Sheet Metal Workers
In short: today's best AI models could do about 18% 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 conversations, in-person and hands-on work, about 82% of the time, stays with people.
Fabricate, assemble, install, and repair sheet metal products and equipment, such as ducts, control boxes, drainpipes, and furnace casings. Work may involve any of the following: setting up and operating fabricating machines to cut, bend, and straighten sheet metal; shaping metal over anvils, blocks, or forms using hammer; operating soldering and welding equipment to join sheet metal parts; or inspecting, assembling, and smoothing seams and joints of burred surfaces. Includes sheet metal duct installers who install prefabricated sheet metal ducts used for heating, air conditioning, or other purposes.
119,770 people do this job in the US. $61,800 a year.
Reviewed on September 30, 2026: what each task is made of, and what handing it over would ask.
A typical 40-hour week
- AI could take on0 h
- A person checks AI's work7 h
- Desk work AI can't do yet0 h
- Stays with people33 h
Sheet Metal Workers, now: of a 40-hour week, AI could take on 0 hours, 7 hours become checking its work, 0 hours are desk work it can't do yet, and 33 hours stay with people.
The largest pieces: Convert blueprints to shop drawings (about 4 h, a person checks); Determine project requirements (about 2 h, a person checks); Commission building systems (under an hour, a person checks).
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.
- 18%
- 0%
- 0%
- 4%
- 78%
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.
What handing it over would take, now
The 18% within reach splits by what a model would need. The dates, pace and reliability follow the controls above. Estimates.
of the working time. The material can be handed to a model and the result handed back.
of the working time. The model has to be connected to the system the work lives in.
is conversation short enough by length. This site leaves conversation with people, so it is not in the total.
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, 6%. Today's estimate: 18%.
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
Of the Claude.ai conversations Anthropic sampled and tied to a work task, no reportable share went to this job's tasks in May 2026; 0 of its 19 tasks showed up. Within each of Anthropic's methods, never across a change: 0.027% to 0.0045% (released February and March 2025); 0.03% to 0.0037% (August 2025 to February 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.027% | 389 of 708 | 1 of 19 |
| Released March 2025 | 0.0045% | 594 of 703 | 1 of 19 |
| The week of Aug 4, 2025 (released September 2025) | 0.03% | 306 of 655 | 1 of 19 |
| The week of Nov 13, 2025 (released January 2026) | 0.0074% | 523 of 672 | 1 of 19 |
| The week of Feb 5, 2026 (released March 2026) | 0.0037% | 599 of 677 | 1 of 19 |
| April 2026 (released June 2026) | none | n/a | 0 of 19 |
| May 2026 (released June 2026) | none | n/a | 0 of 19 |
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 87% 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: 79% in a purpose-built cell, 4% in a structured facility. 0.86 of 3, the 200th highest of the 923 jobs.
- Not physical 13%
- No robot shown 5%
- In a purpose-built cell 79%
- In a structured facility 4%
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.
What changes for the people in this job
From the task-by-task review, by the end of 2028 on the long-run pace at four in five. It says which parts of the work stay with people and which change, not whether anyone keeps a job: that depends on decisions no data set can see.
What stays with people
82% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Fabricate parts on site, the hands-on work, 16%
- Fabricate efficient ducts, the hands-on work, 15%
- Maintain equipment, the hands-on work, 10%
- Install ducts, gutters and flashing, the hands-on work, 8%
- Inspect parts and installations, the hands-on work, 6%
- Fasten seams and joints, the hands-on work, 4%
What becomes checking the machine
18% of the working time: desk work a model could do by then, in tasks where a mistake costs something or a licensed person signs. A person checks it.
- Convert blueprints to shop drawings, costly if wrong, 9%
- Determine project requirements, costly if wrong, 4%
- Commission building systems, costly if wrong, 2%
- Verify green certification standards, costly if wrong, 1%
- Select sheet metal gauges, costly if wrong, 1%
What to learn
- The hands-on craft. About 78% of the time is hands-on work, which no model here does; robots are rated separately on this page.
- Judgment where errors are severe. Some tasks are severe if wrong. Knowing when not to trust an output matters more there than speed.
- Reviewing work you did not do yourself. The tasks under What becomes checking the machine, where a mistake costs something, stay a person's to check: spotting what a draft gets wrong is a skill of its own.
- Briefing a model and checking its drafts. About 18% of the time is documents and data, the part models reach first: writing a clear brief, giving the right material, and checking what comes back.
Your own version of this job: read it against your job description. The headline is free; the full personal report, every task with its timeline and what to learn, is $19 (a sample).
The tasks
Sorted by . 19 tasks from .
