In-House or Outsourced Scanning? A Decision Framework for Aerospace Suppliers
The short answer
Most aerospace suppliers should outsource DR/CT inspection until they can keep a system busy roughly 50% of available production hours on a stable, repeating part mix — and until they have, or can hire, a NAS 410 Level III to own the process. Below that threshold, outsourcing is almost always cheaper and lower-risk. Above it, the economics flip quickly, and the real constraints become staffing, facility, and accreditation rather than the purchase price.
The complication is that most suppliers make this decision on capital cost alone, which is the least reliable input. Here is a framework that weighs the factors that actually determine the outcome.
The 4 questions that decide it
Before any cost modeling, answer these.
If three of the four point the same direction, you have your answer and the spreadsheet is a formality.
- What is your annual scan volume, and is it stable? A supplier running 200 scans a year on twelve different part numbers is in a different business than one running 6,000 scans on three. Volume drives amortization; stability determines whether you can build repeatable, validated techniques instead of re-engineering every setup.
- How diverse is your part mix? A narrow mix of similar geometries, materials, and thicknesses is the ideal in-house case — you can specify one system precisely and automate the routine. A wide mix spanning thin composite skins and thick titanium forgings may need two very different energy ranges, and no single system will serve both well.
- What turnaround do you actually need? This may be the most underweighted variable. If CT results gate a production decision, every day of shipping and queue time is work-in-process sitting idle — or worse, downstream operations performed on parts that later scrap. If CT is a periodic audit or a first-article activity, turnaround barely matters.
- What does your customer require? Some primes mandate a Nadcap-accredited process for radiographic inspection of flight hardware. That requirement doesn’t disappear when you bring the work in-house — it moves onto your shoulders. Read the flowdown before you model anything.
The total cost picture, beyond the system price
The purchase price of an industrial CT system is the number everyone quotes and the smallest share of a five-year cost of ownership. Build your model with all of these:
| Category | What it includes | Frequency |
|---|---|---|
| Capital | System, detector, tube configuration, manipulator, automation | One-time, amortized 7–10 yrs |
| Facility | Shielding or cabinet siting, floor loading, three-phase power, chilled water or HVAC, vibration isolation | One-time |
| Software | Acquisition, reconstruction, analysis, metrology and ADR modules; annual licensing | One-time + recurring |
| Consumables & wear | X-ray tube replacement or refurbishment, filaments, detector calibration | Recurring, plan for it |
| Service | Preventive maintenance and support contracts, spares | Annual |
| People | Operators, a Level III to own the process, backfill for absence and turnover | Annual, and the largest line |
| Qualification | Written procedures, technique development, system performance verification, Nadcap audit preparation and cycle | Upfront + ongoing |
| Data | Reconstructed CT volumes are large. Archival, backup, and retention infrastructure for the required record period | Annual, grows |
Two categories consistently get underestimated. The first is data: a single high-resolution CT volume can run to several gigabytes, and aerospace record retention is measured in years or decades. The second is people, which is significant enough to deserve its own section.
The hidden constraint: qualified operators
You can order a system and take delivery in a matter of weeks — NSI now offers select configurations in as little as 60 days. You may not be able to hire a certified RT professional on the same schedule.
Flight-hardware radiography requires personnel qualified under NAS 410, with a Level III responsible for procedures, technique approval, and interpretation oversight. Depending on your starting point, building that capability internally means:
- Accumulating documented experience hours toward certification
- Formal classroom and practical training, plus examinations
- Access to a Level III — employed, contracted, or shared — who can sign your procedures
- A written practice, and evidence you follow it
- Continuity planning, because a single certified operator is a single point of failure
A capable system with no one certified to interpret its output is an expensive paperweight. This is the most common reason in-house programs stall in year one. If you are heading down this path, structured operator training should be scheduled alongside the equipment order, not after installation — a point covered in more depth in why advanced X-ray training is no longer optional.
A simplified break-even worksheet
Rather than quoting prices that vary enormously by configuration, fill in your own numbers below. Ask your prospective supplier for a quote and your prospective service provider for a rate card — the totals calculate automatically as you type.
Outsourced annual cost
| Line item | Your number | |
|---|---|---|
| A | Scans per year | |
| B | Average cost per scan ($) | |
| C | Shipping, packaging, expedite fees ($/yr) | |
| D | Internal labor to prepare and ship ($/yr) | |
| E | Cost of turnaround delay — WIP carrying cost + scrap discovered late ($/yr) | |
| Outsourced total = (A × B) + C + D + E | $0 | |
In-house annual cost
| Line item | Your number | |
|---|---|---|
| F | System Cost ÷ amortization years ($/yr) | |
| G | Facility amortized ($/yr) | |
| H | Annual service + software licensing ($/yr) | |
| I | Fully loaded operator salaries ($/yr) | |
| J | Level III coverage — employed or contracted ($/yr) | |
| K | Training and recertification ($/yr) | |
| L | Qualification and accreditation upkeep ($/yr) | |
| M | Data storage and archival ($/yr) | |
| In-house total = F + G + H + I + J + K + L + M | $0 | |
| Effective cost per scan = In-house total ÷ A | $0 | |
| At today's numbers: enter your figures above | ||
Then run the sensitivity that matters: recalculate at 60% of your projected volume. In-house economics are dominated by fixed cost, so a volume shortfall hurts far more than it would with outsourcing. If the decision only works at your optimistic forecast, it is not yet a decision — it is a bet.
