I picture a small-ish shipping ship, say something moving 20 or 30 containers only, fully autonomous (100% crewless) with sails and solar only powering it. Solar electric motors handling the near-land operations and "not enough wind" conditions, sails otherwise. If you had autonomous loading and unloading, you would have zero variable/operational cost.
I am certain there are very good reasons this would never work, but I still love the idea.
Crews are cheap. Ships' masters are not.
Part of me starts to wonder now about building boats that are more resilient, but then I remember that salt water is terrible and surely by now someone would have if it were possible, since we spend so much money on humans doing that maintenance!
The number can rise further with more complex cargo.
I think it will do to shipping what two engine, midsized passenger jets (A350/787) are doing to air travel: connect directly between small ports anywhere, instead of using very large ports as hubs. FedEx/UPS could have sea terminals anywhere, possibly automated.
Global shipping will only grow and I don't see a driver for air freight prices to drop enough so that things are truly international (e.g. buy a fridge from AliExpress).
I think it needs to happen, and will happen.
As for very good reasons why this would never work, sometimes that is beside the point. I am not in the running to redesign trade and power on my own terms, but sometimes I can flush out assumptions that don't need to be held.
Im imagining an LED throwie type big magnetic thing with gps RX and TX, and a satellite uplink. Cruise by the autonomous ship on a jetski near a port, thrown the magnet thing on, jetski away, disrupt from afar once the boat is out at sea.