| Task | ||||
|---|---|---|---|---|
| Fabricate or alter parts at construction sites, using shears, hammers, punches, or drills.Shears, hammers, punches and drills on site.Done several times a day by most who do it. Robots: in a purpose-built cell, Iowa Precision Fabriduct NEO / Pro-Fabriduct HVAC Full Coil Line (Mestek Machinery) (sold). Anthropic's rating. | 16% | 1.3 h25 min to 2.5 h | Hands-on 100% | Stays with people |
| Fabricate ducts for high efficiency heating, ventilating, and air conditioning (HVAC) systems to maximize efficiency of systems.Ducts made in the shop.Done more than weekly by most who do it. Robots: in a purpose-built cell, Iowa Precision Fabriduct NEO Full Coil Line (Mestek Machinery) (sold). Anthropic's rating. | 15% | 4.2 h1.7 h to 1 work day | Hands-on 100% | Stays with people |
| Maintain equipment, making repairs or modifications when necessary.Repairs and modifications by hand.Done more than monthly by most who do it. Robots: no robot shown. Anthropic's rating. | 10% | 1.3 h25 min to 3.3 h | Hands-on 100% | Stays with people |
| Convert blueprints into shop drawings to be followed in the construction or assembly of sheet metal products.Shop drawings drawn from the blueprints.Done more than monthly by most who do it. | 9% | 4.2 h1.7 h to 6.7 h | Documents 100% | 99%Half by now |
| Install assemblies, such as flashing, pipes, tubes, heating and air conditioning ducts, furnace casings, rain gutters, or downspouts in supportive frameworks.Assemblies installed in the framework.Done daily by most who do it. Robots: in a purpose-built cell, ABB body-in-white robotic cells (sold). Anthropic's rating. | 8% | 2.5 h50 min to 6.7 h | Hands-on 100% | Stays with people |
| Determine project requirements, such as scope, assembly sequences, or required methods or materials, using blueprints, drawings, or written or verbal instructions.Reading the drawings and instructions and working out the job.Done daily by most who do it. | 6% | 1.1 h25 min to 2.5 h | Documents 60%, in person 40% | 60%Desk part half by now |
| Perform building commissioning activities by completing mechanical inspections of a building's water, lighting, or heating, ventilating, and air conditioning (HVAC) systems.Mechanical inspections on site, with the report.Done yearly or less by most who do it.Robots: no robot shown. Anthropic's rating. | 6% | 5 h1.7 h to 1.7 work days | Hands-on 70%, documents 30% | 24%Desk part half by now |
| Inspect individual parts, assemblies, or installations, using measuring instruments, such as calipers, scales, or micrometers.Calipers and micrometers on the work.Done hourly or more by most who do it.Robots: in a purpose-built cell, ZEISS ATOS / ScanBox automated optical 3D measuring machine (Carl Zeiss Industrielle Messtechnik) (sold). Anthropic's rating. | 6% | 38 min13 min to 1.3 h | Hands-on 100% | Stays with people |
| Fasten seams or joints together with welds, bolts, cement, rivets, solder, caulks, metal drive clips, or bonds to assemble components into products or to repair sheet metal items.Welds, rivets, solder and clips on the seams.Done more than weekly by most who do it. Robots: in a purpose-built cell, Iowa Precision Fabriduct NEO Full Coil Line (Mestek Machinery) (sold). Anthropic's rating. | 4% | 50 min13 min to 2.1 h | Hands-on 100% | Stays with people |
| Transport prefabricated parts to construction sites for assembly and installation.Driving prefabricated parts to the site.Done daily by most who do it. Robots: in a structured facility, Seegrid Palion Lift RS1 / CR1 autonomous lift trucks (deployed). Anthropic's rating. | 4% | 1.7 h25 min to 3.3 h | Hands-on 100% | Stays with people |
| Verify that heating, ventilating, and air conditioning (HVAC) systems are designed, installed, and calibrated in accordance with green certification standards, such as those of Leadership in Energy and Environmental Design (LEED).Checking the system on site against the standard.Done more than yearly by most who do it.Robots: no robot shown. Anthropic's rating. | 3% | 2.5 h50 min to 6.7 h | Documents 50%, hands-on 50% | 48%Desk part half by now |
| Lay out, measure, and mark dimensions and reference lines on material, such as roofing panels, using calculators, scribes, dividers, squares, or rulers.Measuring and scribing the lines.Done daily by most who do it. Robots: in a purpose-built cell, Hypertherm mechanized plasma marking (XPR-class systems with ProNest CAD/CAM) (sold). Anthropic's rating. | 3% | 38 min13 min to 1.7 h | Hands-on 100% | 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. Made of: how the task's time splits among five kinds of doing. Only the documents and systems parts can count as within reach, so a task that is 30% documents can be at most 30% 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.
- Structural Metal Fabricators and Fitters4% now, 4% by 2028
- Aircraft Structure, Surfaces, Rigging, and Systems Assemblers4% now, 4% by 2028
- Structural Iron and Steel Workers3% now, 3% by 2028
- Layout Workers, Metal and Plastic13% now, 13% by 2028
- Drywall and Ceiling Tile Installers5% now, 5% by 2028
- Insulation Workers, Floor, Ceiling, and Wall2% now, 2% by 2028
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
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 Sheet Metal Workers is within reach of AI now?
- An estimated 18% of the working time is within reach of AI models now and 18% 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 Sheet Metal Workers 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: Convert blueprints to shop drawings (99% of the task's time now); Determine project requirements (60% of the task's time now); and Commission building systems (24% of the task's time now, 30% by the end of 2028).
- Which tasks of Sheet Metal Workers stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Fabricate parts on site; Fabricate efficient ducts; and Maintain equipment. Together such tasks are about 72% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Sheet Metal Workers learn as AI takes on more of the work?
- By the job's own task-by-task review, the first things to learn are the hands-on craft; judgment where errors are severe; and reviewing work you did not do yourself. By the end of 2028 on the long-run pace at four in five, about 82% of the working time stays with people, 18% becomes checking a model's work and 0% could be handed over with lighter checks. It says which parts of the work change, never whether anyone keeps a job.
- Do people already use AI for the work of Sheet Metal Workers?
- 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 Sheet Metal Workers?
- Anthropic's robot exposure rating (September 2026) finds 87% 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, 4% in a structured facility. Its index is 0.86 of 3. One example: Seegrid Palion Lift RS1 / CR1 autonomous lift trucks for "transport parts to sites". 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 19 tasks O*NET lists for the occupation, each with an estimated length checked against GDPval and a share of the working time, reviewed one task at a time on September 30, 2026 for what each is made of (documents, business systems, conversation, in person, hands-on), 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.
The full personal report is $19, to print or save as a PDF: every task with when it comes within reach, which ones your description names, the parts of your description the task list does not cover, what stays with you, what becomes checking the machine and what to learn.