Also sanity-check capacity, not just cost. Multiply your realistic cycle time — including load, technique setup, acquisition, reconstruction, and analysis — by annual scan volume, and compare it against available machine hours. If the answer is under about 40–50% utilization, you are buying idle capacity. If it is over 85%, you have no room for growth, maintenance windows, or requalification work, and you should be modeling two systems or a hybrid approach.
ITAR, controlled data, and defense programs
For ITAR- or CUI-controlled hardware, the analysis changes shape. Outsourcing is entirely viable, but the provider must be ITAR registered, with controlled facility access, US Person staffing where required, and a data handling regime that satisfies your program’s requirements. NSI is ITAR registered and operates accredited domestic facilities for exactly this reason.
Where in-house tends to win outright is classified or highly restricted work where transferring hardware off-site is impractical regardless of cost, and where sustained defense program volume justifies a dedicated asset — the case behind purpose-built configurations like the X7000LT for defense programs.
The hybrid path
The framing as a binary choice is the flaw in most versions of this analysis. In practice the strongest programs run both, and the sequence looks like this:
- Outsource during development. Use inspection services to characterize your parts, discover what resolution and technique your acceptance criteria genuinely require, and build a defensible technical case — before you write a capital request based on assumptions.
- Specify the system from real data. You will know your required voxel size, energy range, part envelope, and cycle time from actual scans of actual hardware, not brochure figures. This is the single biggest driver of buying the right system.
- Bring the repeating work in-house. Move high-volume, stable production inspection onto your own system, where fast feedback and cost per scan both favor internal capability.
- Keep a provider for the exceptions. Oversized parts, high-energy work beyond your system’s range, overflow capacity during rate increases, failure analysis, litigation support, and coverage during your own maintenance downtime.
Very few aerospace suppliers reach a steady state of zero outsourcing. Planning for a permanent hybrid is realistic, not a failure of the business case.
|
✓ Signals you are ready to buy
|
✓ Signals you should keep outsourcing
|
Frequently asked questions
How much does an industrial CT system cost?
Configuration drives price far more than brand. Energy range, focal spot, detector, part envelope, automation, and software modules can move the figure by several multiples. Rather than budgeting from a general range, request a quote against a defined part envelope and required voxel size — and model total cost of ownership over 7–10 years, where facility, staffing, service, and data costs typically exceed the purchase price.
When does it make financial sense to bring CT inspection in-house?
Broadly, when projected utilization exceeds roughly half of available machine hours on a stable part mix, and when turnaround time carries real production cost. Below that, fixed costs dominate and outsourcing usually wins. Always re-run the model at 60% of forecast volume before committing.
Can I outsource CT inspection for ITAR-controlled parts?
Yes, provided the service provider is ITAR registered with appropriate facility access controls, personnel requirements, and controlled data handling. Confirm registration status and program-specific requirements before transferring any hardware or data.
Do I need Nadcap accreditation to inspect flight hardware in-house?
It depends on your customer’s flowdown, but many primes require a Nadcap-accredited nondestructive testing process for flight-critical parts. Accreditation is achievable but represents a sustained commitment to written procedures, qualified personnel, and audit cycles. Using an already-accredited provider transfers that burden — NSI holds Nadcap NDT accreditation at multiple facilities.
What is the biggest mistake suppliers make with this decision?
Buying a system specified from assumptions rather than data. Scanning representative parts at a service provider first tells you the voxel size, energy, and cycle time you genuinely need — and frequently reveals that the system you were about to buy was either over-specified or unable to meet the requirement at all.
Ultimately, this is not solely a capital equipment decision. It is a staffing, throughput, and risk decision that happens to involve capital equipment. Suppliers who get it right almost always follow the same order: prove the requirement with outsourced scans, specify the system from that evidence, hire and certify ahead of installation, and keep a provider relationship for the work that will always fall outside their envelope.
North Star Imaging is one of the few organizations positioned on both sides of this decision — we build industrial CT and digital radiography systems and operate Nadcap-accredited, ITAR-registered inspection service labs. That means we can help you scan first and decide second. Explore our aerospace applications, or talk to our experts about modeling your specific case.
Talk to Our Experts