Will Autonomous Ships Make the Seas Safer or Riskier?

The friendliest place on the web for anyone who enjoys boating.
If you have answers, please help by responding to the unanswered posts.
I truly believe that liability is the bigger factor. The potential for a plane crash, car driving on sidewalk, or ship plowing over kayakers may really be the obstacle.

Furthermore, the ethics associated with programming potential accident response might be detrimental to complete autonomy. I'm thinking of the damned if you do, damned if you don't dilemmas. Kill passengers on board, or kill people adjacent to the scene. The trolley car switch thingy...



Well, we get those kinds of dilemmas now, I think, so I'm not sure how the increased use of technology will change that. In fact removing the human factor in the decision might make the outcome more "fair" if such an outcome can be considered that.


I think the concept of "directing mind" would scare the bejesus out of most Boards of Directors. Corporations can act only through its agents and employees. I could conceivably see that they may find themselves culpable or even criminally liable for the "decisions" of an automated system if it fails.

I guess that's why I foresee continuously greater automation, but not complete reliance.

I would also predict that as Artificial intelligence becomes more sophisticated, it will become regulated, as nuclear, chemical, and biological weapons are currently. It will be the greatest existential threat to the human race, if not the animal kingdom.

That said, our hubris will probably take us right up to the edge in order to peek into the abyss.
 
I don't doubt the ability of the 'puter. But throw in the sea, the corrosion of salt water and murphys law and I doubt there is a solution in the offing now. Later? sure. But currently, No.


Oh, I would certainly agree with that, air or sea. Both operating environments have a long way to go before full automation can be totally relied upon.

For example we've just been talking about the airplane's capabilities here, but Air Traffic Control, which is horribly behind the times in many if not most parts of the world, is a whole other challenge. And when you start talking ATC you start talking governments and politics and the whole mess.

I would imagine similar scenarios exist in ocean shipping with politics, economics, national "priorities" and Lord knows what else making it difficult to currently impossible to set up any sort of globally agreed upon ocean shipping control environment.

While I firmly believe that fully automated air and ocean transportation is in man's future, I will be very surprised if anyone participating in this forum will live to see it. Their kids or grandkids may not live to see it, I don't know.

It's not something that's going to be introduced with the snap of a finger. It's going to take time--- a long time. What I find encouraging is that industry--- at least my industry, I have no knowledge of Cappy's--- is starting the journey to get there.

As John Cashman, the chief pilot of the 777 program during its development and flight testing liked to describe the flight test process, it's "baby steps" to find and fix the problems as they arise until one day you've crossed the road and you're there with a certificated airplane ready to go into service.
 
I would also predict that as Artificial intelligence becomes more sophisticated, it will become regulated, as nuclear, chemical, and biological weapons are currently. It will be the greatest existential threat to the human race, if not the animal kingdom.

Or--- is it where human/animal existence is going? Applying the morals, beliefs, philosophies, of today to what might happen tomorrow is a non-starter because that has never worked on this planet.

If we contemplate what might happen in the future as suddenly happening tomorrow, sure, we scare ourselves to death as this thread has demonstrated. All sorts of reasons why fully automated ocean shipping or air transportation can't work, won't ever work, and so on. "I'm never stepping on an airplane with no flight crew," has been a frequent comment.

Well, nobody here will have to. But..... like all the massive changes that have happened on earth, all of them were accepted to be just fine by the time they actually happened.

My mom had a college textbook about something or other and one of the illustrations in it that fascinated me as a little kid was of settlers in the US crossing the country by Conestoga wagon. It was an arresting illustration because it showed the settlers lowering a wagon over a bluff so they could continue on their way. I remember thinking that this was a staggering amount of work just to get from Point A to Point B and it looked really dangerous and it was a really dumb way to get across the country.

So you know that a bunch of years in the future people will be looking at pictures or videos of flight crews in 2015 flying their 787s and A350s with passengers sitting in the cabins behind them and they will think, "What a dumb way to travel. I can't imagine people having to do that. And talk about dangerous. You'd have no clue how those two guys driving it might screw up."

Because by then whatever fully automated, easy-breezy system of transportation they're using will be the norm, and what we accept today as the norm will appear to them to be totally ridiculous and inefficient and dangerous.

