Are driverless boats the future?

We've all heard about driverless cars but what about driverless ships?
26 April 2016

Interview with 

Oskar Levander, Rolls Royce, Jonne Poikonen, University of Turku, Mika Hyvönen, Tampere University of Technology


As Rose mentioned, one of the big problems with shipping currently is the welfare Ship Wheelof the workers on board some of these ships. To deal with this we could of course improve legislation, but a more radical move would be to take people off ships altogether, which would also save money. The idea of driverless ships might sound far fetched, but it's something that the UK-based multinational specialist engineering giants Rolls-Royce are actively developing. Connie Orbach went to meet the team in Helsinki, Finland, where the work is taking place...

Connie - That is the sound of ferry pulling into Helsinki harbour and, of course, there's staff driving and operating the ship - but what if there weren't...

Let us take you on a journey to this same port 30 years from now where huge unmanned ships come to dock...

Oskar - First of all we can imagine that, especially the cargo vessels, they will be fully unmanned.

Connie - That's Oskar Levander leading the charge into the future as Vice-President of Innovation in Rolls Royce's Marine Division..

Oskar - Operating mainly autonomously also with a bit of remote control.  Especially in port, they will be supervised by some shore based central and otherwise when they go out to sea they will be driving by themselves or automatically.

Connie - These ships then, the fully automated ones, would look a bit different.  There would be none of the infrastructure that is put in for people. No mess house, kitchen, sewage works or aircon. Just a solid body for carrying cargo. But onshore things will change too, as some operation of these ships would be remote from a control centre.

Oskar - You could have a control centre here or, actually, the control centre might be on the other side of the world. You never know - that is the beauty. If it was night time here you might want to operate the ship from another time zone where it's daytime, so that nobody has to work during night times.

Connie - Why do we need to do this Oskar? We've got some lovely boats out here, they all seem to be going on alright. Why do we need to make this change?

Oskar - The main driver for this is really lies in the economic.  This is all about making shipping more efficient and, at the same time, safer. If we go towards unmanned ships, the cost of transporting 1 ton of cargo 1, nautical mile will drop more than 20%, just by going unmanned.

Connie - Well that all sounds pretty tempting, but how feasible is this and what do we need to do to get there?

A ship sets sail for America but there's no-one on board. So how does it find it's way without any eyes?  Well, there's a technological equivalent as leader of Technology Research, Jonne Poikonen from the University of Turku explains...

Jonne - Basically we need the technology to perceive the surroundings of the ship.  We'd need different kinds of sensors to guarantee that we can operate in all different kinds of weather conditions and even kinds of situations.  Because, for example, cameras don't work in the dark, they don't work in bad weather conditions and they don't give distance to targets, whereas radar give you distance that works in any condition but you cannot really recognise the optics from the radar data so you have to combine both.  The best properties of different types of sensors.

Connie - With its huge number of sensors combined to make one almighty super eyeball, my boat can now see everything.  But there's something ahead - what is it and how does our ship know what to do? Divert, go on forward, what? Here's co-lead, Tampa University of Technology's  Mika Hyvonen...

Mika - We first, of course, need to classify what they are and we need to recognise where they are based on the maritime rules, so how we need to react.

Connie - The data is all collected in and based on many past experiences and thousands and thousands of photos, an algorithm matches it to buoy or whale...

Mika - How we need to react then defines the method we select and those methods are already available. Those algorithms, of course, they are in aviation and autonomous cars, and when the situation is defined you have to react like this and then you need to ask permission from the ship intelligence.

Connie - OK. That sounded fairly straightforward - just follow the algorithms. And loads of this stuff has already been developed for cars, so it's just a case of repurposing. But then again, a little car on the road seems like a bit of a different beast to a massive great container ship in the middle of the ocean. Jonne again...

Jonne - The most crucial thing is reliability and reliability under every weather condition and every situation. This is something that hasn't been really demonstrated in the automobile side. On the sea it's even worse because the conditions are usually worse than on the highway. On the other hand, because the number of ships in any case will be much smaller than the number of cars, we can also apply the connection to remote control centre which can monitor a ship and aid it if it cannot perform by itself. I can see that this is something that cannot really be done in cars. You cannot have a control centre for million and millions of cars but you can have that for ships. So, in some cases, it might be easier to implement this, in the short term at least, in the marine side than for cars.

Connie - That's the technology side of an automated ship sorted then. There's still a lot of work but it's mainly finetuning. Back to Oskar - I can imagine there are a few other things we're going to need to consider...

Oskar - We also need to have a legal frameworks.  So, if you want to make these ships reality, we need to define the rules, so to speak, that apply to them.  And the thing is, with today's rules, they are not always written in a way that sees the possibility of unmanned operations. So we need to clarify the regulation and the law and also liability. The normal marine liability for the ships are really more shifted towards  a product liability case.

Connie - A similar thing needs to be overcome for automated cars as wel.

Oskar - Exactly. Similar problems or topic but, of course, the challenges are a little bit different.

Connie - And, I guess, when people think about this they're going to be slightly worried about the safety as well. I mean it sounds a bit scary having all these giant automated ships running around the world with no-one to control them.

Oskar - I would actually say the opposite. These ships will be safer than today's ships and, actually, bringing better technology and creating a better situation of awareness and adding some automation on top of that, will improve the safety of vessels. We need to remember that today most marine accidents are related to human error. That would between 75 and 96% depending a little bit on which study you look at. And a big part of these human errors are basically due to fatigue or just the crew not concentrating.  So we can, basically, improve the safety because we need to remember a machine does not get tired - that's the beauty of it.


Add a comment