In response to Amazon Enters Trillion Dollar Ocean Freight Business: How Many Jobs Will Vanish? a reader commented fully autonomous ships will not be here until 2050. Actually, 2020-2025 seems like a realistic timeline.
IEEE Spectrum says, Forget Autonomous Cars—Autonomous Ships Are Almost Here.
My colleagues and I at Rolls-Royce anticipate that the first commercial vessel to navigate entirely by itself could be a harbor tug or a ferry designed to carry cars the short distance across the mouth of a river or a fjord and that it or similar ships will be in commercial operation within the next few years. And we expect fully autonomous oceangoing cargo ships to be routinely plying the world’s seas in 10 or 15 years’ time.
That people should be seriously interested in robotic ships is easy enough to explain: Such ships are expected to be safer, more efficient, and cheaper to run. According to a report published by the Munich-based insurance company Allianz in 2012, between 75 and 96 percent of marine accidents are a result of human error, often a result of fatigue. Remotely controlled and autonomous ships would reduce the risk of such mistakes and along with it the risk of injury and even death to crew members, not to mention the dangers to the ship itself.
The threat posed by piracy to ships and their crews would also be reduced. That’s because uncrewed ships could be built so that they’d be very difficult to board on the high seas. Even if pirates got aboard, access to the controls could be made unavailable. Indeed, the computers in command could immobilize the ship or have it steam in a circle, making it relatively easy for naval authorities to reach it. Recapture would also be easier than is usually the case in such situations because there would be no crew held hostage. And without a captured crew to ransom, the target of the piracy is significantly less valuable.
Another advantage of remotely controlled and autonomous ships is that they can be designed with a larger cargo capacity and lower wind resistance. That’s because, with no crew to accommodate, certain features of today’s ships can be eliminated—for example, the deck house, the crew quarters, and elements of the ventilation, heating, and sewage systems. This will make the ship lighter and sleeker, cutting fuel consumption, reducing operating and construction costs, and facilitating designs with more space for cargo.
Finally, intelligent ships will provide owners and operators with a way to respond to the growing shortage of people who have the requisite maritime skills. With more and more mechanical and electronic systems on board, ships are becoming increasingly complex, needing skilled technicians to keep them working. At the same time, seafaring as a career is growing less attractive, with fewer people from developed nations wanting to spend weeks or months at a time away from home and family. Remote and autonomous operations could facilitate the transfer of jobs requiring high levels of education and skills to ports of call or to operations centers on land, making such careers more interesting to young people entering the industry.
I can’t point to examples in the water, but Rolls-Royce is working now on the specifications and on preliminary designs for the first generation of advanced intelligent ships. My colleagues and I are bringing together researchers at universities, ship designers, equipment manufacturers, and classification societies to explore the economic, social, legal, regulatory, and technological factors that need to be addressed.
The first intelligent ship to go into commercial operation will use mostly technology that already exists. That vessel will likely ply the coastal waters of a single “flag state,” a seafarer’s term for a country that can provide the legal basis for a ship’s operation. The ship could be a ferry, a tug, or other coastal vessel traveling within a very confined area. It could still have a crew on board, although they will be carrying out duties other than navigating the vessel.
Indeed, the testing of such ships is not far off.
Expect such a ship by 2020. By 2025, some forward-thinking shipping companies will be operating remotely controlled, completely uncrewed vessels on the high seas. Five years beyond that, uncrewed oceangoing vessels will be commonplace.
About the Author
Oskar Levander trained as a naval architect at the Helsinki University of Technology. After gaining experience in various other posts, he joined Rolls-Royce in 2012 as vice president of Innovation, Engineering, and Technology, Marine.
Debate Over
Fully autonomous ships are coming sooner than most suspect. They will be faster, more fuel efficient, and cost less to insure.
Accident rates will drop as will odds of piracy. In regards to the timeframe, I suspect Levander is purposely cautious. There are too many benefits for this to take as long as 2030 if such a ship will be operating by 2020 as he states.
By 2025 it will not be just “forward-thinking” shippers using such technology. By 2025 it will be commonpalce, and by 2030 it will be nearly all of them.
Possible Designs
What About Jobs?