And ocean shipping? Well, we could easily say if the Titanic captain had had real-time satellite imagery he could have seen that stupid iceberg long before he was anywhere near it even at night. And the folks in the future looking at videos of the bridge of a 2015 new-build bulk carrier or whatever? "Can you believe those bozos had to actually stand on top of that thing and drive it all the way across the Pacific through storms and stuff? How nuts is that?"

So we never know what's coming.
 
Last edited:
Off Duty--- All I can say is that your way of thinking would have had us all stuck back in the stone age for fear of doing anything because doing anything will expose us to Something Bad Happening.

Fortuanately, your way of thinking--- which is the very definition of burying one's head in the sand, BTW--- is not the prevalent mindset in industries like air transportation and other such technological endeavors.

The challenge is not to give up at the first sign of a seemingly unconquerable problem--- which is what you seem to be advocating--- but to recognize these challenges and then figure out how to overcome them.

Will that lead to more challenges? Of course it willl just as it always has. So at what point do you sugest giving up? Because if one follows your line of reasoning to its logical conclusion, at some point the fear of the challenges will simply cause you to stop and simply stay where you are. Or try to.

The good news is that it's not human nature to do this. People keep trying and keep searching and keep overcoming the challenges one by one and then taking on the next one, and on and on it goes.

There will always be the naysayers and fearmongers and the like who express the sentiments you have expressed. But in the end, they are always bypassed or ignored until they're gone as humans continue their upward climb.


Hey guys. Can I request a caveman smiley? LoL
 
I can see both perspectives - the "go slow" approach and (as a business guy) the benefits to having fewer people on board a commercial ship.

It seems that the barrier to having fully autonomous cargo ships is going to be reasonably high for all the reasons cited.

BUt the more I read this discussion the more I started thinking that the faster adoption area for this "autonomous" technology in boats may be in the smaller to medium size trawler boats, for the simple reason that the risks are relatively low (how much damage can a small boat do to other things?), and the benefit in some specific situations (e.g. overnight passages in open water) could be quite significant. And the technology that could come out of the car industry - say Google's self driving cars - seems relatively transferable to the medium-sized boat industry (think Nordhavns, Selene's, etc.).

I'm thinking that much like people will or are getting used to the semi-automonous features like those currently in Mercedes S class vehicles (that allow the car to basically drive itself when on the freeway) the self-piloting technology in mid-size boats that go off-shore on overnight passages - seems pretty easy.

So - my question to you - for those who do overnight passages on the open ocean... How long will it be before you would accept more complete auto-pilot (sensor driven to watch for non-AIS type objects) on your boat for night passages?

What do you think of this idea? Seems like someone around here in the San Francisco either is, or should be, working on this.

I see it more likely to be a modified version of the self-driving car technology (which as to be pretty cheap to work in $50,000 cars) than a simplified version of the commercial shipping boat autonomous technology.

Also - since it would be off shore, without pedestrians and kayakers to worry about - it seems like this could be the first place of adoption (ahead of cars and ships that will face many legal and political issues).

Your thoughts?
 
The maritime industry has long been right in the thick of adopting technologies that reduce manning requirements. For example, square riggers gave way to fore-and-aft rigged schooners; coal-fired steamships yielded to oil; muzzle-loading cannon gave way to breech loaders, etc. Containerization of ocean freight is the highest form (so far) of that trend. It's how competition drives innovation.

As for automation on smaller vessels, like a passage-making motor yacht, the technology has been getting there, too. I have made short-handed (myself and one crew) offshore passages under sail, and relied on an autopilot plus electronic countermeasures such as a robust radar reflector (Firdell Blipper) and detector (Locata Watchman). Both left me unconvinced that I had gotten my money's worth, and both are now long since out of production. But I reached my destination unscathed - meaning, I guess, that I got away with it that time.

The thing that sobered me up about offshore small-boat passage-making with limited lookout capability is semi-submerged shipping containers. I've seen them by day, barely awash, and imagined the consequences if my 40' fiberglass ketch had plowed into the thing at six or seven knots.

One morning on the 0400-0600 watch, the Sun rose to illuminate hundreds of timbers, approx. 4" x 4" wallowing in the sea around me. They must have fallen off a cargo vessel. Any one of them could have stove-in my hull just at the waterline. How I missed ramming one in the dark is something I still shake my head over.

