Why the press line still runs on people
The cutting, punching and pressing itself was automated decades ago. That is the easy part. The hours in this job go into everything around the stroke of the press: setting up the machine for the next part, changing and aligning dies, feeding and positioning stock, clearing jams, and pulling sample parts to measure against the print. Software does not bolt a die into a ram or feel a slug sticking in a hole.
Setup is the clearest example. A job comes in with a work order and a drawing, and the operator picks tooling, sets stroke and pressure, runs a first piece, measures it with calipers or a micrometer, then nudges the settings until the part is in tolerance. The decisions are small and constant, and they depend on this machine, this coil of steel, this worn punch. When a run goes wrong, somebody has to read the part, trace the cause and fix it without wrecking the tooling.
Our Still needs a human score for this job is 85 out of 100 (higher is safer). That reflects a task mix that is mostly physical and mostly on the machine, not a claim that nothing changes here. The honest pressure in metal and plastic work is fewer openings and leaner crews, not the job disappearing. The Bureau of Labor Statistics counts about 171,200 of these jobs in the US, with employment projected to fall roughly 10.5% between 2025 and 2035, and median pay of $46,330 (BLS, 2025).
What machines run, what AI assists, and what stays with the operator
Start with the tasks our data puts in the “AI does it” group, which covers 0% of task time. These are the routine, repeatable pieces: running a proven program through its cycle, and logging production counts and basic machine data. Once a job is dialed in, the controller keeps the rhythm and the record.
Next, the assisted work, at 8% of task time. Vision systems can flag surface defects or out-of-spec parts on the line. Scheduling and maintenance software can predict when a tool needs sharpening or a bearing needs attention. In both cases the operator still decides whether to stop the run, swap the tooling or accept the part. The method behind that split is explained on our coverage scoring page.
The rest, 92% of task time, sits with people. That includes setting up and adjusting machines for each new part, installing and aligning dies and punches, loading and positioning heavy or awkward stock, clearing jams and misfeeds, and inspecting finished pieces by hand against a drawing. It also includes the safety calls: locking out a machine, noticing a sound that is not right, deciding a run should stop.
What the evidence shows, and what it does not
Our quality-parity grade for this occupation is D. A grade of D means there is no direct published test of AI or robots against a qualified operator doing this job’s real tasks, so we give no parity number at all. Nobody has run a fair head-to-head on die setup, first-article inspection and jam recovery in a working press shop.
What would settle it is specific: a timed, independent trial where an automated cell and a human setter both change over a press to a new part, qualify the first piece, and then handle a deliberate misfeed or a worn punch over a full shift. Until something like that is published, claims that this work is already matched belong in the marketing column, not the evidence column. You can read how we grade evidence from A to D on our quality parity method page, and see the full method at how we score jobs.
When this could change
Most likely after 2046 (8 in 10 of our scenarios). What that window measures, and how we build it, is set out on the replacement year method page.
Two things could pull it earlier. First, hardware: about 90.3% of this job’s exposure is physical, and the robotics tier it needs is mobile robots rather than a fixed arm bolted to one station. Cheaper, more capable mobile units that can move between presses and handle their own tool changes would change the math. Second, cost: automating a full changeover cell is still a capital project for a shop running short, varied runs, and falling hardware prices shift that line.
Two things hold it back. Variety is the big one. Short runs, mixed materials and old presses mean every setup is slightly different, and fully automatic changeover suits long, stable runs. The other is risk. Presses can maim people and destroy tooling in a fraction of a second, so machine guarding, lockout procedures and liability all push shops toward a trained person who is responsible for the machine. Both show up in the blockers listed above.
Good to know: the pressure on this occupation shows up first as fewer entry-level tender roles, while setup and troubleshooting work stays with experienced hands.
How to stay needed on the floor
Lean into the tasks that carry the most weight here. Get fast and reliable at setup and changeover, including die installation and alignment. Own first-article inspection: measuring, reading prints and geometric tolerances, and documenting why a part passed. And get good at diagnosing bad runs, so you can say whether the problem is the tooling, the material or the program.
Two skills compound on top of that. One is programming and editing at the control, so you can adjust a job instead of waiting for someone else. The other is maintenance literacy, including reading the output of monitoring systems and knowing when a predicted fault is real. Both move you from tending a machine to running a cell.
If you want to see where nearby work sits, compare this job with multiple machine tool setters, operators, and tenders, forging machine setters, operators, and tenders, and molding, coremaking, and casting machine setters, operators, and tenders. The wider picture is on our metal and plastic workers family page and the manufacturing sector page. You can also put two jobs side by side, or check the list of jobs expected to shrink before you plan a move.