In my Amazon article, I asked but did not answer the question: How Many Jobs Will Vanish?
In the grand scheme of things, especially from a US perspective, not that many. Crew sizes have been shrinking for some time, and most crews are not US-based anyway.
Thus, the IEEE Spectrum headline beginning with “Forget Autonomous Cars” is incorrect. In contrast to shipping, millions of truck driving jobs will vanish, in the US alone. Moreover, trucking adoption will be nearly instantaneous.
The Rolls-Royce article puts to shame widely-held notions regarding piracy. Similar constructs apply to truck hijacking, as I have been saying for years.
As with shipping, insurance rates will plunge for truckers. The biggest cause of truck driving accidents is human error, typically fatigue.
And instead of having to take mandated rests, autonomous trucks can drive nonstop for days. The cost of the driver vanishes.
Hugely disruptive changes are coming, quickly, and for obvious reasons. The nay-sayers look increasingly foolish with their denials.
Mike “Mish” Shedlock
Well I guess I have to remain looking foolish.
The author conflates navigational autonomy with total autonomy. For the past two decades a ship that navigates itself has been possible, maybe further refinement will mean the captain remains almost totally supervisory. If you want fully autonomous navigation / shipping you will have to have contingency crews stationed at hand… maybe they could follow behind in a super efficient luxury yacht with self aiming catapult technology or robocopters.
Piracy, you forget the ransom is the cargo, people can be kidnapped anywhere more easily. Again, I won’t post how, but there are ways to hold a cargo ship without crew to ransom.
Like disabling the AI autopilot…
Basically you have to be in a position to credibly destroy the ship with no intervention possible. That means just destroying or disabling ships in the hope the owner will pay to be left alone is not what I mean. If the owner is told, and knows, that the ship will be destroyed unless he pays then he will likely pay.
On the other hand, running ships aground just to salvage the goods on board has a long history.
To understand it you have to unmix the hostage/asset combination in a typical ransom. It is difficult to do because the hostages bring the attention but the insurer is liable for the cost of the ship. So resolution will be different, and governments will be able to ‘allow’, as in compensate, ships being destroyed without outcry at loss of life, as a policy to discourage piracy. That does not mean it will be that way though, as for example US registry is less common.
Just saying not to be so sure.
Fake “economic” News
http://www.ibtimes.co.uk/rolls-royce-aims-cut-cost-sea-travel-automated-crewless-vessels-1606147
Probably the only thing keeping it from happening by 2020 is liability issues and liability insurance. Also, pilots will still be required in port, so that has to be handled.
Mish…It’s time to dig out my puffy shirt…But I want to be a pirate…Love your blog…Jerry
“Remotely controlled and autonomous ships” caught my eye. Can you say hacked and hijacked? Just a thought. I totally agree though automation is quickly changing our world.
Crysangle is right; ships have been navigating automatically for years, most deck officers today would be hard put to navigate with a sextant! There are clearly applications for this technology but you have to understand that the ocean is a hard place. I worked on VLCC tankers as an electrician for a couple of years, 7 days a week 12 hours a day was standard, often my hours were longer, just keeping the damn thing running. Maintenance and repair of ships is continuous, the wear and tear is enormous, you have to experience it to believe it.
I try to stop myself from laughing when I read articles like the above. It isn’t that it isn’t ‘possible’ but clearly the author hasn’t much real experience of life at sea. For it to work with no crew, and depending on cargo, they would have to design very complex systems that… well I just give up there as keep it simple is an old motto, just like ships being holes in the water to pour money into. I did post links in the last article of his on how crew have been paired down to minimums already
https://robertjprince.net/2015/02/23/global-container-shipping-bigger-ships-smaller-crews-a-formula-for-disaster/
https://www.quora.com/What-was-the-typical-crew-of-a-cargo-ship-composed-of
And others on crew cost per total.
Crewless…you are going to fly someone 5000 miles and parachute them to tighten a connection… defies logic.
Yep, Mish has almost as much experience in piloting a large ship as he does in driving an 18 wheeler.