Anyway, those are things I have pondered whenever tempted to set the AP and doze off for a few minutes. The technology probably exists to detect such hazards, but at what cost, and with what level of certainty? Running a ship or a boat is a bit like running a business in the sense that risk is ever present. Success means managing risks effectively. Here, the risk ratio presently looks unacceptable.

The sea is beautiful, but indifferent as to whether we cross it safely. Audacity and luck will only get you so far.
 
The thing that sobered me up about offshore small-boat passage-making with limited lookout capability is semi-submerged shipping containers. I've seen them by day, barely awash, and imagined the consequences if my 40' fiberglass ketch had plowed into the thing at six or seven knots.

One morning on the 0400-0600 watch, the Sun rose to illuminate hundreds of timbers, approx. 4" x 4" wallowing in the sea around me. They must have fallen off a cargo vessel. Any one of them could have stove-in my hull just at the waterline. How I missed ramming one in the dark is something I still shake my head over.

I think you make very good points - the risk to boats of smaller but hull-destroying objects is significant.

But I also wonder how well our human eyes and minds are at discerning these things when we're sleep deprived and alone on watch?

Any new autonomous auto-pilot boat technology would have to be as good and hopefully a lot better than our fallible human alternative. Just as in cars today - I can see these being phased in with people still standing watch as basically another set of eyes scanning the horizon looking for risks and then as we build our confidence in them we let them do more and more.

It seems that with these sensors positioned on tall masts (on sailboats) or above the radar on polls - the actual discernibility of the systems might actually be much better than our eyes at close to water level.
 
Interesting - a quick search shows that exactly this is happening:

Autopilot system for recreational boating responds to threats before they occur

Autopilot systems for yachts and inboard boats are a common backup that gives the captain an occasional rest. But these systems are designed to react to changes in conditions after they occur, which may be too late in certain circumstances, whereas the safest, most ideal system will preempt threats and react in anticipation of a coming danger. Google and others have been developing such systems for driverless cars for years, and now a startup spun out of the University at Buffalo hopes to sell preemptive marine autopilot systems to small and mid-size recreational boat owners.

Recreational boating accidents cost hundreds of lives each year. US Coast Guard statistics show over 4,000 accidents and 610 deaths (equivalent to 5.2 deaths per 100,000 registered vessels) from recreational boating in 2014, with operator inattention and inexperience, improper lookout, excessive speed, and alcohol use rated as the top contributing factors.

See the full story:

Autopilot system for recreational boating responds to threats before they occur

Other groups working on Autonomous Marine systems (a short list of many):

Origins | Autonomous Marine Systems

Somerville startup is testing an unmanned catamaran that will be capable of months-long voyages
Somerville startup is testing an unmanned catamaran that will be capable of months-long voyages | BetaBoston
 
Shipboard automation will continue to increase. That's a given, and a good thing.

Going to fully autonomous, i.e., no humans at all, doubtful for a looonng time.

Marin, your praise for the computers and the young engineers is well founded. The tech is truly amazing.

But I sense you do not have deep experience with shipboard propulsion, electrical and other systems. Even with all the automation, much of those systems remains very similar to the less automated systems that preceded. Engines, pumps, turbines, generators, filters, valves, miles of pipe, breakers, switchgear, monitoring gear, and more pumps.

Automate all you want, and those systems are going to remain similar. And things BREAK. A pressure transducer starts spewing oil. A pump shaft seal burns up. A breaker trips below trip current. A contactor coil faults open. A ballast control MOV limit switch sticks.

If you need to look up what a MOV is, I proved my point!!

And airliners are not a good analogy. A turbofan engine is about as complex as a turbine driven feedpump, and has maintenance standing by after a 12hr run if anything acts up.

It will remain economic to have a crew on board to operate and maintain the vessel as it operates. And to handle unforseen events... for a long time into the future.
 
