For about a decade the yachting industry has talked about robots the way it talks about hydrogen — always arriving, never quite here. The record says something more specific. Robots did arrive, just not where the press releases pointed. They live on lines that make the same part over and over, and they stop at the door of the shed where a sixty-metre custom yacht is being built one of one. What separates the two is repetition.
Where the robots already stand
The most robotised boatbuilder in Italy is not a superyacht name. Cranchi runs two overhead Kawasaki robot cells: one at Piantedo spraying resin, catalyst and chopped roving onto parts, another at the Rogolo plant cutting and spraying laminate and topcoat onto parts as large as 24 by 7.1 metres. That last figure is the size of the part, not the length of the boat — a distinction the trade press keeps losing.
Azimut got there first on gelcoat. Its Avigliana plant has run three production robots since the middle of the last decade: one that cuts glass fabric and hands over labelled kits, one riding a crane gantry that sprays gelcoat mostly on night shifts, and one in a closed cell that trims and drills, finding the part in space from reference points an operator sets. Group president Giovanna Vitelli frames the whole programme in one word, and it is not labour: "Automating means quality, because it guarantees repeatability."
Outside Italy the pattern holds wherever volume does. At Yamaha's Vonore plant in Tennessee, trimming the deck of an eighteen-footer took four people forty-five minutes; two robot arms do it in thirteen. Sea Ray has been cutting and trimming hulls with robots for over two decades. Mercury Marine runs twenty-five robots in machining alone and grinds ninety-five per cent of its propellers robotically. Hanse fills a building in Greifswald with CNC robots, including a five-axis machine that shapes complex wooden forms and runs unattended at night, switching timber species on a schedule. Riviera lets a robot lay up to eight coats of lacquer on doors and cabinet fronts. Grand Banks mills foam core on an eight-axis KUKA cell sixteen hours a day and gets a month of work done in days.
| Builder | What is automated | Result |
|---|---|---|
| Cranchi | Laminate and topcoat spray, Kawasaki cells | Parts up to 24 × 7.1 m handled robotically |
| Azimut (Avigliana) | Fabric cutting, gelcoat spray, trim and drill | Gelcoat runs on night shifts, unmanned |
| Yamaha (Vonore) | Deck trimming | 45 min × 4 people → 13 min × 2 robots |
| Sea Ray (Tellico) | Hull cutting and trimming | 3–4 people × 1 hour → under 45 min |
| Mercury Marine | Machining, propeller grinding | 95% of props robot-ground, hand-finished |
| Hanse (Greifswald) | Five-axis joinery, unattended night shifts | Line capable of a finished boat every 4 hours |
| Riviera (Coomera) | Lacquer application | Up to 8 coats, hand sanding between each |
| Grand Banks (Johor) | Eight-axis KUKA core milling | 16+ hours a day, a month of work in days |
What a robot costs and what it buys back
Yamaha paid one to one and a half million dollars per robot. The cell only made sense at about nine boats a day, earned its second wave at twelve, and the plant now runs twenty-eight. Below that line the machine idles, and an idle machine is a loan payment.
The clearest money in the industry sits in fairing. Filling and fairing a superyacht hull by hand runs twenty to twenty-five man-hours per metre of length and can take a year or more, at a cost of €800,000 to €1.5 million per hull. A Robosurf installation costs upwards of a million euros and covers up to ten square metres a minute once a surface map is loaded; scanning the hull of a seventy-metre yacht in Ancona took less than a day. The vendor puts payback inside one build, two at most — its number, not an independent measurement. Collaborative arms, the light ones a person can work beside and program from a phone, sit in the low tens of thousands instead.
Printing the part instead of the mould
The sharpest single case belongs to Ferretti. The side air intake grille of the Pershing GTX116 — 4.2 metres long, 40 kilos, glass-reinforced ASA — came off a Caracol robotic extrusion platform in 72 hours. Against the conventional route it is fifteen per cent lighter, halves the lead time and cuts waste by sixty per cent. Most of that comes from what never had to be built: a grille like that normally needs a mould, and the mould needs weeks of hand work before a single part exists.