5 years afloat as USCG Deck Officer on a 378′ High Endurance Cutter to commanding my own Patrol Boat. Have sailed all over the Pacific, Bering Sea, East Coast and the Carribean. Piloting a ship is more than just keying in a couple of waypoints into a computer and then going to sleep. An experienced Captain or Master can hold bare steerage way in a hurricane or extra tropical cyclone, can navigate along or in Long Island sound using visual aids to navigation or needs to be able to utilize line of sight aids (some maintained and others improved) to navigate a channel. The US navy has recently begun teaching celestial navigation again as they are concern that Satellite aided navigation may not always be available.
What about anchoring off shore or handling a major casualty? Lots of very large complex machinery that must be maintained. More to this than Mish realizes,
I practiced celestial navigation with an Almanac and calculator, that was not easy at all, though I suppose nowadays there are applications to take care of most of the calculation. How about watch keeping? Yachts lost in wave clutter or background, that no radar will find and having to respond to flare warnings ? If an autonomous ship might be forgiven for not responding to an incident it did not cause, it is certainly responsible for responding to one it has caused. The list is endless…
It will happen and cargo will be loaded autonomously. Policing will be in place around the lanes. None of the objections are insurmountable.
They will also be hybrid diesel-electric and covered in solar panels.
With the Baltic Dry where it is operators will look to get margin back wherever they can.
It isn’t that it won’t move that direction, but a little realism.
All you are saying, that the article is saying, is possible, but it would not make sense to try to implement it as suggested. Maybe crew can ultimately be reduced to three or four people after some of the many engineering problems are made close to reliable, but that is not related to automation, in fact automation needs a seperate new set of skills beyond the traditional technical ones to be maintained on board.
So a long long way off from being anything standard even with few crew, in my opinion, but it is an interesting venture all the same.
Fully Autonomous Mish Almost Here
It would be an improvement.
What???….no more gold reco’s?….oh, that’s right….it’s done nothing but go down from $1800 for the past six years. What a money loser….but, but,,,,millions of trucking jobs!!! Ugh…who cares.
This concept is more realistic than millions of self-driving cars navigating the nation’s freeways. But you’d never find me on an trans-Atlantic cruise with a robot at the helm. No more than I would fly on a passenger jet with no human in the cockpit. My tolerance for risk just isn’t that high. Regardless of the deal offered.
Mish, you keep writing that millions of jobs will disappear soon due to automation, but you never talk about what will be next. Would you mind sharing your thoughts about that? What will happen to the millions of people who are currently employed in these industries?
They would become refugees and flee to China.
in 1800, about 90% of the labor force worked on farms
https://www.agclassroom.org/gan/timeline/farmers_land.htm
People have been pointing to situations like these for decades.
I have no idea where the jobs will come from.
Imagine stepping back to 1800 where 90% or whatever % of the population worked on farms.
Tell them only 2% will work on farms.
They would have thought you were nuts.
They might have asked “where will the people work?” and no one would have had an answer.
No one thought of railroads, airlines, banks, cars.
Is it different now?
I don’t know, but certainly the Fed, Congress, and Nixon exacerbated the problem by crony capitalism, min wages, public unions, closing the gold window, etc. etc. etc.
I could see the creation of new industries. Like Somosa deliverers for the exploding population of H1B robotic engineers. Street vendor careers would flourish too.
You are on to something there LF.
After pet psychology, maybe something like robotic interaction enhancement training would not be too far out of norm… after all, you would not want to bite the bot that feeds you.
Mostly Filipinos anyway!
I agree with the autonomous car sooner than expected but from my experience with people involved in large cargo shipping, they are always dealing with one problem after another. These are massive pieces of equipment with an enormous amount of moving parts, and unlike a car that might have a problem on Interstate 95, no more than 30 miles from assistance, a problem on one of these puppies, 1.500 miles from the nearest port, is a PROBLEM!
The problem with self-driving cars has nothing to do with mechanical breakdowns. It has to do with millions of them navigating the same treacherous roadways while trying to avoid colliding w/ other cars, bicyclists and pedestrians. Comparing the self-driving cars to isolated ships on the oceans is a bad comparison. Cars would create exponentially more problems.
But you’re correct. What happens if an autonomous fully loaded cargo ship goes limp 1500 miles from land on rough seas? I guess they haven’t thought that far in advance.
I saw it as mutually exclusive arguments.