Autonomous Ships

This may touch a raw nerve in some, and if so I apologize in advance. I will try not to refer to Star Wars.
Someone above mentioned cost factors, and the biggest cost factor is humans. They only want to work 8 hrs a day, 5 days a week, with paid holidays and vacations, and time and a half and double time on Sundays even if they don't go to church, and time off to get married then demand a better wage to raise a family, and health insurance and retirement and time to eat and a nice place to eat in ,and bathrooms, and a good place to nurse their babies, and time off if one of their little computer chips get sick. Most places devote 99% of their entire budget to payroll related expenses.
Robots only demand a bit of oil and energy to work their servo units and if one gets in the way of bigger ones and get crushed, smashed, sliced diced or otherwise eviscerated then his/her/its family won't sue for wrongful death, unsafe work environment or loss of potential income. Just pick up the pieces and replace it...uh..him...or her....
Labor unions and OSHA have proven repeatedly that the strictly business approach to corporate America cares only about the bottom line that begins with $$ dollar signs, and they had to be forced to care about the human factor.
If a company can replace 5 million dollars of payroll related expenses with a two million dollar robot, the only question is, why not?
 
And airliners are not a good analogy. A turbofan engine is about as complex as a turbine driven feedpump, and has maintenance standing by after a 12hr run if anything acts up.

Having spent the last 36+ years living and working with airliners and airlines all over the planet I can tell you that the modern high-bypass turbofan is amazingly attention-free and is usually the one thing on the plane the maintenance folks don't have to worry about.

I have directed shoots of thousands of airplane turnarounds all over the world, from 727s at Fedex on up to Emirates' huge 777 fleet (haven't done any 787s yet, though). The plane comes in, people and cargo get off, the plane is fueled, people and cargo get on, and the plane leaves. That's it. A line guy checks the oil in the engines and adds some if needed and that's all the attention that engine needs.

Sure, there are maintenance folks on-site at the major airports. But unless something actually breaks--- very, very rare on the current generation of turbofans--- they are not needed. At least not for the engines.

More often than not what they're needed for is to fix the bloody IFE system (in-flight entertainment) which most people do not know is the single most complex (and hated) system on a modern jetliner.

And equipment is improving in reliability all the time, even the IFE systems.:) Potential breakdowns are taken care of on an airliner with redundancy. No reason (except economics) not to apply the same principle to ships.

The bottom line is that humans are expensive, not only in the fact they have to be paid, but even moreso in the fact that when they screw up, either in the air or on the ocean, the financial penalties can be staggering.

As component reliability continues to improve, and sensors and data input systems become more and more comprehensive, and airspace/oceanspace control systems become comprehensive, coordinated and reliable, the point will come when operating a plane or ship under full autonomous control will be less expensive than operating it with humans.

Because humans will always be accident prone. That's what things that are "alive" do, be it an ant, a squirrel or a human. They screw up.

Machines only screw up when they aren't developed to the point where they can't screw up or there are other machines that automatically take over.

But once that point is reached, you get results like the BMW assembly plant in Greenville, SC turning out 1,200 X-cars a day according to the CEO of the plant who gave me that figure a few years go, to supply half the world's demand for BMW X-cars. (The other half come from a plant in Germany). The SC plant can do this because it is almost fully automated.

You're absolutely correct, Ski, in that we're still a long, long way from fully automated operations in aviation and shipping, although at the continuously accelerating rate of technology development it may not be as long a way as we think. But the day will come because there is no reason that it can't.

The thing humans bring to the table--and will always bring to the table-- is vision. The vision to conceive an idea, the vision to figure out how to make the idea a reality, and the vision to actually make the idea work. In that respect, there will always be humans involved in air and ocean transportation because there will always be ways to improve it. But there is no reason that humans have to keep driving the machines.

They have to now, but they won't have to in the future.
 
Last edited:
It's not so much operating it, it is fixing things when they break. Hard to do that with automation.
 
It's not so much operating it, it is fixing things when they break. Hard to do that with automation.

That's what redundancy is for. The failed component can be replaced or repaired at the destination, airport or seaport. And, as time passes, components are improved to become even more reliable.
 
Redundancy is a wonderful thing, but it does not assure complete reliability. Some process lines need valves, so how is that handled? Two valves in parallel? Two valves in series? In parallel any one can open, in series any one can close. So you can have redundancy to establish flow or stop flow. Hard to get both. Same with breakers and contactors.

So what do you do when a ballast control valve sticks open?
 
Technology has not proven to be perfect. Until and unless it can be shown to have complete 100% redundantly redundant perfection then there will always be a need to have a (poor, stupid, ignorant, feeble) Human as a backup to take over when the ship hits the phan. (so to speak)

My particular example had a computer failure the last day I was aboard. The Techies were falling over themselves because the symptoms had never been seen before. Yet, they couldn't stop the symptoms, nor could they replicate the occurrence. The solution was to replace the computer.

