Barely a day goes by now without news of some initiative around autonomous or unmanned ships. That they are intriguing possibilities is not in doubt but how many of the serious questions around their use have been asked let alone answered? One of the main benefits of unmanned vessels that has been put forward is that they will eliminate the possibility of human error and so make shipping safer. After all aren’t more than 75% of all accidents down to human error? That may well be a true statistic as regards actual incidents that have been investigated but how many more were avoided by the presence of a human somewhere? Spotting a floating hazard in a confused sea, smelling the distinctive scent of an electric short circuit or hot oil, realising that an area of deck was warmer than it should be, noting a crack opening across a deck, seeing that distinctive orange of a lifejacket in the water. All things that are not uncommon at sea but which can prevent an incident occurring or save lives. Then there is the simple matter of maintenance. It may be easy to navigate a ship from the shore but half the crew on a typical ship are engineering staff concerned with maintaining the fabric of the ship. They are there 365 days a year but if they are replaced by shore gangs when the vessel is in port how many times will the ship be delayed from sailing because an essential repair is not completed? Bad enough in a non-tidal port but potentially even worse if the ship gets neaped for a week. And what about the autonomous ship’s interaction with other vessels? Fine and dandy when things are as they should be but leaving aside the possibility of a major power outage, what happens when something as simple as a navigation lamp bulb blowing occurs?