Anthropic's robot exposure index rates operating engineers at 1.6 out of 3 and electricians at 0.33

Wed, 30 Sep 2026 19:27:56 -0400

Anthropic’s economics research team published “What work can robots do?” on September 30, 2026, with an index that scores every US occupation on how much of its work robots can do today. The authors, Russell Legate-Yang and Maxim Massenkoff, had Claude, Anthropic’s AI model, search the web for robots able to do each of the 7,594 physical tasks in O*NET, the Labor Department’s database of job tasks. The index runs from 0, where robots can do none of a job’s tasks, to 3, where they can do all of them in unstructured places such as a jobsite. Operating engineers and other construction equipment operators score 1.6. Electricians score 0.33, and construction managers score 0.

What the index measures

Each physical task gets a tier by where a robot can do it. E0 means nowhere. E1 means a setting built around the machine, such as a fenced cell where fixtures hold each piece in a known position. E2 means a structured workplace built for people, such as a warehouse or a hospital. E3 means a place taken as found, and the rubric names “a construction site” as an example. An occupation’s index averages its task tiers, with E1 counting 1 and E3 counting 3, weighted by the share of work time each task takes.

Claude splits each task into typical instances, rates each one, and gives the task the highest tier that covers at least half the weight. For “Dig trenches,” Claude rated straight trenching in open ground, 25% of the task, at E3, citing the Built Robotics Exosystem kit that lets a hydraulic excavator dig on its own. Hand digging around buried pipes and cables, another 20%, is E0. More of the task looks like the careful digging, so the task is E0.

Robots do the least in E3 settings. Weighted by work time, 26.3% of physical work is E0, 49.8% is E1, 22.0% is E2 and 1.9% is E3. Counting every tier, the authors report that robots can do 74% of physical work somewhere, which is 34% of all work hours.

Robot exposure tiers, share of US work time
Tier	Where a robot can do the task	Example setting in the rubric	Physical tasks rated	Share of physical work (%)	Share of all work (%)
Not physical	Cognitive and interpersonal work				54
E0	No environment		2,419	26.3	12
E1	Purpose-built robotic work environment	A fenced machine-tending cell, a bottling line, an automotive paint booth	3,967	49.8	23
E2	Structured human work facility	A hospital pharmacy, a hotel corridor, an airport concourse, a container port	1,103	22.0	10
E3	Unstructured environment	A private home, a construction site, a public sidewalk	105	1.9	1

Source: anthropic.com

Where construction occupations land

The data covers 49 construction occupations in SOC group 47, the federal code for construction and extraction jobs, leaving out extraction. Their median score is 0.88. Seventeen score 1.0 or higher and eight score below 0.5. The top of the whole index is taxi drivers at 2.22.

Equipment and finish work lead. Paving, surfacing and tamping equipment operators score 1.85, and the authors cite an XCMG fleet of unmanned pavers and rollers. Operating engineers score 1.6, tapers 1.59 and brickmasons and blockmasons 1.42. The mechanical and electrical trades sit low, with plumbers and pipefitters at 0.51, HVAC mechanics at 0.46 and electricians at 0.33. Carpenters score 0.70. Cost estimators score 0.07 and construction managers 0.

Pile driver operators score 0.09, which looks odd next to the appendix’s mention of a pile-driving robot on solar farm sites. This post uses the occupation file, which gives one score per occupation without the task detail.

