■ SAFE RISK ■ Construction & Trades
No. A trade with roughly a handful of working squeri in Venice and a boat that is deliberately asymmetric is not an automation target — it is a survival-of-apprenticeship problem.
“Each gondola is hand-carved from 280 pieces of 8 types of wood. AI isn't touching this.”
Our AI replacement risk score — how we score jobs
Work in a squero is boatbuilding as slow craft. A gondola is built from oak, larch, fir, cherry, walnut, elm, mahogany and lime, each chosen for a specific structural job, and assembled without a full set of plans — the hull shape lives in templates, in the eye, and in the master's hands. It is asymmetric by design, curved so that a single oarsman on the right side drives it straight, and it is fitted to the weight and stroke of the gondolier who commissioned it. Beyond new builds, most of the year is maintenance: hauling out, scraping, recaulking, repainting the seven coats of black, replacing rotted frames and fitting the forcola and ferro.
Automation has touched the edges. CNC routers and laser cutters can produce fittings, and some yards use digital scanning to record hull geometry, which is genuinely useful for training and for reproducing a known good shape. Fibreglass and composite hulls exist and are cheaper — that is the actual commercial threat, and it is decades old, not new. Software handles the ordering, the tourism-facing side and the licence paperwork.
The core resists for unglamorous reasons. The volume is tiny, so no one will amortise a robotic cell against a few dozen boats a year. The wood is variable and hand-selected, the joinery is adjusted in place, and the finishing steps are judged by feel. The customer is buying provenance as much as a boat, which means a machine-built gondola would be worth less, not more. The genuine risk here is demographic: too few apprentices willing to spend years on low pay to learn a trade with a shrinking master population.
Automatability: our editorial assessment of current and near-term AI capability
Nothing automates this trade in any relevant timeframe — the batch size alone rules it out, and buyers pay specifically for hand construction. The real clock is generational: if apprentice numbers keep falling, the skill thins by the 2040s regardless of technology. Our risk score of 16 measures automation exposure, not the very real risk that the craft simply runs out of successors.
Parts of one, yes — fittings and some components are already cut by CNC. A full hull is another matter: the boat is asymmetric, built to a specific oarsman, assembled from hand-selected variable timber and adjusted during construction rather than from a complete drawing set. With only a few dozen boats built a year, nobody will ever pay to automate it.
Endangered, but from demography rather than technology. The number of working squeri in Venice is small and the master craftsmen are ageing, while the apprenticeship is long and poorly paid relative to the alternatives. Demand for boats and for restoration remains, so the constraint is people entering the trade, not customers leaving it.
Digital scanning and photogrammetry to record hull shapes and templates, CNC for metal fittings and repetitive trim, and straightforward business software for commissions and maintenance records. None of it changes how a hull is built; it protects institutional knowledge and frees hours that currently go into repetitive cutting and admin.
Considerably. Composite hulls have existed for decades and are cheaper to buy and maintain, and they are the reason to worry about the traditional build. What protects the wooden gondola is regulation, tradition and the fact that tourists and gondoliers alike treat the hand-built boat as the real thing — an authenticity premium a moulded hull cannot claim.