Why upholstery work stays with people
Upholstery is a one-off job on a one-off object. Every chair arrives with its own frame, its own damage and its own history. Before anything else, an upholsterer strips the old covering, webbing and padding off the frame and reads what is underneath: a cracked rail, a loose joint, springs that have lost their tension. Software cannot see that from a photo, and no machine today can take a sofa apart by feel.
The build side is just as physical. Tying and installing cushion springs, stretching fabric over curves, and tacking or stapling material tight without puckers are hand skills learned over years. Small corrections happen constantly, by touch. That is why people keep most of the task time on this job: 85%.
There is a customer side too. Upholsterers discuss fabrics, colors and styles with clients, measure the piece, and estimate what the work will cost. AI tools now help with parts of that conversation, which is the real story here — tasks shifting, not the trade disappearing. The question of whether AI will replace upholsterers is answered mostly by how much of the job happens with hands on a frame.
What AI does, what it assists, and what stays manual
The slice AI can run on its own is small: 0%. It lands on paperwork and visuals. Drafting a quote from measurements, keeping job records, and generating a mock-up of a fabric on a chair are the kinds of tasks software can finish without a person steering every step.
A similar slice is assisted rather than automated: 15%. Fabric visualizers let a customer see a pattern on their own couch before ordering yards of material, which cuts guesswork on color and scale. Yardage math, pattern-repeat planning and cut lists are also faster with software, though the upholsterer still checks the numbers against the actual piece.
Everything else is manual. Removing old covering and padding, repairing or replacing frames and springs, cutting and fitting new material, sewing rips and tears, and building custom pieces from a sketch all need a person in the room. Coverage, our measure of the share of task time AI can handle today, reads 9 out of 100 for this job; you can read how that figure is built on the coverage method page.
What the evidence shows
No one has tested a machine against a working upholsterer on a real piece of furniture. That is why the quality-parity evidence grade here is D, and why this page gives no parity number. A grade like that means not measured, not measured and failed.
What would settle it is specific: a trial where a robot or an AI system strips a used chair, assesses the frame, and fits new covering to a finish a customer would accept, scored next to a qualified upholsterer on time, fit and durability. Factory sewing cells and automated cutting tables exist, but they work on identical parts in a known position, which is the opposite of reupholstery. Our estimate of the physical share of this job runs high, and the robotics tier it would take is a dexterous humanoid — hardware that is still in testing, not in workshops. The quality-parity method explains how we grade this kind of gap.
The labor data points the same way without AI entering the picture. The Bureau of Labor Statistics counts about 20,140 upholsterers in the US with median pay of $46,340, and projects employment down 2.3% from 2025 to 2035 (BLS, 2025). That drift comes mostly from cheap replacement furniture, not from software.
When the picture could change
Most likely after 2046 (8 in 10 of our scenarios). The reasoning behind a window that wide is set out on the replacement-year method page.
Two things could pull it earlier. General-purpose robot arms with reliable touch sensing would open up repetitive steps like stapling and trimming on standard shapes. And if more work shifted to factory-made modular furniture with standard frames and zip-on covers, the variable part of the job would shrink.
Two things hold it back. The frames are all different, so a machine has no fixed starting point, and fabric behaves unpredictably under tension. Cost is the other brake: our estimates put a capable robotic setup well above what it costs to employ a person for the same work, which is a hard sell for a shop of two or three people.
How upholsterers stay needed
Lean into the tasks that stay manual. Frame and spring repair is the clearest one, because it turns a worn piece into a sound one and no visualizer can do it. Custom work from a client’s sketch is the second: odd shapes, antiques, boat and vehicle interiors. Third is fitting and finishing patterned fabric, where matching repeats across a curve separates a good job from an obvious one.
Two skills to add. First, estimating and client consultation, using fabric visualizers to help people decide quickly rather than letting a cheap sofa win by default. Second, restoration knowledge: period construction, horsehair and traditional stuffing, correct tacking. That work is priced on judgment, not on speed.
What to do: photograph before-and-after on frame and spring repairs, not just finished covers, so customers can see what they are paying for.
Nearby jobs worth comparing are tailors, dressmakers, and custom sewers, hand sewers and furniture finishers. You can put any two of them side by side on the compare tool, or see the wider group on the textile, apparel, and furnishings workers family page and the manufacturing sector page. This job’s Still needs a human score is 84 out of 100 (higher is safer); the trades it sits near show up on our list of jobs that mostly need a person, and the full scoring approach is on the methodology page.