A device can pass every design review and still be commercially fragile.
What breaks first usually isn’t the design. It’s the supply chain holding the design up. And when that happens, the delay gets labeled a “supply chain problem,” even though the root cause was an engineering decision made two or three years earlier, often for good reasons, by people who had no way of seeing the downstream cost.
We see this pattern constantly as a cable assembly manufacturer working with medical device OEMs. Some of the most expensive program delays we’ve helped untangle didn’t start on the production floor or in purchasing. They started in a requirements document.
This post walks through four places where that happens most often and ends with a short set of questions R&D teams can ask before design release, while there’s still flexibility to address it.
The Decisions That Quietly Create Supply Risk
Most single-source exposure isn’t chosen deliberately. It accumulates through choices that each look reasonable on their own:
- Specifying a component that only one supplier makes, because it was the one on the bench during feasibility
- Requiring a connector geometry that only one vendor can produce
- Defining a cable construction that sits at the edge of what most manufacturers can reliably build
- Writing verification requirements so tightly around one part that qualifying an alternate means repeating large portions of testing
- Selecting materials with thin or fragile supply chains
None of these are mistakes in the moment. The problem is that by the time a shortage, an end-of-life notice, or a price increase shows up, the design is frozen, the verification evidence is built around the original part, and engineering flexibility is gone.
Lesson 1: “Interchangeable” Components Rarely Are
On a recent program, an OEM needed an alternate source for a patient-contact electrode after the original supplier became a supply risk. On paper, the two parts looked like near drop-in equivalents. Similar construction, similar materials, similar dimensions.
In practice, switching to the alternate required a full qualification effort: a production part approval package, documentation and drawing updates, risk file review, labeling review, and verification activity before a single part could enter production builds. None of that work was wasted. It was necessary. But none of it had been planned for, because the original design never anticipated a second source.
The takeaway for R&D: equivalence on a datasheet is not equivalence in a design history file. Every component change carries documentation and verification costs, and those costs are mostly determined by how the original design and its requirements were written.
Questions worth asking during development:
- Is there a second supplier that could realistically provide this part today?
- If this component disappeared next year, what would it take to replace it?
- Are the requirements written around the function we need, or around the specific part we happened to pick?
(We covered the qualification side of this in more depth in How Medical Device OEMs Can De-Risk Supplier Changes Without Slowing Product Launches.)
Lesson 2: Design to Demonstrated Capability, Not Theoretical Capability
Cable specifications are a common place where physics and manufacturing collide.
We’ve seen programs where the requested twist rate for a cable pair (twists per inch) was higher than suppliers could consistently produce while also holding the outer diameter, shielding coverage, and electrical performance the design required. Each requirement was achievable on its own. Achieving all of them at once, lot after lot, was a different question. The result was multiple rounds of supplier quotes, raw material changes, and requirement renegotiation, all of it happening late in the program.
The lesson is simple but often skipped: build requirements around demonstrated process capability, not what’s theoretically possible on a good day. A spec that only one supplier can hit, and only at the edge of its process window, is a single-source decision even if the drawing doesn’t say so.
Before locking a cable construction, ask:
- Can more than one supplier build this, today, with their existing equipment?
- What does the actual process window look like, and where does our requirement sit in it?
- Is our tolerance stack-up realistic once raw material variation is included?
- Which requirement is driving the difficulty, and is it truly necessary at that level?
Involving your cable manufacturer at this stage costs very little. Discovering the problem during design verification costs a lot.
Lesson 3: The Connector Is Often the Highest-Risk Component in the System
On a recent high-contact-count extension cable project, the connector had to satisfy a long list of requirements at the same time:
- Electrical isolation at elevated voltage
- Shielding for signal integrity
- Compatibility with the intended sterilization method
- Mating cycle durability
- High pin count in a compact form factor
- Clean signal performance across every channel
Any one of those is manageable. Together, they narrowed the field of viable connector options to a very small number of vendors, and in some configurations to one.
This is why the connector so often becomes the critical path item in a cable-based system. It’s where electrical, mechanical, sterilization, and reliability requirements all converge, and it’s usually the component with the longest lead time and the fewest substitutes.
The practical recommendation: bring manufacturing and supply chain into connector selection earlier than feels necessary. A short conversation about lead times, alternate part availability, and assembly process implications before the connector is locked can prevent a much longer conversation after it is.
Lesson 4: Design for Qualification, Not Just Manufacturability
Most engineering teams are familiar with Design for Manufacturability. Fewer think explicitly about what could be called Design for Qualification: how hard will it be to change this design later, once it’s controlled and in production?
We recently supported an alternate configuration for an existing assembly where the change itself was modest. The downstream work was not. The production part approval package, drawing revisions, inspection updates, and supporting documentation added up to far more effort than the physical change suggested. That’s normal. It’s also predictable, which means we can design for it.
Design for Qualification asks questions like:
- How difficult will change control be for the components most likely to change?
- How difficult will supplier substitution be, given how we write our requirements and verification?
- What evidence will our quality and regulatory teams need to approve a change, and can we make that evidence easier to produce?
This doesn’t mean designing everything to be generic. Some requirements are genuinely unique and should be. It means knowing which parts of the design are locked to one source, and making that a conscious decision rather than an accident.
A Five-Question Check Before Design Release
FMEAs are good at capturing how a design can fail in use. They’re less consistent at capturing how a design can fail commercially. Before release, it’s worth adding five supply-focused questions to the review:
- Which component has only one practical supplier? Not one approved supplier. One supplier that could realistically make it.
- Which specification is hardest to manufacture? Ask your manufacturing partners. They usually know immediately.
- Which component has the longest replacement timeline? Include qualification time, not just lead time.
- What change would trigger re-verification or re-validation? Know this before you need to make one.
- What part shortage would stop production tomorrow? If there’s an answer, there should be a plan.
None of these questions take long to answer. They’re valuable because you can answer them while the design can still change.
The Bottom Line
Many supply-chain delays are really design decisions surfacing late. The teams that avoid them aren’t the ones who never specify a unique part. They’re the ones who know exactly where their single-source exposure is, made those choices on purpose, and brought their manufacturing partners in early enough to keep options open.
If your team is approaching design freeze on a cable- or interconnect-heavy system, now’s the time to ask your manufacturing partner which requirements will be hard to build, hard to source, or hard to change. It’s much easier to have that conversation now than after launch.