Navigational, Measuring, Electromedical, and Control Instruments Manufacturing (NAICS 3345)

A short Histometrics rollup primer for public-market and private investors. This industry group has exactly one child industry, so this level and that child are the same set of companies. We give this level's own ground-truth federal statistics and then point you to the full child primer for detail. Figures are reported facts with citations; statements about the future are labeled outlook or judgment.

1. Overview

NAICS 3345 is the U.S. instruments industry group — the plants that build the machines that sense, measure, test, image, guide, and control the physical and biological world. NAICS is the North American Industry Classification System, the federal scheme for grouping businesses; this is its 4-digit "industry group" level. It spans pacemakers and MRI (magnetic resonance imaging) scanners, fighter-jet radar and missile guidance, the thermostat on your wall, the flow meter in a refinery, the water meter on your house, the oscilloscope on an engineer's bench, the mass spectrometer in a drug lab, the CT (computed tomography) scanner and cancer-treatment accelerator in a hospital, and the radiation detector in a nuclear plant.[1][2]

Why it matters to an investor: taken together these are the "picks and shovels" of the entire economy and the health system — high-value, engineering- and patent-intensive products, protected by hard-won certifications, that seed decade-long tails of consumables, service, calibration, spares, and software. Because the products serve defense, health care, industrial capital spending, utilities, semiconductors, and construction, the aggregate is steadier than any single piece.[2]

2. What's inside — and why this level equals its one child

An "industry group" (4-digit) normally rolls up several "industries" (5-digit). This one is unusual: NAICS 3345 has a single child, NAICS 33451, and nothing else. So the two codes describe the identical set of companies, the same shipments, and the same workforce — the numbers below are simultaneously 3345's and 33451's.[1][2]

The real internal structure lives one level further down. NAICS 33451 splits into nine 6-digit sub-industries that behave very differently as businesses. By 2022 shipments they run: search, detection, navigation and guidance (334511) at $53.5 billion, 31.7% of the level; electromedical and electrotherapeutic apparatus (334510) at $39.6 billion, 23.5%; analytical laboratory instruments (334516) at $19.35 billion, 11.5%; industrial process control (334513) at $14.2 billion, 8.4%; the "other measuring" catch-all (334519) at $12.87 billion, 7.6%; electrical test and measurement (334515) at $11.66 billion, 6.9%; irradiation apparatus (334517) at $9.18 billion, 5.4%; totalizing fluid meters (334514) at $5.29 billion, 3.1%; and automatic environmental controls (334512) at roughly $3.0 billion, 1.8%.[2]

They differ on nearly every axis an investor cares about. Value and headcount sit in different places — the largest sub-industry has only 404 firms while the fifth-largest has 789, so firm count does not track size.[2] Pay and productivity vary roughly two-fold: average pay runs from ~$82,000 in metering to ~$112,000–121,000 in defense electronics, analytical labs, electromedical, and irradiation, and output per worker from ~$443,000 in metering to ~$730,000 in irradiation.[2] Gross margins span roughly 32% to 69%, ordered by how much of the value is not the box (Section 5).[2] And the direction of travel is genuinely divergent — defense electronics in an up-cycle, the health sub-industries steady, analytical labs muted but recovering, process control cyclical, environmental controls mid-transition, and electrical test split against itself.[2] That contrast among the nine is the whole story of this level — and it is told in full in the 33451 child primer, which this page does not duplicate. For any depth on the sub-industries, read 33451.

3. How big it is — the rollup

Our ground-truth federal statistics for NAICS 3345 (identical to 33451 because it is the only child):

The revised child primer confirms these totals bottom-up: on a like-for-like 2022 Census and 2023 CBP basis the nine sub-industries' employment sums exactly to 414,694, their establishments exactly to 5,050, and their receipts to ~$168.7 billion — the nine are the complete set. Firm counts are the one line that does not add (~4,581 across the nine against 4,427 here), because a company active in two sub-industries is counted once at this level but appears in each of them.[2] (One wrinkle: the electromedical primer now leads with an AIES 2023 employment figure of 94,349 rather than the CBP figure of 94,361 that makes the sum close. The two come from different Census programs; the reconciliation above holds only on the CBP basis.[2])

The economics are the signature of a capital- and intellectual-property-intensive sector: ~$407,000 of output and ~$106,000 of pay per worker, well above the manufacturing average — value comes from engineering and certification, not headcount.[1]