The logistical challenge of breakdowns being one, navigational feasibility being the other.
I personally don’t see the navigational feasibility problem with either of the two. Liability of the auto one will need to be worked through the system, but I bet that laws are passed favoring corporate interests.
As in if you are on a bike and you ride in front of a car; since the auto has illustrated X number of safety factors, it’s your damn fault.
I’m cheering for the cars before I get old. For now I prefer manual transmissions.
Everything will be recorded, black box, not difficult.
When an accident occurs any judge would have access to full replay from all autonomous vehicle logs for a while before accident happened.
They could see what the vehicle saw and how it reacted.
To a degree some cars already allow that.
have you considered the roadways are treacherous because of humans?
Technological advances have been automating and replacing jobs for hundreds of years. It is an unstoppable trend. Yet we have more people employed on this planet than ever before. We continue to create more jobs than we eliminate.
No one can predict what the new jobs created will be. Nor can we say for certain that we will continue to create more jobs than we eliminate. We can only assume that the trend will continue.
I personally believe that the “winners” in the future will be those (countries, governments, individuals, etc.) that embrace the technological change and take advantage of it. The “losers” will be those that resist the change.
Stuff breaks. Stuff that floats breaks even more. Uncrewed ships won’t happen until Skynet comes on line.
As harsh as the sea may be, space is a much harsher environment. Which is why the only ships that ply those lanes, are in fact unmanned.
I realize the space guys do play with somewhat less restrictive budgets, but “breaking down” is hardly the biggest issue with any competently designed machine. Interaction with other actors, especially those that want to take advantage of it, is. Ships are much more sheltered from that, than cars are.
Ships might not be my first idea for automation. There are more obvious targets: I would start by replacing the FED with IBM Watson. Watson could beat the FED on all aspects of it’s purported policies:
FED Watson
politically independent yes(?) yes
data driven yes(?) oh, yes
revolving door yes no comprendo
collusion no(?) no comprendo
cost of maintenance expensive cheap
A simple bomb crater would soundly beat the Fed at absolutely everything of value. No Watson required.
This will be painful in the short run, but really great in the long run. Human beings will finally have time to solve big problems such as moving to a new solar system instead of wasting times doing stupid and repetitive tasks over and over again. New industries will be created to support these automated things and that will create other types of jobs. I say bring it on.
“…between 75 and 96 percent of marine accidents are a result of human error”
I find this sort of argument for automation hilarious. Sure, no doubt many accidents are due to human error, but this is because humans are the ones controlling the ships (cars, airplanes). Put computers in charge and it will simply change to “… between X and Y percent of accidents are a result of computer errors”.
The real question, of course, is can any fully-automated system actually do any better than human operators. I doubt it, because current AI lacks the flexibility to understand and handle unexpected events that so often cause accidents.
The best approach is for technology to assist human operators, like collision avoidance systems in cars, lane departure warnings etc. I bet technology-assisted human operators will outperform fully automated systems for a very, very long time.
You’re very likely to be right about that. Over time, the level of technological assistance will increase, and the situations that require human intervention will become fewer. But there will almost always be corner cases that are beyond the scope of what the machines will react optimally, or even better than a trained human, to.
They will need a helipad to land the mechanics on board when they need repairs.
Well, this is interesting; those thirty thousand ton displacement automated ocean going vessels that take ten miles to stop after running over a, eight ton thirty foot sailing yacht at night will not have the crew to pickup the survivors, if any, assuming the collision registers with Captain Hal, the computer. Let’s ask Hal if he would stop: I’m sorry Mish, I’m afraid I can’t do that.
Automated ships (or cars, trucks, transports, planes, etc.) will never be perfect. But they will be vastly superior to what we currently have. Cost of building, operating, and insuring these methods of transport will drop significantly. They will become commonplace faster than you think. Remote operation or onboard AI will only get better over time. Ethics are being programmed into the AI systems, which means sometimes HAL will have to make an ethical choice, just like a human. I foresee the ethical programming to be the biggest area of controversy. There is no debate that this technology is going to be commonplace very soon, in spite of some of the problems mentioned here.
I always laughed at the techno-optimists. So much hype over innovation these days, while in reality, the pace of innovation almost grinded to a halt (just look at TFP)
If it takes 10 miles to stop and turn around how many survivors would there be?
