■ HIGH RISK ■ Transportation & Logistics
Autonomous and remotely-operated tugs are real, funded, and already doing demonstration jobs in major ports. Captains won't vanish this decade — regulation and liability move at anchor-dragging speed — but crew sizes will shrink and remote operation centers will start absorbing the wheelhouse.
“Autonomous tugs don't need a harbor pilot's license.”
Our AI replacement risk score — how we score jobs
Tug work is precision violence: putting thousands of horsepower against a container ship's hull in wind, current, and darkness, inches from disaster, on the radioed instructions of a harbor pilot. A captain's day mixes ship-assist jobs, barge tows, line handling coordination with deckhands, engine and winch monitoring, and long stretches of waiting punctuated by minutes where a mistake writes off a hull or a life. The skill is feel — knowing how this ship will behave in this current before it does.
That feel is precisely what autonomy vendors are chasing, and they're further along than most mariners like to admit. Major towage operators have run remote-controlled and semi-autonomous tug demonstrations in real harbors, and navigation autonomy handles the transit legs — the boring miles between jobs — quite well already. The near-term model isn't a ghost fleet; it's reduced crewing, with one master supervising automation aboard, then shore-based remote operation centers where a single operator oversees several tugs' transits and takes the conn for the delicate parts.
The resistance is institutional as much as technical. Ship-assist work is a liability nightmare — a bad push at the wrong moment means a ship into a berth — and ports, insurers, unions, and coast guard regulators all have to sign off before an unmanned tug noses up to a loaded tanker. Close-quarters work in degraded conditions, gear failures underway, and towing jobs that go sideways still favor a human aboard. Our 58 score reflects a trade where the technology is arriving ahead of the permission, and the permission is coming.
Automatability: our editorial assessment of current and near-term AI capability
Remote and semi-autonomous tug trials are happening in commercial ports now, and reduced crewing is the immediate battleground. Expect shore-based remote operation of transit legs to become normal at forward-leaning ports by around 2030, with full unmanned ship-assist trailing years behind because regulators and insurers must bless every step. Captains nearing retirement are fine; captains starting out should plan for a hybrid career.
In trials, yes. Several major towage companies have demonstrated remotely operated and semi-autonomous tugs in real harbors, mostly on transit legs rather than live ship-assist work. Commercial routine use is still limited by regulation, insurance, and port rules — but the demonstrations prove the technology works well enough to keep investing in.
Full replacement is not a this-decade event; ship-assist is too high-consequence for regulators to unman quickly. The realistic path is crew reduction and shore-based remote operation of routine legs emerging by around 2030, with experienced masters moving into supervisory and remote roles. Captains early in their careers should expect the job to change shape mid-career.
With eyes open, yes. Mariner shortages are real, pay is strong, and the transition will take years. But choose training that transfers — remote operations, systems management, pilotage knowledge — rather than betting your whole career on a wheelhouse chair existing in 2045. Our 58 risk score says pressured, not doomed.
The messy parts: a parted towline in a seaway, an engine failure mid-push, fog plus a confused recreational boater in the channel, and any situation requiring physical intervention on deck. Autonomy handles clean scenarios; tugs earn their money in dirty ones. That's why a human stays in the loop for the foreseeable transition.