The undercount caveat — read this before quoting $168.6 billion. This is what U.S. establishments physically ship, not the size of the business the household-name companies actually run, and the revised child makes the gap concrete on four fronts:

1. The giants are booked under other codes, and single segments now rival whole sub-industries. A company is classified by its primary product, so GE HealthCare's global Imaging segment alone (~$9.2 billion in 2025) is about the size of the entire irradiation sub-industry's $9.18 billion of U.S. shipments; Honeywell's Building Automation segment booked $7.367 billion against environmental controls' ~$3.0 billion; Thermo Fisher's Analytical Instruments segment was $7.55 billion against the whole analytical sub-industry's $19.35 billion; and Keysight books $5.4 billion worldwide against electrical test's $11.66 billion. Medtronic, Abbott, and Boston Scientific report most revenue under other medical codes, and radar and guidance dollars are largely counted inside aircraft (336411) and guided-missile/space (336414) manufacturing.[2]
2. Trade cuts both ways — the old blanket "imports aren't in it" line was only half right. Controls, meters, imaging systems, and test gear are heavily import-supplied, so U.S. shipments understate U.S. consumption there. But electromedical runs the other way: U.S. receipts of $45.5 billion (AIES 2023) exceed an estimated U.S. domestic market of about $26.3 billion in 2024, because America is a net exporter of high-end implants and devices — even though the broader medical-instruments trade line (HS 9018) showed $35.8 billion of exports against $41.3 billion of imports in 2024. Treat this level as a mix of net-export and net-import pieces, not one import-swamped whole.[2]
3. Global sales dwarf U.S. output, including at firms with no U.S. listing — privately held Endress+Hauser passed €4 billion of 2025 sales, privately held Rohde & Schwarz booked €3.16 billion in fiscal 2024/25, and Elekta's fiscal 2025/26 net sales were SEK 16.7 billion (~$1.7 billion). None of that is U.S. shipments.[2]
4. The recurring streams sit outside manufacturing. Consumables, service, calibration, and software subscriptions are frequently classified as services: GE HealthCare split 2025 into $13.7 billion of product and $7.0 billion of service, and ended the year with $10.7 billion of service remaining performance obligations against $5.0 billion of product obligations; Keysight's services reached 24% of revenue and Agilent's 29%.[2]

A vintage note in the same direction: where newer Annual Integrated Economic Survey (AIES) 2023 figures exist they run well above the 2022 Census — electromedical at $45.5 billion versus $39.6 billion, environmental controls at $4.196 billion versus ~$3.0 billion. The 2022 vintage is used throughout because it is the only one available for all nine sub-industries and for this level; read the newer numbers as evidence the baseline is conservative, not as replacements.[2]

This is a code-structure undercount, not a hidden small-proprietor one. The sector is corporate and incorporated, and even "small" means something unusual here: SBA size standards for the nine sub-industries run from 600 to 1,350 employees,[3] and the FDA's own analysis of the irradiation sub-industry found 98 of its 113 employer firms (87%) qualified as small under a 1,200-employee cutoff.[2] The private/family long tail is real but concentrated in the catch-all (334519), environmental controls (334512), and process instruments (334513), and it does not distort the totals much.[2]

4. The investable universe — where value concentrates

There is no single "instruments" stock and no ETF (exchange-traded fund) for NAICS 3345; access is child-by-child, and value concentrates in three places (all detailed in the 33451 primer):

• Defense electronics and health care hold the biggest, cleanest public franchises — reachable through aerospace-and-defense primes and funds, and medical-device, imaging, and radiation-therapy large-caps.
• The industrial-measurement middle is dominated by diversified serial acquirers where instruments are one segment, with a widening set of near-pure-plays for focused exposure — in test, lab tools, meters, radiation detection, surveying, and controls.
• Several category leaders you cannot buy publicly — the best process, test, metering, lab, and detection franchises are private or foreign-listed.

One update from the revised child: the shallow middle is getting less shallow. The same conglomerates that spent decades rolling these niches up are now splitting them back out, and each break-up manufactures a new pure instrument company (Section 8) — with the caveat that a newly focused pure-play concentrates the downside as well as the upside.[2] Tickers, yields, and multiples belong to Section 10 and the child primer; at this level the point is that public access is deepest in defense and health care and shallowest, though improving, in the industrial-measurement middle.

