EXPOSED
The modal worker in this catch-all category builds and instruments test setups, runs bench and field tests, calibrates equipment, collects data, and writes it up for the engineers who sign off. The hands-on half — wiring a test rig, mounting sensors, troubleshooting why a prototype fails on the floor — is genuinely hard to automate; the documentation, data reduction, spec-checking, and report-writing half is already being absorbed by AI and automated test software. No license stands between this role and automation, and the engineer, not the technician, owns the consequential call.
Most of this decline happened after 2021 — it is not the pandemic dip.
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing.
BLS projection, 2024–2034
+1.5% 67,300 → 68,300 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +1.5% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~5,700 openings a year on average, including replacing people who leave.
TesterScannerAir TesterGun TesterTechnicianAir AnalystJob CaptainMetal TesterSheet TesterSteel TesterFiber SplicerForm DesignerSteel AnalystDrawing TracerLaser OperatorSpectrographerSpectroscopistStress AnalystTensile TesterX Ray OperatorDrawing CheckerHardness TesterMarine SurveyorSand Technician
The BLS uses Engineering Technologists and Technicians, Except Drafters, All Other for work that doesn't fit any named occupation, so it covers roles that have little in common with each other. Two consequences worth knowing before you read anything below:
Holding it up: embodiment . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier An 11 reflects the split shift: LabVIEW/TestStand sequences, automated DAQ, and LLM-drafted test reports have already eaten the data reduction, spec-compliance checking, and write-up portions of the day, while fabricating a fixture, routing thermocouples on a hot prototype, and diagnosing intermittent connector faults still need a person on the bench — which is why it sits at the top of the mixed band rather than in the resistant one.
Hands-on in uncontrolled environments At 13 the work happens where the hardware is — climbing over a test stand, pulling cable in an unfinished plant, taking field readings on equipment you did not install and cannot fully control — but a real fraction of hours is still spent at a desk reducing data and updating drawings, which keeps this off the 17-plus tier occupied by field service and rigging trades.
No licence, no signature requirement A 3 is the honest floor: nothing in this role requires a PE or state licence, employer-side credentials like ASNT NDT Level II or an ETA cert are asked for by some shops and waived by others, and the seal on the test report belongs to the supervising engineer — the technician's initials on a data sheet carry no statutory weight.
Meaningful discretion An 8 covers the calls you actually make unsupervised — deciding a run is invalid, changing instrumentation placement when the plan does not fit the part, flagging an out-of-family reading instead of logging it — all inside a written test procedure that someone else approved and can override, so the ambiguous high-stakes decision never lands on your desk.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (11/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (3/20) is whether the law requires a licensed human to sign. Trust premium (6/20) is whether buyers specifically pay for a person. Judgment and accountability (8/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 17 of this occupation's 41 points (41%).
Embodiment (13/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 59/100, still EXPOSED.
Task-mix shift: as automated test software and LLM report generation absorb data reduction, spec-checking and write-up, what remains is rig fabrication, sensor mounting, fixture design, and diagnosing why a prototype fails on the floor — work that is inherently non-repeating and instrument-specific. This raises the score only for technicians whose employers keep them on the bench rather than shrinking headcount to match the reduced documentation load.
Certification schemes that name the technician, not just the supervising engineer, on the test record: ASNT NDT Level II/III certification is already required by name on radiographic and ultrasonic inspection reports under ASME BPVC and API 510/570, and NETA ATS acceptance testing requires a certified technician's signature. If more test categories (battery safety per UL 9540A, EMC per FCC/CISPR accredited labs under ISO/IEC 17025 personnel-competence clauses) are pulled into named-signatory regimes, an AI-generated result becomes unusable without a certified human attesting to setup and execution.
A17025-style requirement that the person executing the test declare measurement uncertainty and whether the setup was valid for the stated conditions — i.e., owning the 'this data is trustworthy' call rather than handing raw numbers to an engineer. Accreditation bodies (A2LA, ANAB) already audit personnel competence records; tightening those clauses so the technician's judgment on setup validity is the auditable artifact moves the call to the bench.
Migration of the role toward field test and commissioning work — grid interconnection witness testing, EV battery pack abuse testing, aircraft ground vibration testing — where the environment is uncontrolled and the instrumentation must be installed and debugged on unfamiliar hardware. FAA and IEEE/UL test protocols already require on-site instrumented runs by qualified personnel.
Narrow route only: defense and nuclear supply chains where NAVSEA/NQA-1 or ITAR programs require cleared, badged personnel to perform and witness tests on-site. This is a security and provenance premium on a specific human body, not a general preference for human work.
The limit. The catch-all SOC code is the binding constraint: the certification-signatory route (ASNT, NETA, 17025) genuinely exists but only covers specific test disciplines, so it lifts a subset of these 60k workers, not the category. For general-purpose R&D bench technicians there is no licensing body positioned to claim signature authority and no buyer who cares whether a human mounted the accelerometer. Realistic ceiling for the modal worker is around the mid-50s; the certified NDT or high-voltage acceptance-testing subset can reach higher.
| Houston-Pasadena-The Woodlands, TX | 2,860 | $76,120 -3% |
| Los Angeles-Long Beach-Anaheim, CA | 1,770 | $69,890 -11% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 1,660 | $99,540 +27% |
| San Diego-Chula Vista-Carlsbad, CA | 1,450 | $92,480 +18% |
| Dallas-Fort Worth-Arlington, TX | 1,290 | $71,930 -8% |
| Bremerton-Silverdale-Port Orchard, WA | 1,160 | $108,010 +38% |
| New York-Newark-Jersey City, NY-NJ | 1,120 | $78,890 +1% |
| San Francisco-Oakland-Fremont, CA | 1,090 | $83,940 +7% |
| Portland-South Portland, ME | 750 | $108,850 +39% |
| Bremerton-Silverdale-Port Orchard, WA | 1,160 | $108,010 +38% |
| Baltimore-Columbia-Towson, MD | 930 | $107,140 +37% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 41. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.