The Factory Floor and the Balance Sheet

What if the most important question in defense manufacturing is no longer simply how to produce more, but where to put the next dollar?

Deputy Secretary Steve Feinberg’s recent Fostering One Strong Industrial Base memorandum addresses another part of the challenge. It directs the Department to reduce government-unique accounting, audit, and compliance barriers, broaden competition, and rely more heavily on financial controls and audits companies already maintain. The memorandum also calls for negotiated margins to better reflect value delivered, risk carried, and private capital invested.

Together, these efforts address two important parts of the scale problem: creating the conditions for greater production and reducing barriers between defense and the broader commercial economy. Industry still has to convert demand and capital into output.

The answer depends on where the real constraint resides. Artificial intelligence is giving senior leaders better tools to find it.

Defense companies are hardly starting from zero.

Across aerospace and defense, manufacturers have invested heavily in digital engineering, connected supply chains, advanced production, automation, integrated planning, and increasingly artificial intelligence.

The Aerospace Industries Association and EY recently found that three-quarters of surveyed aerospace and defense organizations are implementing digital thread in some form. Yet only 14 percent reported having fully deployed it across the enterprise. Just 29 percent said their enterprise data were sufficiently standardized, governed, and accessible, while 71 percent expected predictive analytics and AI-enabled insights to deliver the greatest future value from digital thread.

Those findings point to an important transition.

The question is no longer simply whether finance and operations should be connected. Increasingly, they already are. The more important question is what the integrated enterprise should optimize for.

Most companies naturally manage different objectives through different functions.

Operations focuses on throughput. Supply chain focuses on material flow and supplier performance. Finance focuses on cash, margins, and returns on capital. Programs focus on schedule and delivery.

In defense, another requirement sits alongside all of them: resilience. That makes the optimization problem harder.

The most financially efficient supply chain may minimize inventory. A resilient defense supply chain may deliberately carry more. The lowest-cost sourcing strategy may concentrate volume with one supplier. Resilience may favor a second source. The highest near-term return on capital may argue against excess capacity. Surge requirements may justify preserving it. The leanest balance sheet is therefore not necessarily the balance sheet best suited to defense production. Nor is the factory with the highest utilization necessarily producing the most capability.

The challenge for senior management is understanding how these variables interact.

A company can improve working capital by reducing inventory and simultaneously increase production risk. It can maximize utilization of one asset while creating congestion somewhere else. It can spend millions adding capacity without increasing output because inspection, skilled labor, material availability, or a sub-tier supplier is actually limiting production.

What appears optimal inside one function can be suboptimal for the company as a whole. That is where AI becomes strategically interesting.

The next frontier in industrial AI is not simply optimizing the factory floor. It is optimizing the relationship between production and the balance sheet. And in defense, that optimization must account not only for efficiency, but also for resilience.

Consider a manufacturer preparing for a major production ramp.

Management has several ways to deploy incremental resources. It can buy equipment, increase strategic inventory, automate inspection, qualify another supplier, add personnel, or help alleviate a constraint deeper in the supply chain.

Traditional capital planning can evaluate the expected return from each investment. But a production system is more complicated than a collection of standalone projects.

Another machining center creates little additional output if inspection is already the bottleneck. More inventory may be valuable if a long-lead component is limiting production, but unnecessary if skilled labor is the constraint. Capital deployed to a critical supplier may generate more incremental output than another investment inside the company’s own facility.

The better question is therefore not simply which investment offers the highest standalone return. It is: Where does the next dollar create the greatest increase in resilient capability?

Answering that requires seeing the business as a system.

Production schedules, equipment utilization, workforce availability, supplier performance, inventory, backlog, purchasing commitments, cash requirements, and capital plans all contain pieces of the answer.

AI creates the possibility of analyzing those variables together, modeling their interactions, and testing alternatives before scarce resources are committed.

A senior management team considering a production increase could ask not only whether sufficient capacity exists, but what happens across the business if output rises.

Which resource becomes constrained? Where does inventory accumulate? When does cash usage peak? Does another investment actually relieve the bottleneck? Would the same dollar create more output somewhere else?

The point is not to replace executive judgment with an algorithm. It is to give leaders a more complete operating picture against which to exercise that judgment.

Defense manufacturing makes this challenge especially clear because resilience sometimes requires accepting costs that a purely efficiency-driven model would eliminate.

The goal is therefore not the leanest possible enterprise. It is to deploy capital in a way that produces the required output with the right level of resilience.

AI increasingly allows senior leaders to evaluate those tradeoffs across functions rather than optimizing each function independently.

Lower inventory releases cash. What does it do to delivery risk? New capacity sounds attractive. Does it address the limiting factor?

Senior executives do not need another collection of KPIs. They need a clearer understanding of the relationships among the ones they already use.

The industry’s investment in digital thread provides much of the foundation.

AIA and EY define digital thread as the seamless flow of trusted information across design, production, and operations. Their research also suggests that enterprise value depends increasingly on governance, standardized data, performance measurement, and execution rather than technology alone.

Connecting information, however, is not the same as deciding what to optimize.

For defense manufacturers, one useful management lens is capital-to-capability conversion: understanding where incremental resources actually produce additional reliable output. This is not another accounting ratio. It is a way of thinking about the enterprise.

A CEO, COO, CFO, program leader, and operations executive should increasingly be able to examine the same proposed investment and understand its consequences for production, cash, schedule, suppliers, and resilience.

That creates a more powerful management system than optimizing those variables independently. The factory floor and the balance sheet stop being separate conversations. They become different views of the same operating system.

Government still plays an important role.

Credible demand signals influence investment. Multiyear procurement can change the economics of capacity expansion. Contracting speed and payment structures affect the amount of capital industry must carry.

Duffey’s focus on production scale and Feinberg’s effort to reduce barriers to commercial participation address parts of that environment. But Washington cannot determine whether a manufacturer should buy another machine, carry additional inventory, automate a process, or deploy resources somewhere else inside the business. Those are management decisions.

Industry ultimately decides how people, equipment, suppliers, data, and capital are combined to create output. AI offers leaders a new ability to make those decisions with a more complete understanding of the system.

For much of the past generation, industrial advantage was built by relentlessly improving efficiency. The next era will require something more complicated.

Defense companies will have to increase output while strengthening resilience, integrating new technologies, managing constrained workforces, expanding supplier networks, and preserving financial performance.

Those objectives will not always point in the same direction.

The next industrial advantage may therefore belong not simply to the companies with the most capacity, the largest backlogs, or the greatest access to capital. It may belong to those that understand their operating systems well enough to know what is actually limiting output and where incremental resources will create the greatest effect.

The next industrial advantage will come from understanding the enterprise well enough to answer a deceptively simple question: Where does the next dollar create the most capability?

Ryan P. McDermott is an Iraq War veteran, former Pentagon official, and author of the book, Downriver: Memoir of a Warrior Poet. His views do not reflect those of his employer or any affiliated organization.