← Risk register SOC 47-2011 · reviewed 2026-08-11

Boilermakers

10,190 US workers · median $76,410/yr · Construction

SAFE

Boilermakers assemble, weld, and repair pressure vessels, boilers, and tanks inside confined, hot, awkward spaces at refineries, power plants, and shipyards — work no current robot can do at cost. AI touches only the thin paper layer: reading blueprints, weld procedure documentation, and inspection reports. The real constraints are ASME code welding qualifications, rigging skill, and diagnosing corrosion and leaks by hand and eye during tight turnaround windows.

10-year outlook: Employment stays small and cyclical with power-plant and refinery outage cycles, but the hands-on pressure-vessel work is essentially untouched by AI through 2035, and retirements keep qualified welders scarce.

US employment, 2019–2025-35.6%
15,82010,190 workers

Part 2020 shock, part continued decline in the years since.

Median pay $63,100 → $76,410 -3.1% in real terms (nominal +21.1%, less ~25% US inflation over the period)

The job count is not the verdict

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. The marked year is 2020.

BLS projection, 2024–2034

-2.4% 10,400 → 10,100 on the projections basis

Hard to automate, but shrinking anyway

The work resists current AI, yet the BLS projects -2.4% by 2034. Whatever is shrinking this occupation, the evidence does not point to automation — demand, demographics, offshoring and industry decline all shrink jobs that no machine could do.

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.

~800 openings a year on average, including replacing people who leave.

One email if this score changes. Watch as many occupations as you like from the same address — no account, and nothing is sent on a schedule, only when a verdict actually moves.

Also known as — 24 job titles this covers

Titles reported by people doing this work, from the US Department of Labor's O*NET survey. If your job title is here, this page is about your work even though the name doesn't match.

FitterErectorAssemblerBoilermakerBoiler MakerBoiler FitterBoiler SetterBoiler TesterBoiler WelderBoiler ErectorBoiler RelinerBoiler MechanicPressure TesterBoiler InstallerBoiler RepairmanBoilermaker FitterBoilermaker WelderBoiler Shop MechanicBoilerhouse MechanicBoilermaker MechanicBoilermaker JourneymanIndustrial BoilermakerLocomotive BoilermakerBoilermaker Pipe Fitter

Score — 67/100 resistance

Holding it up: embodiment (20/20). Weakest point: trust premium (7/20).

Five dimensions, 0–20 each, summed. Higher means more protected. The arithmetic is shown so you can check it: 18 + 20 + 10 + 7 + 12 = 67. · Scored 2026-08-11, and re-examined when evidence accumulates rather than on a schedule.

Task resistance 18/20

Tasks largely resist digitisation Rolling and bending shell plate, aligning drums with jacks and wedges, back-gouging a root pass overhead inside a mud drum, and replacing tubes through a 16-inch manhole are tasks that require a body in the vessel rather than a shop-floor gantry, so only the takeoff, weld-map, and NDE paperwork sit within reach of software — that's why this is 18 and not a mixed 12.

Embodiment 20/20

Hands-on in uncontrolled environments You work inside the vessel: confined-space entry permits, 120°F stack temperatures during a boiler outage, scaffolds and stagings you build yourself, torch and stinger in hand, in geometry that changes at every plant — this is the ceiling case for uncontrolled physical environments, not merely a field job.

Liability shield 10/20

Certification preferred, not legally required You hold real credentials — ASME Section IX procedure qualifications, an R-stamp or NB shop's welder continuity records, sometimes state boiler-repair endorsements — and a failed radiograph is traced to your stencil, but the Authorized Inspector signs the R-1 form and the certificate holder carries the code liability, which puts this at 10 rather than up with a PE or licensed operating engineer.

Trust premium 7/20

Some relationship component Contractors and plant maintenance managers request specific hands by name for turnaround crews and boilermakers get called back to the same units year after year, but you're dispatched out of a Local hall on manpower calls and the client is buying qualified stencils and headcount, not you personally — hence 7.

Judgment & accountability 12/20

Meaningful discretion Deciding whether wall thinning warrants a pad weld or a full tube section, calling a rigging plan for a 40-ton drum through a boiler house door, and judging fit-up and preheat when the drawing doesn't match the as-built vessel are all discretionary calls — but WPSs, lift plans, and the Authorized Inspector's hold points bound them, so 12 rather than a 16 where you'd own the ambiguity alone.

Scored twice. An independent second run returned 69/100 — SAFE, agreeing with the verdict above.

This score sits on a verdict boundary. At 67/100 it is one point from EXPOSED. Re-scoring moves results by a point or two, so here the score is more informative than the label.

Confidence: high · reviewed 2026-08-11 · how scoring works

What this job involves — and which parts are yours

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.

