Boilermakers
In short: today's best AI models could do about 3% of this job's working time (projected from METR's April 2026 measurement and Epoch's November 2026 index) if the work were set up for them, and about the same by the end of 2028 if progress keeps its long-run pace. The people and hands-on work, about 97% of the time, stays with people.
Construct, assemble, maintain, and repair stationary steam boilers and boiler house auxiliaries. Align structures or plate sections to assemble boiler frame tanks or vats, following blueprints. Work involves use of hand and power tools, plumb bobs, levels, wedges, dogs, or turnbuckles. Assist in testing assembled vessels. Direct cleaning of boilers and boiler furnaces. Inspect and repair boiler fittings, such as safety valves, regulators, automatic-control mechanisms, water columns, and auxiliary machines.
10,190 people do this job in the US. $76,410 a year.
A typical 40-hour week
Not yet reviewed task by task- AI could take on1 h
- A person checks AI's work0 h
- Desk work AI can't do yet0 h
- Stays with people39 h
Boilermakers, now: of a 40-hour week, AI could take on 1 hours, 0 hours become checking its work, 0 hours are desk work it can't do yet, and 39 hours stay with people.
The largest pieces: Study blueprints to determine locations, relationships, or dimensions of parts (about 1 h).
Estimates for a typical week of this job, if progress keeps its long-run pace, counting work the best models finish four times in five. Every task is below.
of the working time. Now is projected from METR's April 2026 measurement and Epoch's November 2026 index.
- 3%
- 0%
- 97%
of the work: Anthropic's measure of what people already bring to Claude at work (2025). It counts help with part of a task, so it can be higher than the share within reach.
On the chosen pace and reliability. Estimate.
Share of the working time within reach, now to 2030
Change assumptionsHide assumptionsLong-run pace, four in five
What each choice means: , , , , . Every pace starts from .
At this date and pace, models finish tasks up to 2.4 work days long four times in five.
How far this has come
With GPT-4, measured in March 2023, 0% of this job's working time was within reach at . With Claude 3.7 Sonnet, measured in February 2025, 0%. With Claude Mythos Preview (early), measured in April 2026, 3%. Today's estimate: 3%.
A backcast: today's task data with the best model METR had measured by each date. It shows how fast the models have moved, not what anyone forecast at the time.
Measured use over time
None of this job's 18 tasks showed up in the Claude.ai conversations Anthropic sampled in any of its 7 releases, from February 2025 to May 2026.
Measured, not estimated: Anthropic's samples of Claude.ai conversations, each tied by a model to an O*NET task, from the Anthropic Economic Index. It says where Claude's use goes, not how much of this job's working time it covers, so it is not on the same scale as the reach above. The releases differ in their sample, their classifier and the smallest share they report (the June 2026 file rounds each task to a hundredth of a percent, so a smaller task reads as none), which is why a change of method starts a new line. At least 96% of each older release's task-tied use matched this site's tasks, and all of the newest. Claude.ai only, so use through other tools and other companies' models is not in it. Every job ranked.
Robots and the physical work
Anthropic rates 94% of this job's working time as physical work. For 82% of the working time, a robot has been shown doing most of the task somewhere: 76% in a purpose-built cell, 6% in a structured facility. 0.88 of 3, the 189th highest of the 923 jobs.
- Not physical 6%
- No robot shown 12%
- In a purpose-built cell 76%
- In a structured facility 6%
Anthropic's rating, not this site's estimate: Claude Opus 5, searching the web, rated each physical O*NET task by the kind of setting in which a robot it could cite (deployed, sold or demonstrated) does most of the task, weighted by its own estimate of each task's share of the time (What work can robots do?, September 30, 2026, data CC BY 4.0; every job ranked). A robot shown somewhere is not a robot in your building: cost, setup and safety decide that. The share within reach on this page still counts no body work. The robot behind each rated task is under the task below.