Any?
By the way, what the hell is that craft doing in front of the liner in the first place?
And what makes you think Radar will not see the vessel, if it is of any size.
You are throwing up ridiculously easy to dispute nonsense, without bothering to think for one second.
You are very very wrong there Mish, on every count, I won’t even start explaining though some of the answers are above, the rest you would find out if you took up some form of ocean boating, or even simply read up on practical accounts and the questions faced by the maritime community .
I’ll be even harsher on you in fact, I dislike your attitude and would happily see to it that people of your nature were not allowed near the water without having some kind of experience hammered into them first. Even if you were only to endanger yourself you would also be endangering others, as that is the spirit of seafaring. No one, as in no one, lasts long at sea with the kind of arrogance you just displayed.
Mish – I have been a deck officer on USCG and USN frigate size vessels. All US Military vessels deploy human lookouts 24/7 as there are many vessels out there that are accounted for by radar. Ever here of “flotsam”? I personally have taken action to avoid collision several times with objects not shown on radar until too late.
I see technology as a great aid to maritime navigation but not a substitute for manned vessels. Quite possible that the occupants of an 8 ton vessel survive a collision with a larger vessel. The US Coast Guard and merchant vessels pull these folks out of the water frequently
This sounds like the typical wet dream from a “Naval Artichoke” as they are known in the industry.
Look at the Emma Maersk, Danish container ship, and tell me the deck house is a source of wind resistance? It is surrounded by containers fore and aft. They are the wind resistance. When Oskar gets through designing his ship for unmanned operation, the cost of it will go through the roof and outweigh the crew costs. Crew costs are minimal compared to fuel. Burning 336 ton/day, the Emma costs $102,000/day to operate just for fuel!!! When you add millions and millions to the initial costs to automate for crewless operation, only to have to scrap the vessel in 20 years because it is worn out, does not make economic sense.
please be serious
I think that is what he is saying to you Mish.
Mish: What part of my post is not serious to you??
I wish I could be there in 50 years when robot teachers are using posts from this column as examples that Luddites will be with us forever, no matter how advanced we get! Keep ’em coming Mish!
Root: Do you believe in the economics of scale? Ships are getting larger and larger. I used the Emma Maersk as an example. What do you think crew costs are versus fuel per day?? The 300+ tons of consumption is right on from the manufacturer of the diesel that propels it. The engine would amaze you landlubbers. https://www.google.com/search?q=emma+maersk+engine&biw=1280&bih=591&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwi9o82Rh4zSAhVB7WMKHTZUBdwQsAQIIA#imgrc=E5c069NLLjLbBM:
It weighs 2300 tons and develops 109,000 HP. A ships fuel consumption varies as the cube of the speed. Doubling speed increases fuel consumption by a factor of 8.
The crew numbers 13. In addition to the main engine, the Emma has 5 diesel generator sets. Maintenance on the gen sets on an unmanned vessel would require pulling it out of service to accomplish this. Off hire, no revenue.
Mish: I will give you a hypothetical case. The Emma has a crankcase explosion or ruptures a fuel line, the computer has to shut down the engine because a fire has started in the engine room. Now, who has to replace the fuel line or replace the crankcase doors? You are unmanned!
Oh, one more little problem, the Emma is in the middle of the South China Sea and raging Typhoon with 150 knot winds and 60 foot waves are bearing down on it, but you have no main engine. Steerage is maintained by electric motors supplied by the gen sets. Now, do you think the computer can judge the waves and swells, to supply the proper heading into the Typhoon to ride it out without the vessel foundering?
Lastly, let me know when Maersk lines decides to build some unmanned ships, then I’ll know some serious progress has been made. Until then it is still some wild ass Naval Artichoke’s wet dream!
These posts on autonomous operation seem to always deal with cars and trucks. Now this one with ships, which for many of the reasons people with maritime experience have mentioned, seem to be the least likely. One I haven’t seen yet, but would seem to be closer to prime time than the others, is trains. They are already down to 2 crew members and operate in a relatively closed environment.
Trains have too few personnel to matter
Passenger trains have more personnel but more need to keep them
Trucks first for reasons long discussed