5. How the money works

The nine sub-industries share one economic engine, so the level's model is easy to state: an initial device sale seeds a multi-year, higher-margin annuity of consumables, spares, calibration, service, and software (a "razor-and-blade" pattern — recurring revenue is roughly 83% of sales at Thermo Fisher and 81% at Danaher, and GE HealthCare's service backlog is more than double its product backlog).[2] Once an instrument is "specified" into a plant, hospital, ship, or lab and validated for a regulated workflow, switching costs and installed-base lock-in protect pricing — visibly, in the 6.7% rise in the producer price index for process-control instruments between April 2025 and April 2026.[2]

Three refinements the revised child adds. First, the gross-margin ladder is the thesis in one line: margins run from roughly 32% to 69% across the level, with implantables and monitoring at the top, high-end test and precision metrology next, lab instruments behind them, then detection and metering, capital imaging around 40%, and OEM radiation hardware at the bottom — an ordering that tracks how much of the value is not the box. Itron shows the same ladder inside one company: 31.2% gross margin on devices, 39.1% on networked solutions, 39.6% on outcomes analytics.[2] Second, R&D intensity is wider than previously stated here — roughly 6–19% of sales, not 8–18% — with medtech lower than assumed (GE HealthCare ~6%, Medtronic ~8.5%) and test higher (Keysight at 19%).[2] Third, operating margins spread wider than gross margins, from the mid-20s to low-30s at the best segments down to a cluster near 12%, with some participants losing money outright.[2] For the capital-goods sub-industries, book-to-bill (new orders versus shipments) above 1.0 signals expanding backlog — but backlog is not revenue, since customers can often cancel or reschedule without penalty.[2] What varies across the nine is the customer and the cycle, which is exactly why the aggregate is diversified.

6. What drives demand

Because the level is a portfolio of nine niches, demand is a portfolio of separate drivers rather than one cycle. Defense budgets and the threat environment anchor the largest piece — about $900.6 billion for national defense in FY2026, of which roughly $295 billion is procurement plus RDT&E, the buckets that buy sensors and guidance.[2] Demographics, chronic disease, and cancer care drive the health sub-industries: 61.2 million Americans aged 65 and over in 2024 (18.0% of the population) and an estimated 2.1 million new U.S. cancer cases in 2026.[2] Industrial and energy capital spending moves process control and part of test, with process industries accounting for 55–60% of instrumentation demand.[2] Semiconductors and AI are the loudest current driver — semiconductor test revenue at the category leader rose 18.8% in 2025 — and data centers have become a demand line of their own, contributing just under half of one controls specialist's 2025 sales growth.[2] Utility budgets and infrastructure funding underpin metering, where North American water AMI (advanced metering infrastructure) penetration was still only about one-third at end-2022 against an EPA-assessed $625 billion of 20-year drinking-water need.[2] Pharma and research funding drive analytical labs (NIH's FY2025 appropriation was $48.5 billion).[2] Two newer drivers the parent previously lacked: the nuclear revival tied to AI data-center power, with technology companies committing over $10 billion to nuclear partnerships, and the regulatory shift away from ethylene oxide sterilization toward electron-beam and X-ray, which should roughly double the sterilization-services market that buys those machines by 2030.[2] Cutting across all of them, steady replacement and calibration cycles run largely independent of the business cycle, and technology jumps pull replacement demand forward.[2]

7. Regulation

Across the level, regulation is more often a moat and demand creator than a constraint. Measurement traces through an unbroken calibration chain to national standards at NIST (the National Institute of Standards and Technology), and accredited calibration is itself a recurring-revenue business; FDA (Food and Drug Administration) device pathways and CMS (Centers for Medicare & Medicaid Services) reimbursement gate the health sub-industries; export controls — ITAR (International Traffic in Arms Regulations) and EAR (Export Administration Regulations), plus NRC licensing for nuclear-related equipment and advanced-chip controls for the most sophisticated testers — bind defense, high-end test, and nuclear instruments; and functional-safety, environmental, and metering-accuracy standards force customers to buy more instruments.[2]