The tasks
Sorted by . 18 tasks from .
| Task | Kind | |||
|---|---|---|---|---|
| Assemble large vessels in an on-site fabrication shop prior to installation to ensure proper fit.Done more than weekly by most who do it.Robots: in a purpose-built cell, PEMA Assembly & welding station (Pemamek) (sold). Anthropic's rating. | 24% | 1.7 work days3.3 h to 4.2 work days | Stays with people | |
| Shape seams, joints, or irregular edges of pressure vessel sections or structural parts to attain specified fit of parts, using cutting torches, hammers, files, or metalworking machines.Done more than monthly by most who do it. Robots: in a purpose-built cell, PeddiBot-1200 robotic thermal cutting machine (Staubli robot arm, HGG-powered) (Peddinghaus Corporation) (sold). Anthropic's rating. | 11% | 3.3 h1.3 h to 6.7 h | Stays with people | |
| Examine boilers, pressure vessels, tanks, or vats to locate defects, such as leaks, weak spots, or defective sections, so that they can be repaired.Done more than weekly by most who do it. Robots: in a structured facility, Gecko Robotics TOKA crawlers + Cantilever software (deployed). Anthropic's rating. | 8% | 2.5 h50 min to 6.7 h | Stays with people | |
| Repair or replace defective pressure vessel parts, such as safety valves or regulators, using torches, jacks, caulking hammers, power saws, threading dies, welding equipment, or metalworking machinery.Done more than yearly by most who do it. Robots: in a purpose-built cell, PEMA Robotic Nozzle Welding Station with WeldControl 300 SCAN (Pemamek) (sold). Anthropic's rating. | 6% | 3.3 h1.7 h to 6.7 h | Stays with people | |
| Lay out plate, sheet steel, or other heavy metal and locate and mark bending and cutting lines, using protractors, compasses, and drawing instruments or templates.Done more than weekly by most who do it. Robots: in a purpose-built cell, Hypertherm plasma marking on mechanized CNC cutting tables (with ProNest CAD/CAM) (sold). Anthropic's rating. | 6% | 1.7 h50 min to 3.3 h | Stays with people | |
| Shape or fabricate parts, such as stacks, uptakes, or chutes, to adapt pressure vessels, heat exchangers, or piping to premises, using heavy-metalworking machines such as brakes, rolls, or drill presses.Done more than yearly by most who do it. Robots: in a purpose-built cell, Faccin 4HEL/4HEP plate rolls with PGS-Ultra CNC and Wind Tower Package (deployed). Anthropic's rating. | 6% | 3.3 h1.7 h to 1.3 work days | Stays with people | |
| Position, align, and secure structural parts or related assemblies to boiler frames, tanks, or vats of pressure vessels, following blueprints.Done more than weekly by most who do it. Robots: no robot shown. Anthropic's rating. | 6% | 2.5 h1.3 h to 6.7 h | Stays with people | |
| Inspect assembled vessels or individual components, such as tubes, fittings, valves, controls, or auxiliary mechanisms, to locate any defects.Done more than weekly by most who do it. Robots: in a purpose-built cell, Gecko Robotics TOKA 3/4/Flex crawlers (sold). Anthropic's rating. | 5% | 1.7 h38 min to 4.2 h | Stays with people | |
| Bell, bead with power hammers, or weld pressure vessel tube ends to ensure leakproof joints.Done more than monthly by most who do it. Robots: in a purpose-built cell, SwiftRoll X1 automatic rolling & machining robot (KRAIS Tube Expanders) (sold). Anthropic's rating. | 5% | 2.5 h50 min to 1 work day | Stays with people | |
| Conduct pressure tests on vessels, such as boilers.Done yearly or less by most who do it. | 4% | 2.5 h50 min to 6.7 h | Stays with people | |
| Bolt or arc weld pressure vessel structures and parts together, using wrenches or welding equipment.Done more than monthly by most who do it. Robots: in a purpose-built cell, PEMA boiler panel lines / tube shop welding lines (Pemamek Oy) (deployed). Anthropic's rating. | 4% | 3.3 h1.3 h to 6.7 h | Stays with people | |
| Clean pressure vessel equipment, using scrapers, wire brushes, and cleaning solvents.Done more than yearly by most who do it. Robots: in a purpose-built cell, ICM boiler tube wall climbing cleaning robot (demonstrated). Anthropic's rating. | 3% | 1.3 h25 min to 3.3 h | Stays with people |
Within reach, per task: the share of the task's time whose instances are short enough for a model to finish at the chosen date, pace and reliability. Half by: the date half of that time comes within reach.
