■ MODERATE RISK ■ Personal Services
No — the entire point of a numbers-matching restoration is that obsessive humans did it by hand. AI will find your parts and match your paint; it won't be trusted to weld a quarter panel on a seven-figure Ferrari.
“3D scanning finds replacement parts. But rebuilding a '67 Mustang is a labor of love.”
Our AI replacement risk score — how we score jobs
Restoration work is archaeology with a torque wrench. A shop day might involve stripping fifty years of paint and body filler to find out what the seller lied about, hunting a discontinued trim piece across forums and swap meets, machining a part that hasn't been manufactured since the Johnson administration, and documenting every step for a customer who will absolutely notice a wrong-year hose clamp at a concours judging. Projects run months to years, and the business side — quoting the unquotable, managing customer expectations as rust bills triple — is half the job.
Technology has quietly revolutionized the parts problem, which was always the worst part. 3D scanning and printing recreate unobtainable trim, brackets, and even complex castings; AI image search identifies mystery components from a photo; spectrophotometers and formulation software match sixty-year-old paint chemistry. Diagnostic AI helps with the growing niche of 'youngtimer' classics that have early ECUs. Expect research, sourcing, documentation, and quoting to keep compressing — the knowledge-hoarding that let some restorers charge for what they knew rather than what they did is evaporating.
The hands stay employed. Metal-shaping a compound-curve fender, laying period-correct lacquer, rebuilding a carburetor by feel, and knowing that a '67 block should carry this casting number and that shade of engine enamel — this is judgment plus craft plus provenance, and the collector market pays specifically for human authorship. Meanwhile the customer base is a genuine long-term question: the generation that lusts after tri-five Chevys is aging out, and future collectors may fetishize different machines. That demographic drift, more than AI, is what our 35 is pricing.
Automatability: our editorial assessment of current and near-term AI capability
The tooling shift is underway now — 3D scanning, printed parts, and AI-assisted sourcing are already standard in serious shops — and will keep compressing research and fabrication time through the 2030s. The craft core faces no credible automation on that horizon. Watch the customer demographics instead: as the muscle-car generation ages out, demand will migrate toward newer classics, favoring restorers who can handle 1980s-2000s electronics alongside carburetors.
The trade is healthier than its demographics. High-end restoration shops have waitlists, and technology has made the worst part — parts hunting — dramatically easier. The real long-term question is the customer base: buyers who covet 1960s muscle are aging, and the next collector wave wants 80s and 90s machines. Restorers who follow the market rather than their own nostalgia will stay busy for decades.
Mostly by killing the scavenger hunt. AI image search identifies mystery parts, marketplaces and databases surface rare components in hours instead of months, 3D scanning and printing recreate what can't be found, and software matches decades-old paint formulas. The welding, metal-shaping, and assembly remain entirely human — which is exactly what concours judges and collectors are paying for.
Not remotely. Restoration is the opposite of manufacturing: every car arrives as a unique puzzle of rust, bad prior repairs, and missing pieces, and the work is one-off problem solving on irregular, fragile material. Robots excel at repeating a known task a million times. There is no near-term path to automating a job where the task is different every single hour.
A 3D scanner before another English wheel. Digital fabrication for unobtainable parts, formulation software for paint, and AI-assisted sourcing are where hours and margins are won now. Add electronics competence for younger classics, and document everything on camera — provenance content markets your shop and adds real value to the finished car.