Robot exposure index, construction occupations, 0 to 3
SOC code	Occupation	Group	Robot exposure index (0 to 3)
47-2071.00	Paving, surfacing, and tamping equipment operators	Equipment operators	1.85
47-2073.00	Operating engineers and other construction equipment operators	Equipment operators	1.60
47-4061.00	Rail-track laying and maintenance equipment operators	Equipment operators	1.21
53-7021.00	Crane and tower operators	Equipment operators	0.92
47-2072.00	Pile driver operators	Equipment operators	0.09
47-2082.00	Tapers	Finishes and masonry	1.59
47-2021.00	Brickmasons and blockmasons	Finishes and masonry	1.42
47-2053.00	Terrazzo workers and finishers	Finishes and masonry	1.25
47-2022.00	Stonemasons	Finishes and masonry	1.20
47-2081.00	Drywall and ceiling tile installers	Finishes and masonry	1.08
47-2161.00	Plasterers and stucco masons	Finishes and masonry	1.07
47-2051.00	Cement masons and concrete finishers	Finishes and masonry	1.03
47-2043.00	Floor sanders and finishers	Finishes and masonry	1.00
47-2142.00	Paperhangers	Finishes and masonry	0.96
47-2042.00	Floor layers, except carpet, wood, and hard tiles	Finishes and masonry	0.94
47-2131.00	Insulation workers, floor, ceiling, and wall	Finishes and masonry	0.93
47-2044.00	Tile and stone setters	Finishes and masonry	0.91
47-2041.00	Carpet installers	Finishes and masonry	0.75
47-2141.00	Painters, construction and maintenance	Finishes and masonry	0.67
47-4091.00	Segmental pavers	Finishes and masonry	0.32
47-4051.00	Highway maintenance workers	Sitework and structure	1.38
47-2061.00	Construction laborers	Sitework and structure	1.05
47-4031.00	Fence erectors	Sitework and structure	1.04
47-2151.00	Pipelayers	Sitework and structure	1.02
47-2221.00	Structural iron and steel workers	Sitework and structure	0.93
47-2121.00	Glaziers	Sitework and structure	0.92
47-2171.00	Reinforcing iron and rebar workers	Sitework and structure	0.92
47-4071.00	Septic tank servicers and sewer pipe cleaners	Sitework and structure	0.76
47-4041.00	Hazardous materials removal workers	Sitework and structure	0.72
47-2031.00	Carpenters	Sitework and structure	0.70
47-2181.00	Roofers	Sitework and structure	0.51
47-4099.03	Weatherization installers and technicians	Sitework and structure	0.51
49-9051.00	Electrical power-line installers and repairers	Mechanical, electrical and plumbing	0.92
47-2011.00	Boilermakers	Mechanical, electrical and plumbing	0.88
47-2211.00	Sheet metal workers	Mechanical, electrical and plumbing	0.86
47-2152.04	Solar thermal installers and technicians	Mechanical, electrical and plumbing	0.59
47-2231.00	Solar photovoltaic installers	Mechanical, electrical and plumbing	0.53
47-2152.00	Plumbers, pipefitters, and steamfitters	Mechanical, electrical and plumbing	0.51
49-9021.00	Heating, air conditioning, and refrigeration mechanics and installers	Mechanical, electrical and plumbing	0.46
47-2132.00	Insulation workers, mechanical	Mechanical, electrical and plumbing	0.36
47-2111.00	Electricians	Mechanical, electrical and plumbing	0.33
47-4021.00	Elevator and escalator installers and repairers	Mechanical, electrical and plumbing	0.26
47-3011.00	Helpers, brickmasons, blockmasons, stonemasons, and tile and marble setters	Helpers	1.31
47-3012.00	Helpers, carpenters	Helpers	1.28
47-3013.00	Helpers, electricians	Helpers	0.85
47-3015.00	Helpers, pipelayers, plumbers, pipefitters, and steamfitters	Helpers	0.80
47-3016.00	Helpers, roofers	Helpers	0.79
47-3014.00	Helpers, painters, paperhangers, plasterers, and stucco masons	Helpers	0.78
51-4121.00	Welders, cutters, solderers, and brazers	Fabrication shop	0.96
51-2041.00	Structural metal fabricators and fitters	Fabrication shop	0.93
47-1011.00	First-line supervisors of construction trades and extraction workers	Supervision and office	0.73
17-3031.00	Surveying and mapping technicians	Supervision and office	0.56
17-2051.00	Civil engineers	Supervision and office	0.29
47-4011.00	Construction and building inspectors	Supervision and office	0.20
17-1022.00	Surveyors	Supervision and office	0.17
13-1051.00	Cost estimators	Supervision and office	0.07
47-1011.03	Solar energy installation managers	Supervision and office	0.05
47-4011.01	Energy auditors	Supervision and office	0.04
11-9021.00	Construction managers	Supervision and office	0.00
53-3054.00	Taxi drivers	Reference, outside construction	2.22

Source: huggingface.co

A trade can score high while much of its work stays manual. Pressing paper tape into joint compound is E0, since mud and tape are messy materials for a robot. Applying the later coats is E3, citing the Canvas drywall finishing robot, which scans walls, sprays the coats and sands them. By work time, 41% of tapers’ tasks are E3 and 26% are E0. For carpenters, no robot can yet install windows, frames, flooring or trim.

Construction tasks the paper rates by name
Occupation	Task statement (O*NET)	Tier	Robot the authors cite	Where the paper discusses it
Paving, surfacing, and tamping equipment operators	Start machine, engage clutch, and push and move levers to guide machine along forms or guidelines and to control the operation of machine attachments.	E3	XCMG unmanned paver and roller fleet	Figure 4
Tapers	Apply additional coats to fill in holes and make surfaces smooth.	E3	Canvas drywall finishing robot	Robot exposure section
Tapers	Press paper tape over joints to embed tape into sealing compound and to seal joints.	E0	None	Robot exposure section
Not named in the paper	Weld components in flat, vertical, or overhead positions.	E1	Path Robotics welding cells	Measuring exposure section
Painters, construction and maintenance	Apply paint, stain, varnish, enamel, or other finishes to equipment, buildings, bridges, or other structures, using brushes, spray guns, or rollers.	E1	Robot painters not named in the appendix	Appendix C.2
Carpenters	Erect scaffolding or ladders for assembling structures above ground level.	E0	None	Figure 2
Carpenters	Install structures or fixtures, such as windows, frames, floorings, trim, or hardware, using carpenters' hand or power tools.	E0	None	Appendix C.1
Not named in the paper	Dig trenches.	E0	Built Robotics Exosystem, for straight trenches in open ground	Measuring exposure section

Source: cdn.sanity.io

Helpers score higher than the trades they work for. Helpers of carpenters score 1.28 against 0.70 for carpenters, helpers of electricians 0.85 against 0.33, and helpers of roofers 0.79 against 0.51. Those are the three helper categories that cover a single trade. One reading is that the robot-capable work on a crew sits with the entry-level positions, where firms train their next journeymen.

