Why this work stays in the trench
Ask whether AI will replace plumber helpers and the answer starts with where the work happens. Helpers dig and backfill trenches, carry and stage pipe, cut and thread lengths to the marks a journeyman gives them, and hold fittings in place while joints are made. None of that is text or data. It is weight, mud, tight crawl spaces and a hundred small judgments about whether a length will clear a joist.
The second reason is the shape of a job site. A software model can read a plan. It cannot feel that a trench wall is loose, notice the smell of sewer gas, or decide that the pipe run has to shift four inches because the framing is not where the drawing says. Helpers also clean work areas, pass tools, and keep materials moving so the licensed tradesperson never stops working. That coordination is physical and social at the same time.
Third, this job is a training post. Helpers learn by standing next to someone doing the work, which is also how the trade replaces retiring plumbers and pipefitters. The Bureau of Labor Statistics counted about 44,330 people in this occupation with median pay of $42,360, and projects employment growth of 2.4% from 2025 to 2035 (BLS, 2025). Modest growth, but growth.
What AI handles, what it assists, and what people keep
Our task review puts no part of this job in the group AI can do on its own, so the share of task time AI handles today reads 0%. There is no helper task that a model finishes end to end without a person present.
The assist group is also empty, at 0%. That does not mean software is absent from the site. Scheduling apps, digital plans and photo-based inspection notes all show up around this crew. They sit with the contractor and the licensed tradesperson, not with the tasks a helper is paid to perform.
Everything else stays with people: 100% of tracked task time. Cutting, measuring and threading pipe to a fitter’s marks stays human because the tolerance is physical. So does digging and grading trenches for sewer and water line, where soil, utilities and shoring change hour by hour. Fitting valves and couplings, and holding assemblies while they are soldered or glued, needs two sets of hands in the same small space. Our coverage score for this job is 3 out of 100, which tracks how much of that task time AI can handle today. The coverage method page explains how that share is built.
What the evidence actually shows
There is no published head-to-head test of AI against people doing helper work in plumbing or pipefitting. Our evidence grade for quality parity is D, which is the grade we use when nothing has been measured directly, so we publish no parity number for this job. Guessing one would be worse than leaving it blank.
What would settle it is specific: a timed trial of a mobile robot trenching, cutting and staging pipe on a live residential or commercial job, scored for joint quality, rework and safety against a first-year apprentice. Until that exists, the honest read is an engineering problem, not an open question about language models. Our full scoring approach sits on the methodology page, and how parity is graded explains why a grade can exist without a score.
Good to know: every task in this job is classed as physical work, and the automation that exists for pipe handling today is fixed automation in a shop, not a machine that follows a crew into a basement.
When this could change
Most likely after 2048 (8 in 10 of our scenarios). Two things could pull that window closer. Mobile manipulation is improving faster than most forecasts assumed, and trench and pipe-laying machinery on large utility projects already automates parts of the dig. If modular, prefabricated pipe assemblies become standard on more buildings, more cutting and fitting moves into a controlled shop where machines work well.
Two things push it back. The hardware cost curve is the hard one: our cost notes show AI software for this job is cheap next to labor, but a machine that can do the physical half is not software, and it is the machine that is missing. Site variability is the other. Older buildings, buried utilities, code inspections and union apprenticeship rules all slow the swap. The replacement-year method sets out what the window does and does not claim.
How to stay needed on site
Lean into the parts of the day nobody can do remotely. First, trench and excavation work, including shoring, grading and spotting buried utilities. Second, precise cutting, threading and dry-fitting so the licensed tradesperson joints clean the first time. Third, material staging and site readiness, which is the skill that gets a helper asked for by name.
Two skills to build on top of that. One, read plans and isometrics well enough to anticipate the next move instead of waiting for instruction. Two, get into a registered apprenticeship and document hours toward a journeyman license, because the license is the part of the trade that is both regulated and in demand.
Nearby work worth comparing: plumbers, pipefitters and steamfitters is the direct next step up; helpers for electricians and pipelayers share most of the same day. You can put any two of them next to each other on the job comparison tool, read the wider picture on the construction trades helpers family page and the construction sector page, or see where hands-on trades sit on the list of jobs that mostly need a person.