Three things the revised child sharpens. Cybersecurity is now a regulated requirement, not merely a risk, in four of the nine sub-industries: FDA section 524B for "cyber devices" in force since March 29, 2023; DFARS 252.204-7012 and the CMMC framework flowed down to defense subcontractors; IEC 62443 pull in process control; and mounting scrutiny of water-metering networks.[2] The FDA's Quality Management System Regulation (QMSR) took effect February 2, 2026, incorporating ISO 13485:2016 and opening management-review, quality-audit, and supplier-audit records to inspection.[2] And fiscal policy now cuts both ways with dated cliffs — the 25C residential efficiency credit expired for equipment placed in service after December 31, 2025, the 179D commercial deduction terminates for construction beginning after June 30, 2026, and the Build America, Buy America waiver that keeps federally funded AMI water-meter projects buildable runs only to December 19, 2027.[2] The net effect is still uniform: certification raises barriers to entry, protecting incumbents, while creating approval, recall, and compliance-cost risk for any single product line.

8. Consolidation

Three patterns run across the whole level, and the third is new.

First, the aggregate looks unconcentrated but every piece of it is more concentrated than the whole. The level's HHI of 163.2 is lower than any individual sub-industry's, its CR4 of 19.0% is lower than the lowest sub-industry's (23.5%), and every sub-industry for which Census publishes a top-50 share sits above this level's 58.9% — up to 96.2% in defense electronics and 93.2% in metering.[1][2] That is arithmetic, not competition: the code pools nine industries that do not compete with one another. In the real markets, pacemakers, hearing aids (six firms hold roughly 80%), high-end imaging (the top five vendors hold ~96% of the U.S. market, one at ~40%), semiconductor test (an ~80% two-firm duopoly), mass spectrometry, and surveying (effectively three names) are each controlled by two or three players.[2] Concentration is also a moving target rather than a fixed fact: the smallest sub-industry, environmental controls, now carries the third-highest CR4 in the level at 40.8%, roughly double its 24.8% in the 2002 Census.[2]

Second, the same diversified serial acquirers straddle multiple sub-industries — Emerson, Honeywell, AMETEK, Fortive, Roper, Danaher, Thermo Fisher, GE HealthCare, Siemens Healthineers, Xylem, ITW — fed by a pipeline of venture- and private-equity-backed innovators that mostly exit by acquisition rather than IPO (initial public offering). The scale is large: the Waters–BD Biosciences & Diagnostic Solutions combination (~$17.5 billion) closed February 9, 2026, alongside Siemens–Varian (~$16 billion), Blackstone–Copeland (~$14 billion), J&J–Shockwave (~$13.1 billion), and Emerson–National Instruments (~$7.8 billion).[2]

Third — and this is the update — the level is now splitting apart about as fast as it consolidates. Honeywell is separating into three companies, Fortive spun out Ralliant in June 2025, Danaher spun Veralto in 2023, GE spun GE HealthCare in 2023, Johnson Controls sold residential and light-commercial HVAC to Bosch for ~$8.1 billion, and Xylem sold its non-North-American metering arm.[2] Private equity has become a first-class owner class alongside the strategics.[2] For investors that creates three repeatable plays rather than two: own the acquirers, own or back the targets, and buy the spin-offs when a conglomerate finally surfaces a pure instrument business — remembering that Ralliant took a $1.44 billion goodwill impairment in its first full year.[2]

9. Risks

The main risks carry over from the child, several now with hard numbers attached. Cyclicality of customer capital spending is real and recent — one major general-purpose test business fell 14.5% in 2025 and another's revenue dropped 9% in 2024 with operating margin falling from 24.8% to 16.7%.[2] Budget, reimbursement, and policy risk spans distinct public payers (the Pentagon, CMS and insurers, utility rate cases, research funding) and now includes three dated cliffs in 2026 and 2027.[2] Tariffs are a quantified, level-wide margin headwind rather than a hypothetical: GE HealthCare estimated tariffs cut its 2025 operating income by approximately $245 million and cash flow by approximately $285 million, and Siemens Healthineers and Philips each projected impacts in the hundreds of millions of euros.[2] Supply-chain pinch points are specific and hard to substitute — semiconductors, rare earths, gallium and germanium, helium and iodine, cobalt-60 on reactor harvest schedules, sole-source detector materials, long-life lithium batteries — and the binding problem is qualification, since swapping a component can require redesign, revalidation, and sometimes a new regulatory submission.[2] Add customer concentration (one OEM supplier's top five customers were about 40% of fiscal 2025 revenue), fixed-price contract execution (one prime booked 73% of fiscal 2024 revenue on fixed-price work; another took a $600 million charge on a terminated development contract), technology disruption and structural decline — environmental-controls employment fell 63.5% between 2000 and 2024, from 35,549 to 12,965 — foreign and low-cost competition, skilled-labor scarcity in cleared RF engineering, field service, and calibration, and concentration and currency (FX) exposure from global sales.[1][2]

