and Camille Baroni Center for Government Contracting, Costello School of Business, George Mason University, 2024), 34, https://business.gmu.edu/news/2024-10/balloon-goes-mobilizing-defense-industrial-base-now-prepare-future-conflict.  

[14] Jeffrey Jeb Nadaner and Tara Murphy Dougherty, “Numbers Matter: Defense Acquisition, U.S. Production Capacity, and Deterring China,” Govini (2024), 6, https://www.govini.com/insights/numbers-matter-defense-acquisition-u-s-production-capacity-and-deterring-china. 

[15] Speer and Davidson, Inside the Third Reich, 216.

[16] U.S. Congress, Senate, Investigation of the National Defense Program: Additional Report of the Special Committee Investigating the National Defense Program, Pursuant to S. Res. 71 (77th Congress), A Resolution Authorizing and Directing an Investigation of the National Defense Program, 77th Cong., 2d sess., S. Rep. 480, Part 5, 9, https://hdl.handle.net/2027/coo.31924069038861. 

[17] U.S. Department of Defense, Office of the Under Secretary of Defense for Research and Engineering, Systems Engineering Guidebook (Washington, DC: Department of Defense, 2022), 14, https://ac.cto.mil/wp-content/uploads/2022/02/Systems-Eng-Guidebook_Feb2022-Cleared-slp.pdf.

[18] U.S. Department of Defense, Systems Engineering Guidebook, 14

[19] Maier, “Architecting Principles for Systems-of-Systems,” 267-284.

[20] U.S. Department of Defense, Systems Engineering Guidebook, 14.

[21] U.S. Department of Defense, Systems Engineering Guidebook, 14; Maier, “Architecting Principles for Systems-of-Systems,” 267-284.

[22] Klein, A Call to Arms, 373-382.

[23] Klein, A Call to Arms, 373-382.

[24] Klein, A Call to Arms, 372-387.

[25] Klein, A Call to Arms, 380-385.

[26] Klein, A Call to Arms, 380-385.

[27] Maier, “Architecting Principles for Systems-of-Systems,” 267-284.

[28] Maier, “Architecting Principles for Systems-of-Systems,” 267-284.

[29] Maier, “Architecting Principles for Systems-of-Systems,” 267-284.

[30] Anduril Industries, “Rebooting the Arsenal of Democracy: Anduril Mission Document,” Medium (June 6, 2022), 24, https://blog.anduril.com/rebooting-the-arsenal-of-democracy-anduril-mission-document-67fdbf442799. 

[31] Shyam Sankar, “The Defense Reformation,” Palantir Technologies, October 31, 2024, https://www.18theses.com/. 

[32] “Executive Order 14265 of April 9, 2025, Modernizing Defense Acquisitions and Spurring Innovation in the Defense Industrial Base,” Code of Federal Regulations, title 3 (2025): 15621–15624, https://www.govinfo.gov/content/pkg/FR-2025-04-15/pdf/2025-06461.pdf.

[33] Executive Order 14265.

[34] Michael Brown and Pavneet Singh, Why Venture Capital Is Indispensable for U.S. Industrial Strategy: Activating Investors to Realize Disruptive National Capabilities (San Francisco, CA: Institute for Security & Technology 2024), 13, https://securityandtechnology.org/wp-content/uploads/2024/10/Why-VC-is-Indispensable-for-U.S.-Industrial-Strategy.pdf. 

[35] David Novick, Melvin Anshan, and W. C. Truppner, Wartime Production Controls (New York: Columbia University Press, 1949), 105-106. 

[36] Klein, A Call to Arms, 454.

[37] “The State of Industrial DevOps 2024,” Copia Automation (2024), https://www.copia.io/resources/the-state-of-industrial-devops-2024.    

[38] Jason Wolff, “The Department of Defense’s Digital Logistics Are Under Attack,” (Washington, DC: Brookings Institution, July 2023), https://www.brookings.edu/articles/the-department-of-defenses-digital-logistics-are-under-attack/. 

[39] Shyam Sankar, “Rebooting the American Industrial Base: Software and the Future of Manufacturing,” American Affairs VIII, no. 3 (Fall 2024), 70-77, https://americanaffairsjournal.org/2024/08/rebooting-the-american-industrial-base-software-and-the-future-of-manufacturing/.

[40] Suzanne Berger and the MIT Task Force on Production and Innovation, Making in America: From Innovation to Market (Cambridge, Massachusetts: The MIT Press, 2013), 155-177.

[41] Riley Van Steenberg, “Why No One Knows the True Size of the Defense Industrial Base (UPDATED),” National Defense Magazine, December 24, 2024, https://www.nationaldefensemagazine.org/articles/2024/12/24/ndia-policy-points-why-no-one-knows-size-of-defense-industrial-base.

[42] Matt Sheehan, “Remaking ‘Made in China’: Beijing’s Industrial Internet Ambitions,” MacroPolo, February 21, 2021, https://macropolo.org/china-manufacturing-industrial-internet/?rp=e. 

[43] Sheehan, “Remaking ‘Made in China.’”

