■ HIGH RISK ■ Manufacturing & Production
For the repetitive line role, yes — this is automation's oldest and most successful project, and industrial robots plus machine vision keep widening what counts as repetitive. Humans remain in factories, but increasingly as technicians and exception-handlers, not as hands on the line.
“The factory of the future has two employees: a human and a dog. The human feeds the dog. The dog keeps the human away from the machines.”
Our AI replacement risk score — how we score jobs
'Factory worker' covers a lot of floor: machine tending, assembly line stations, packing and palletizing, material handling, quality checks, and the thousand small tasks between them. The common thread is standardized work in a controlled environment — the same part, the same motion, the same takt time, shift after shift. That standardization is the whole point of a factory, and it's also a precise description of what robots are for.
The displacement is generations deep and still accelerating. Industrial robots own welding, painting, and rigid assembly; collaborative robots now work unfenced beside humans on lighter tasks; machine vision inspects faster and more consistently than human eyes; and autonomous mobile robots ferry materials that forklift drivers and line feeders once moved. Global robot installations keep setting records, with density highest in exactly the industries — automotive, electronics — that employ the most line workers. The newest frontier is dexterity: AI-trained robots handling deformable materials, mixed bins, and variable products, which erodes the last technical excuse for many human stations. Labor shortages since 2020 flipped the politics too; plants automate now because they can't hire, not just to cut costs.
Yet factories still employ millions of humans, and the reason is instructive: exceptions, changeovers, maintenance, and judgment. Short production runs make robot programming uneconomical; new product launches need flexible hands; machines break and jam; and quality problems require someone who can ask why. The factory jobs growing — robot technician, maintenance mechanic, process engineer, quality analyst — all supervise or repair automation rather than compete with it. Our 75 risk score is really a fork in the road: the repetitive-motion factory job is going, while the factory-technician job it funds is hiring.
Automatability: our editorial assessment of current and near-term AI capability
This is the longest-running automation story in economics, and the current chapter is brisk: record robot installations, AI-driven dexterity gains, and post-2020 labor shortages have pushed adoption into mid-size plants that once couldn't justify it. Through this decade, expect repetitive line roles to keep converting to machine-tending and technician roles, with the pure manual station becoming rare in well-capitalized plants by the early 2030s.
Fully 'lights-out' factories exist but remain rare, because exceptions are stubborn: changeovers, jams, maintenance, new products, and quality mysteries all still need humans. The realistic trajectory isn't zero workers — it's far fewer workers per unit of output, with the remaining roles skewed toward technicians, maintenance, and engineering. A plant that employed a thousand line workers may employ two hundred people, almost none doing repetitive manual tasks.
The ones that fix, adjust, and improve the robots: industrial maintenance technicians, controls and PLC techs, tool and die makers, process and quality engineers, and skilled trades like industrial electricians. Changeover-heavy and short-run production roles also persist longer than high-volume repetitive stations. The pattern is simple — jobs defined by variability survive; jobs defined by repetition don't.
Maintenance and automation technician paths are the highest-leverage moves, and many plants will pay for the training because they're desperate for techs. Start with whatever your employer offers: PLC or robotics courses, apprenticeships, quality certifications. If your plant offers nothing, community colleges run industrial maintenance programs, often part-time. The goal is to be the person the robot calls for help, on the payroll where the robots are arriving.
The opposite — they're the accelerant. Plants that can't fill shifts buy robots out of necessity, not strategy, and vendors now market automation explicitly as a staffing solution. This cuts both ways for workers: today's shortage gives you leverage and training opportunities, but it's also financing the equipment that shrinks tomorrow's headcount. Use the leverage window to reskill on the company's dime.