Robot exposure index, trades and their helpers, 0 to 3
Trade	Trade index	Helper category	Helper index	Helper category covers only this trade
Carpenters	0.70	Helpers, carpenters	1.28	Yes
Electricians	0.33	Helpers, electricians	0.85	Yes
Roofers	0.51	Helpers, roofers	0.79	Yes
Plumbers, pipefitters, and steamfitters	0.51	Helpers, pipelayers, plumbers, pipefitters, and steamfitters	0.80	No, also pipelayers
Painters, construction and maintenance	0.67	Helpers, painters, paperhangers, plasterers, and stucco masons	0.78	No, also paperhangers and plasterers
Brickmasons and blockmasons	1.42	Helpers, brickmasons, blockmasons, stonemasons, and tile and marble setters	1.31	No, also stonemasons and tile setters

Source: huggingface.co

What the costs say

The authors estimate cost in a separate step. Claude estimated the full annual cost of the robots cited for each task, and the authors compared it with BLS total compensation for the share of the job robots can do. On that basis, robots cost less than people for 0.3% of all work.

Welders are the closest occupation to construction in the cost table. Robots can do 90% of their work by time, which at median total compensation of $73,400 is about $66,100 of labor a year. The robots would cost an estimated $334,600 a year, about five times as much, and the example system is a Zeman steel-beam assembler line at $2.7 million upfront.

Robot and labor cost for five physical occupations
Occupation	Employment (thousands)	Median total compensation ($ thousands)	Exposed tasks (%)	Labor cost of exposed tasks ($ thousands, computed)	Annual robot cost ($ thousands)	Fixed ($ thousands)	Variable ($ thousands)	Robot cost as a multiple of labor cost (computed)	Example robot	Example robot upfront cost ($ thousands)
Packers and packagers, hand	560	49.0	97	47.5	45.4	22.2	23.2	0.96	Douglas top-load case packer	160.0
Taxi drivers	41	56.8	89	50.6	57.8	27.4	30.4	1.14	Waymo robotaxi	135.0
Dishwashers	477	45.0	100	45.0	172.0	97.9	74.1	3.82	Armstrong Robotics dishwashing system	165.0
Janitors and cleaners	2,210	47.7	73	34.8	280.0	143.7	136.3	8.04	Somatic-class cleaning mobile manipulator	100.0
Welders, cutters, solderers, and brazers	416	73.4	90	66.1	334.6	160.0	174.6	5.07	Zeman steel-beam assembler line	2,700.0

Source: anthropic.com

Robot prices have fallen about 3% a year since the 1990s, in the data the authors use. At that rate, robots would be cheaper than labor for 0.8% of all work in about seven years and for 10% in about 40 years. The authors find regulation matters less for construction tasks than in healthcare or education.

Evidence so far

The evidence on the index comes from its authors at Anthropic, which makes the model that did the rating. They rated 1977 job tasks against 1977 robots and followed each occupation through 2024. Compared with an unexposed occupation in the same industry, one whose tasks robots could all do had wages 7.1% lower 20 years later and lost 34.2% of its jobs, with a 95% confidence interval of 16.4% to 52.0% on the job figure (the range the data support at that level of certainty).

Claude could count a robot that does a related task, such as the solar pile driver. Dropping those ratings cuts the exposed share of physical work from about three-quarters to about half. The appendix also notes that a bricklaying robot now spreads the mortar bed for block, a task no robot could do in 1977.

How to measure it on your projects

The data joins on SOC codes, so a contractor can score its own self-performed work in Excel or Power BI:

Export the last 12 months of craft hours by labor classification from payroll. That is the baseline.
Map each classification to a SOC code, such as 47-2111 for electricians.
Multiply each classification’s hours by its index, add the results, and divide by total craft hours.

For a pilot of a robot the paper cites, such as a drywall finishing robot, track labor hours per 1,000 square feet finished, by coat, with the baseline from the last three comparable jobs. Keep taping and finish coats in separate cost codes, since the paper rates them at opposite ends of the scale. A first step is to score the three largest crafts on one active job.

Availability and cost

The paper and data are free. The data is in Anthropic’s Economic Index dataset on Hugging Face under a CC BY 4.0 license, with the index for 923 occupations and ratings for 18,796 tasks, including Claude’s reasoning and the robots it cited.

Caveats

The index counts a task as exposed when a robot can do it anywhere, as of 2026. Adoption on a given job depends on cost and site conditions. The counts in this post weight each occupation equally, and the cost figures are model estimates that the authors call suggestive.

Robotics | Construction Metrics