10. How to invest and the outlook

How to invest. Because there is no single instruments ETF, assemble the exposure child-by-child from the 33451 primer: aerospace-and-defense funds and primes for defense electronics; medical-device and imaging large-caps for health care; diversified industrial/automation and life-science-tools funds — or the growing set of pure-plays — for the industrial-measurement middle. Every available fund carries meaningful non-33451 content, so none is a tracker.[2] On the private side, the opportunity is consistent: venture capital into next-generation-instrument startups (exit mostly by strategic acquisition, so timing tracks the majors' M&A appetite), private-equity roll-ups of contract manufacturers, calibration and test-lab services, and niche makers prized for sticky aftermarket revenue, and direct ownership of the many private or foreign specialists. The child primers converge on one diligence checklist: separate genuinely recurring revenue from deployment-period maintenance, test whether backlog is cancelable, verify that certifications and export classifications transfer on a change of control, and map sole-source component exposure.[2]

Outlook (forward-looking judgment). The appeal of NAICS 3345 is precisely that it is a diversified basket of durable, certification-moated, recurring-revenue businesses riding different cycles. Defense and sensor demand and AI-driven semiconductor and memory test are up now; the health-care sub-industries and utility metering are steady and anchored to non-discretionary budgets; analytical labs are muted but turned back up in 2025, with process control cyclical yet structurally intact and demonstrating pricing power; electrical test is genuinely bifurcated, with AI silicon pulling one half up while general-purpose and EV-exposed lines fall; and building controls is mid-transition from hardware to subscription software while absorbing a policy air pocket.[2] The reliable base case is continued portfolio churn — cash-rich acquirers buying innovation while conglomerates split themselves into purer instrument companies — and a steady migration of value from the box to the software and services around it, now measurable in the roughly eight-point gross-margin gap between selling a meter and selling the analytics that run on it.[2] The offsets are inherent cyclicality, a patchwork of budget and reimbursement dependencies with several dated cliffs, and tariff and supply-chain exposure that is no longer speculative. For the full sub-industry-by-sub-industry treatment, see the NAICS 33451 primer. These are informed judgments, not guarantees.

---

Sources

1. Histometrics ingested federal statistics for NAICS 3345 — U.S. Census Bureau, 2022 Economic Census (receipts $168.6B; 4,427 firms; CR4 19.0%, CR8 27.9%, CR20 43.4%, CR50 58.9%; HHI 163.2) and County Business Patterns 2023 (5,050 establishments; 414,694 employees; $44.0B annual payroll). Because NAICS 33451 is the only child of NAICS 3345, these figures are identical at both levels. https://www.census.gov/programs-surveys/economic-census.html; https://www.census.gov/programs-surveys/cbp.html
2. Histometrics child primer — NAICS 33451, Navigational, Measuring, Electromedical, and Control Instruments Manufacturing (the sole child industry; full treatment of the nine 6-digit sub-industries — 334510 electromedical $39.6B, 334511 search/detection/navigation/guidance $53.5B, 334512 environmental controls ~$3.0B, 334513 industrial process control $14.2B, 334514 fluid meters $5.29B, 334515 electrical test $11.66B, 334516 analytical lab $19.35B, 334517 irradiation $9.18B, 334519 other measuring $12.87B — with per-child concentration, pay and productivity, gross and operating margins, R&D intensity, the AIES 2023 vintage comparison, the two-way trade picture, segment-versus-code scale examples, the QMSR and §524B/CMMC/IEC 62443 regulatory frontiers, the acquire-and-split consolidation record, quantified tariff and supply-chain risk, and the investable universe). Draws on Census 2022 / CBP 2023 / AIES 2023, BLS, SEC filings, FDA, CMS, NRC, NNSA, NIH, EPA, DOE, GAO, and industry-research sources.
3. U.S. Small Business Administration, "Table of Small Business Size Standards," NAICS 33451 child codes (600–1,350 employees), 2023. https://www.sba.gov/document/support-table-size-standards