Jobs that share skills with this one
From O*NET's related occupations. Within reach now and by the end of 2028, long-run pace, four in five.
- Plumbers, Pipefitters, and Steamfitters23% now, 23% by 2028
- Millwrights0% now, 0% by 2028
- Helpers--Pipelayers, Plumbers, Pipefitters, and Steamfitters0% now, 0% by 2028
- Engine and Other Machine Assemblers9% now, 9% by 2028
- Stationary Engineers and Boiler Operators21% now, 22% by 2028
- Aircraft Structure, Surfaces, Rigging, and Systems Assemblers4% now, 4% by 2028
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
- Utility System Construction2,130 people
- Building Equipment Contractors1,830 people
- Nonresidential Building Construction1,630 people
- Fabricated Metal Product Manufacturing (3323 and 3324 only)840 people
- Commercial and Industrial Machinery and Equipment (except Automotive and Electronic) Repair and Maintenance740 people
Where the most of these jobs are
The metro areas with the most people in it (BLS, May 2025).
Questions about this job
- How much of the work of Boilermakers is within reach of AI now?
- An estimated 3% of the working time is within reach of AI models now and 3% by the end of 2028, on METR's long-run pace at four in five. Within reach means a model could finish work that long at that reliability. It is not a forecast of jobs lost or of savings.
- Which tasks of Boilermakers come within reach first?
- Weighted by the time each takes, the tasks with the most within reach now (projected from METR's April 2026 measurement and Epoch's November 2026 index) are: Study blueprints to determine locations, relationships, or dimensions of parts (100% of the task's time now).
- Which tasks of Boilermakers stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Assemble large vessels in an on-site fabrication shop prior to installation to ensure proper fit; Shape seams, joints, or irregular edges of pressure vessel sections or structural parts to attain specified fit of parts, using cutting torches, hammers, files, or metalworking machines; and Examine boilers, pressure vessels, tanks, or vats to locate defects, such as leaks, weak spots, or defective sections, so that they can be repaired. Together such tasks are about 97% of the working time.
- Do people already use AI for the work of Boilermakers?
- In the Claude.ai conversations Anthropic sampled for May 2026, no reportable share went to this job's tasks. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 0%. These are measures of use, not of how much of the work AI could finish, and they cover Claude only.
- Can robots do the physical work of Boilermakers?
- Anthropic's robot exposure rating (September 2026) finds 94% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 82% of the working time, 6% in a structured facility. Its index is 0.88 of 3. One example: Gecko Robotics TOKA crawlers + Cantilever software for "examine boilers, pressure vessels, tanks, or vats to locate defects, such as leaks, weak spots, or defective sections, so that they can be repaired". A robot shown somewhere is not one in every workplace; this site's own estimates still count no physical work as within reach.
- How is this estimated?
- From the 18 tasks O*NET lists for the occupation, each with an estimated length checked against GDPval and a share of the working time, each classed as desk, people or body work, and METR's measurements of how long a task the best AI models can finish, carried forward to the date from the latest, Claude Mythos Preview (early) in April 2026. Data checked November 10, 2026.
This job in your state or metro area
Embed these figures on your site
A small card with the share of working time within reach now and at the end of 2027, 2028 and 2030, the date and a link back. Free under CC BY 4.0; the card carries the credit.
Your own version of this job
The page above is the occupation. Your job is your title and what you actually do: give both, and they are read against these tasks in your browser, nothing sent anywhere. The headline is free: how much of your working time is within reach now and by the end of 2028, and how many of the tasks your description names.
This occupation has not been reviewed task by task yet, so the full personal report is not offered for it; 402 of 923 occupations are reviewed so far, largest first.