There’s an old proverb: “For want of a nail, the shoe was lost; for want of a shoe, the horse was lost; for want of a horse, the rider was lost; and for want of the rider, the battle was lost.” It’s often quoted as a lesson in cascading failure. In modern defense and industrial systems, it represents an unfortunate reality.
In my 28-year Marine Corps career, I saw this firsthand more often than I care to remember. As a depot commander, I presided over production lines stalled for weeks or months over a single unavailable part. As a logistician in the field, I saw weapon systems sit deadlined for the same reason. Usually not complex components. Usually not overly expensive ones. Just…unavailable.
And this isn’t an isolated problem. It’s a steady drag on mission-capable rates across the entire Joint Force. In January of 2026, I was at an event where Chief of Naval Operations Admiral Daryl Caudle said publicly that if the U.S. Navy could solve parts delays and shortages needed for shipyard maintenance, the Fleet would be a third larger. Then in February, LTG Christopher Mohan, Commanding General of U.S. Army Materiel Command, publicly remarked that the Army had over 34,000 parts onbackorder, with more than 19,000 delayed more than six months. Those are staggering statistics.
That’s the uncomfortable truth: Readiness usually doesn’t break at the system level. It breaks at the part level. And the system we rely on to fix that problem isn’t built to recover fast enough.
What People Think Is Happening
When a system goes down due to a missing part, the perceived cause is usually a supply chain disruption. Maybe a supplier went out of business or discontinued a part.
Perhaps a shipment was delayed, or demand spiked unexpectedly. The response follows:
Find another vendor, buy more inventory, build buffer. All reasonable, all standard. But that’s not really where the bottleneck is.
What’s Actually Happening
In many cases, the delay is just the symptom, not the problem itself. Rather, the root cause of that delay is that viable supply no longer exists.
Parts become obsolete. Suppliers exit the market. Many components receive no bids. The Defense Logistics Agency alone releases a monthly “No Bid List” that routinely includes tens of thousands of parts. Across large organizations like the U.S. Department of War (DoW), the scale is staggering. And as such, this dynamic requires a significant ongoing process to manage “diminishing manufacturing sources and material shortages” (DMSMS). The problem is significant enough that there are entire offices, programs, and policies to govern it.
This challenge is especially acute for platforms and systems designed to last 20, 30 or even 40+ years. Consider the venerable “Humvee,” for example. First produced in 1985, these rugged vehicles are expected to last decades. During that time, the base of suppliers shrinks. Original manufacturers move on. Tooling disappears.
And critically, this isn’t about complexity. The lack of a bracket, connector, or housing can be just as disruptive as a missing electronic component. At that point, it doesn’t matter how complex the system is. It’s now constrained by a single unresolved part. No approved substitute. No clear path to produce. Everything behind it waits.
Why Can’t the System Respond Fast Enough?
By now, the reader might be tempted to accuse me of “admiring the problem,” as often happens in government and industry. But to get to the correct solution, we must properly frame the problem. And in fact, the problem is well understood. The response
just doesn’t scale.
First, the data problem. Engineering data exists, but it’s often incomplete, outdated, or unusable. Drawings live in PDFs. Technical data packages are fragmented. Information is scattered across depots, program offices, and legacy systems, or locked behind data rights issues. Even when the data exists, it’s rarely usable in a way that supports quick decisions and action.
Second, there’s no repeatable path from “need” to “solution.” Each part becomes a bespoke problem. Can it be manufactured? With what process? By whom? Under what approval pathway? Answers vary, and ambiguity slows everything down.
Third, the process depends heavily on expert judgment. Engineers must evaluate manufacturability — geometry, materials, tolerances, qualification requirements — one part at a time. That expertise is invaluable, but it doesn’t scale across portfolios of thousands of parts.
So organizations compensate. They make lifetime buys. They over-order. They build inventory buffers. These strategies tie up billions in working capital and consume precious resources that could be put to much better use. And still parts go obsolete,
inventory gets scrapped, and critical systems remain vulnerable.
The Real Constraint
The irony is that manufacturing capability exists. Suppliers exist. Technologies like additive manufacturing, or 3D printing, are more capable than ever. The real constraint is decision making. Organizations struggle to determine quickly and defensibly which parts can be produced, how to produce them, and who can do so. And they struggle to do this consistently at scale.
What Needs to Change
Fixing this starts with a better decision infrastructure, not more tools. Decisions need to be structured, repeatable, and scalable. Workflows must connect engineering, procurement, suppliers, and qualification teams. Data must become usable — not just accessible, but decision-ready. And critically, decision contexts must persist so organizations don’t have to start from scratch every time a part fails.
What’s missing is a way to make those decisions consistently, carry the context forward, and avoid solving the same problem from scratch each time. This is where better workflow, better data structure, and AI-enabled decision support can materially change the equation. Not as a silver bullet, but as a way to systematize what is currently fragmented, manual, and slow. By structuring how organizations ingest data, assess options, and route parts to qualified suppliers — whether through additive manufacturing, conventional production, or repair — viable pathways can be established for parts that would otherwise remain unresolved.
Building more inventory, or bigger iron mountains, is not the solution to readiness. Rapid, repeatable response to supplier failure is. Today, that response is too slow, too inconsistent, and too dependent on heroic individual effort. The result is
predictable: Systems sit idle, waiting on parts that should not be difficult to produce. Ultimately, readiness comes down to a simple question: How quickly can you resolve a single constraint?
Because in the end, the proverb still holds. For want of a nail — or a bracket, or a valve, or a connector — the system stops. And if you can’t fix that quickly, nothing else matters.
Kirk Spangenberg is President of Accio3D’s Aerospace & Defense Division and a retired Marine Corps Colonel with 28 years of experience in defense logistics and sustainment.