Steel is the only place headcount actually falls
Where the work is welding, the numbers move people. Marketex Marine in Tallinn took double-bottom production from twelve welders to two operators on a twin-robot portal. Fincantieri built MR4Weld with Comau — a six-axis arm on tracks that finds the joint by vision, works without fencing and needs one operator — and claims up to three times manual throughput. Damen put a PEMA robotic micro-panel line into its Galati yard in Romania, and its managing director said plainly why: "We need more welders than there are currently available."
The half of the story nobody puts in a press release
Beneteau built the sanding robot and then refused to install it. Inside COROMA, a European research project with sixteen partners and a €7.26 million budget, a Stäubli arm on a mobile platform demonstrated that a cobot could handle up to eighty per cent of hull sanding. The group's innovation director, Bruno Merchie, explained the decision without ceremony: the return on investment would not have been obvious. Beneteau also said why in production terms — on large models the cycles run so long that keeping a robot continuously loaded is the hard part.
The machine worked. The business case did not.
Full automation has already been tried at the extreme and it failed on economics, not physics. Genmar's VEC line promised a hull in an hour and twenty thousand boats a year. The company went under in 2009. Rob Parmentier, who ran boats built on the process, was blunt afterwards: it was a great process making great boats and it cost about ten per cent more than open moulding, so when volume fell it was doomed. Larson gave up on VEC in 2017 and went back to open moulds.
The Dutch state is paying into the same conclusion. STREAM, a €4.5 million project with €2 million of government money, exists because standard welding robots are close to unusable on Dutch yards: project variation is so high that programming only pays off in series. Kranendonk, the country's leading shipbuilding robot supplier, lists Royal IHC, NASSCO, Austal USA and offshore fabricators among its references — and not one yacht yard.
The biggest Dutch builder answers the same labour question a different way. Oceanco puts the construction of a yacht at seventy to eighty per cent human labour, and its response has been a network of co-maker subcontractors rather than manipulators. Royal Huisman bought Ottobock exoskeletons for hot work — automating the human instead of replacing him. Baglietto caps itself deliberately at six or seven deliveries a year and 65 metres to preserve the craft. Sanlorenzo and The Italian Sea Group have spent heavily on capacity without a single public statement about production robotics.
The most honest voice belongs to a yard that did buy one. Viking installed its first custom robotic sanding station in July 2025, working with Viam, and reported back in October that it does not reach Viking's finish quality without a person: think of them as tools first and robots second, and do not expect a technology you can walk away from. It handles rough, high-removal passes. Final fairing stays in human hands.
Volume is the gate on automation in boatbuilding. Where the same part comes down the line hundreds of times a year the robot is already there and has been for twenty years. Where each hull is a prototype, the arithmetic collapses — and the yards that ran the numbers properly, Beneteau among them, said no to machines that provably worked.
The push comes from demography
industriAll Europe puts forty per cent of the European shipbuilding workforce into retirement by 2030. Confindustria and BCG estimate Italy's sea economy needs roughly 175,000 workers between 2026 and 2030, split between growth and replacement, with about forty-three per cent of shipbuilding vacancies going unfilled. Welders top every hard-to-hire list; painters, laminators and joiners follow.
The wording splits by venue, though. On technology stages the industry talks about quality and repeatability; on labour stages it talks about shortage. No yacht yard chief has said on the record that robots were bought because the welders could not be found — the closest are a shipbuilder, a Romanian steel yard and a robot vendor. Treat the link as our reading of the evidence rather than an industry position.
Outside steel welding, the headcount holds. Mercury says robots added people — programmers, quality control — and pulled the average age of the department down. The work moves rather than disappears: whoever used to drill the holes now checks the machine that drills them.
What it means if you are buying the boat
A robot-finished production boat buys you sameness in the good sense. Gelcoat thickness, lacquer film, drilled positions and trimmed edges land within a tolerance a tired human cannot hold at the end of a shift, and hull number four hundred matches hull number four.
A custom yard sells the exact opposite, and sells it as a strength. Baglietto restricting itself to six or seven hulls a year, Oceanco calling three-quarters of the build human labour — that is the product. When a sixty-metre yard tells you every one of its yachts is different, it is also telling you why there is no robot in the shed. Even at the automated end nobody let go of the hands: Riviera sands between all eight robotic coats, Mercury finishes propellers manually, Hanse still lays up its glass by people. The robot took the dull, dusty and toxic passes. The judgement calls stayed where they were.
Photo: Robosurf.