AI already does these at usable quality

These still need a person

Active moats on the surviving side: embodiment, licensure, judgment

How to future-proof this job

Where to go deeper on what this job runs on: edX — operations management and process monitoring courses free to audit · OSHA Outreach Training — the 10- and 30-hour cards most employers ask for low · Coursera — critical thinking and logic, audit free free to audit · Apprenticeship.gov — industrial maintenance and millwright programs paid to train · MIT OpenCourseWare — circuits, systems and diagnostic method, free and ungated free · Apprenticeship.gov — paid, employer-sponsored trade apprenticeships, searchable by trade and ZIP paid to train

All 35 skills ranked by how many jobs they open →

What would move this back up — beyond any one person

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 81/100, still SAFE.

5 specific changes that would raise this score
  • already happening liability shield +3

    Extension of ASME Section IX welder performance qualification requirements to explicitly cover machine/automatic welding operators with the same continuity-of-qualification and re-test burden as manual welders, so an automated orbital or robotic weld on a pressure boundary still needs a named qualified human accepting it. Already partially in motion via Section IX machine-welding variables and API 577 practice.

  • already happening task resistance +2

    Two-tier shift: if fabrication-shop panel and tube-bank work continues migrating to automated shop welding cells, the field remainder concentrates in confined-space repair, distortion control, and discovered-corrosion diagnosis — the tier no automation touches. This raises measured resistance of the surviving job without any regulatory change, while shrinking headcount.

  • already happening trust premium +2

    Nuclear Section III / NRC 10 CFR 50 Appendix B work and Navy shipyard SUBSAFE specs where owners write manual-process requirements and named-welder traceability into the contract. This is a narrow segment premium for a specific qualified human, not a general buyer preference, and it does not extend to refinery or industrial boiler work.

  • plausible liability shield +4

    State/jurisdictional boiler codes tie liability to the vessel's ASME Code stamp and the National Board R-stamp holder, not to the individual welder. If NBIC or state boiler-safety boards (e.g. the Nat'l Board's push for traceable weld records) added a rule that the named welder-of-record must personally sign each pressure-retaining weld log and remain identified for the life of the repair — the way NDE Level II/III technicians already sign radiographs — the shield would become personal rather than corporate.

  • plausible judgment accountability +3

    Turnaround fitness-for-service calls (weld out vs. replace a corroded section, accept a leaking tube plug) currently sit with the API 510/570 inspector or plant engineer. If insurers underwriting boiler-and-machinery coverage (FM Global, Travellers Equipment Breakdown) began requiring a documented sign-off from the lead craft on discovered-condition scope changes as a condition of coverage, the boilermaker's judgment becomes a recorded, consequential call rather than an informal recommendation.

The limit. Embodiment is already maxed at 20 and cannot rise. The realistic ceiling here is roughly the mid-70s: the binding risk for boilermakers is not AI substitution but demand contraction — coal plant retirements, refinery closures, and shop-prefabrication moving field hours into factories. A higher score on this register would not protect against that.

These are conditions, not forecasts — what would have to happen, not what will. Specific rules, cases and bills are named so you can go and check whether they exist and where they stand; verify before relying on any of them. Nothing here is legal or financial advice.

Where this work is, and what it pays there

BLS metro figures for 20 areas. The verdict above does not change by city — the rubric judges what the work involves, not where it happens — but pay and headcount do, and the national median hides a very wide range.

Most of these jobs

Houston-Pasadena-The Woodlands, TX 1,170 $77,090 +1%
Baton Rouge, LA 680 $74,730 -2%
New York-Newark-Jersey City, NY-NJ 330 $82,420 +8%
Los Angeles-Long Beach-Anaheim, CA 320 $110,370 +44%
Beaumont-Port Arthur, TX 230 $73,060 -4%
Corpus Christi, TX 170 $64,660 -15%
Philadelphia-Camden-Wilmington, PA-NJ-DE-MD 140 $86,910 +14%
New Orleans-Metairie, LA 120 $77,290 +1%

Best paid

San Francisco-Oakland-Fremont, CA 50 $135,290 +77%
Los Angeles-Long Beach-Anaheim, CA 320 $110,370 +44%
Chicago-Naperville-Elgin, IL-IN 90 $109,640 +43%

Percentages are against this occupation's national median of $76,410. Counts are jobs in that metro, not vacancies. Metros where the BLS suppressed the cell are absent rather than shown as zero.

Who is actually doing this — nobody, on the record

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 67. 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.

Read that as a gap in the reporting we can see, not proof of absence — the dispatch runs on English-language feeds and misses plenty. If you know of a case, tell us, or add a field report from inside the job.

Quick take — do you do this job?

Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.

Self-reported and unverified — a sentiment signal, not a survey. One response per person per occupation; you can change your answer.

Field reports — what people say has changed

No field reports yet. A written account takes a paragraph rather than a tap, goes to an editor before it appears, and is the one thing on this page the rubric cannot produce on its own.

File a field report

Concrete beats general: a tool that arrived, a task that moved, a headcount decision you watched happen. Don't include anything that identifies you or your employer if that would put you at risk.

Watch this verdict
Kept current

Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.