[44] Liza Tobin et al., “Beyond Fusion: Preparing for Systems Rivalry with China,” War on the Rocks, August 13, 2024, https://warontherocks.com/2024/08/beyond-fusion-preparing-for-systems-rivalry/; David Dorman, “Industrial Internet Development Boosted by Military-Civil Fusion,” Digital China Wins the Future, April 13, 2022, https://digitalchinawinsthefuture.com/2022/04/13/new-miit-issues-industrial-internet-work-plan-for-2022-military-civil-fusion-highlighted/.   

[45] Mike Bechtel, “How the US Air Force ‘STITCHES’ Together Legacy Systems,” Deloitte, November 3, 2022, https://www2.deloitte.com/us/en/insights/focus/tech-trends/2023/stitches-us-air-force-defense-interoperability-interface.html.

[46] Kristian McCann, “Why Does Manufacturing See the Most Cyberattacks?,” Cyber Magazine, August 12, 2024, https://cybermagazine.com/articles/why-does-manufacturing-see-the-most-cyber-attacks.

[47] Nadaner and Dougherty, “Numbers Matter,” 6.

[48] Jeffrey Engstrom, Systems Confrontation and System Destruction Warfare: How the Chinese People’s Liberation Army Seeks to Wage Modern Warfare, RRA1708 (Santa Monica, CA: RAND, 2018), 11, https://www.rand.org/pubs/research_reports/RR1708.html.    

[49] Patrick Tucker, “Attacks Against Defense Industrial Base Increasing, NSA Chief Warns,” Defense One, June 25, 2024, https://www.defenseone.com/technology/2024/06/attacks-against-defense-industrial-base-increasing-nsa-chief-warns/397652/.

[50] Victor Reyes et al., “Creating Pathways for Tomorrow’s Workforce Today: Beyond Reskilling in Manufacturing,” Deloitte and The National Association of Manufacturers (2021), https://www2.deloitte.com/us/en/insights/industry/manufacturing/manufacturing-industry-diversity.html. 

[51] Franklin D. Roosevelt, “Fireside Chat.,” October 12, 1942, online by Gerhard Peters and John T. Woolley, The American Presidency Project, https://www.presidency.ucsb.edu/node/209991. 

[52] Glenn Leibowitz, “Apple CEO Tim Cook: This Is the No. 1 Reason We Make iPhones in China (It’s Not What You Think),” Inc., December 21, 2017, https://www.inc.com/glenn-leibowitz/apple-ceo-tim-cook-this-is-number-1-reason-we-make-iphones-in-china-its-not-what-you-think.html.

[53] “The Lights-Sparse Versus the Lights-Out Factory,” White Paper, Siemens Digital Industries Software (2021), https://resources.sw.siemens.com/en-US/white-paper-the-lights-sparse-versus-the-lights-out-factory?bc=eyJwYWdlIjoiM3hKQ2VqeXZHRHoxSmc4b21BTVBvRSIsInNpdGUiOiJkaXN3IiwibG9jYWxlIjoiZW4tVVMifQ==. 

[54] Daniel Küpper et al., “The Factory of the Future,” Boston Consulting Group, December 6, 2016, https://www.bcg.com/publications/2016/leaning-manufacturing-operations-factory-of-future. 

[55] “Anduril Raises $1.5 Billion to Rebuild the Arsenal of Democracy,” Anduril Industries, August 7, 2024, https://www.anduril.com/article/anduril-raises-usd1-5-billion-to-rebuild-the-arsenal-of-democracy/. 

[56] Aaron Mehta, “General Atomics Signs 3D Printing Agreement with Automotive-focused Startup Divergent,” Breaking Defense, February 15, 2023, https://breakingdefense.com/2023/02/general-atomics-signs-3d-printing-agreement-with-automotive-startup-divergent/. 

[57] Chris Walti, “Demo Hour: Mytra,” Presentation at Special Competitive Studies Project AI+Robotics Summit, 2:00, October 26, 2024, https://m.youtube.com/watch?v=Kbp7R6ok7g0.

[58] Saman Farid and Audrow Nash, “Boosting Factory Utilization: The Key to Reviving US Manufacturing,” Audrow Nash Podcast, April 16, 2024, https://music.amazon.com.br/podcasts/918b9fd7-1de6-4d26-89ea-baf7e7ad7317/episodes/c3e001fe-623f-425c-91c9-bb38d38ac5c9/audrow-nash-podcast-boosting-factory-utilization-the-key-to-reviving-us-manufacturing. 

[59] Trevor Sesnick, “Starbase Tour and Interview with Elon Musk,” Everyday Astronaut, August 11, 2021, https://everydayastronaut.com/starbase-tour-and-interview-with-elon-musk/. 

[60] Dan Carney, “Tesla’s Switch to Giga Press Die Castings for Model 3 Eliminates 370 Parts,” Design News, August 13, 2021, https://www.designnews.com/automotive-engineering/tesla-s-switch-to-giga-press-die-castings-for-model-3-eliminates-370-parts.