Find a computer that can 'replace a computer' and I will buy into the concept. It hasn't happened as yet. Now add in the variables of the ocean, humidity, short circuits and vibration. The industries not ready (yet).

I would ask: Is brain surgery completely 'human free'? Sure, computers, TV screens, robots, and other aids. But who does the final decision? If they can't completely trust a robot.... Until ALL boats are antonymous, there will always be a 'human' element. There is a term for this: left handed navigation.

Wrong decisions don't make correct solutions.

Good thing you don't insist that humans be perfect; we'd never get anywhere.

For every technology failure, I'll bet there's a dozen human failures. Will the technology ever be perfect? Of course not. But that's not the correct question.

The correct question is, "Will technology reduce accidents?"
 
The correct question is, "Will technology reduce accidents?"

Exactly. Which is why the aviation industry, at least, is working on reducing the likelihood of human errors, which are the cause, directly or indirectly, of the vast majority of aircraft accidents, by eliminating humans from the equation.

All of the human traits that can contribute to an accident--- fear, poor judgment, insufficient training, cultural or gender conflicts among the flight crew, confusion that leads to taking the wrong action, emotional stresses that can lead to all sorts of abnormal behavior, egos that prevent a flight crew member from admitting or even realizing he or she is wrong, social status that prevents a first officer from correcting a captain in some cultures, and a whole lot more--- go away with automation.

I have to assume that the same situation exists in the maritime world, too although this is an industry about which I have almost no knowledge or experience. Perhaps the forum participants who work in this industry can comment on the degree of risk human traits pose to maritime safety.
 
Last edited:
So, are you saying there will be no need for humans to maintain the ship's systems while underway? Don't think that will happen until humans are obsolete.
 
So, are you saying there will be no need for humans to maintain the ship's systems while underway?

Nobody maintains an airplane's systems while they are "underway." Redundancy takes care of that and there is no reason that it can't with a ship either.

And you are judging this by the reliability of today's equipment. Along with computer, data acquisition and manipulation, and artificial intelligence developments that will make autonomous operations practical will come vastly more reliable on-board systems. Comparing what is coming with what we have now is like comparing today's vehicles with Model T's, or probably more accurately, horses and wagons.

When I hired into Boeing they were designing airplanes with slide rules. They built three full-size physical mockups-- Class I, II and III-- for every new airplane model to test the fit of all the components. The last planes that were created using this ancient process were the 757 and 767. Today, only 36 years after I hired in, we design planes using 4th-plus generation computer applications and we test the fit and function of the entire plane the same way. The first plane we did this with was the 777 and the applications we used to create that plane are staggeringly crude to what we use today.

Today, for example, we have applications that use 3D models of people--- and you can define the person: thin, fat, tall, short, male, female--- and tools to see how easy or hard it will be for a maintenance person to remove a specific component anywhere in the plane. You give the "person" the command to climb into the space and remove such-and-such a component and the "person" selects the necessary tool(s), climbs into the 3D plane, works his or her way into the space and removes the component, all in real time. So unscrewing screws, removing nuts and bolts, whatever it takes. And you can move the "camera" around to view the action from any angle while it's happening.

If there is not enough room to move the tool or fit a hand or arm into a particular position or space the program calls this out and illustrates it.

We use another program from Dassault in France called DELMIA to determine if the airplane we designed will actually fit together in assembly. Same thing as above--- the 3D components, from wings to landing gear, are put together using 3D models of the tooling that's been designed to support and move them. Any interferences are highlighted so the design engineers and the tooling folks can see the problem and resolve it be it an airplane problem or a tooling problem.

The time interval often used here to define a total change in the way things are done is twenty years. As in "Working on airplanes as an engineer here at Boeing twenty years from now is not going to be anything like it is in today's world," (a direct quote from one of our senior engineers today).

So don't make the mistake of thinking about the future in terms of how we'll do things and the equipment we'll do it with by using what we do and use today. Because it will all be totally different in ways today's people, particularly the older ones, can't even conceive. The young ones can, though, which is why this is such an exciting time in this industry as the older generations are rapidly being replaced with the younger ones.

And again, everything that applies to the air transportation industry in this regard is applicable to the ocean transportation industry. In both, the driver is economic incentive.
 
Last edited:
Good luck with that. What happens when things break? Or a fire? Or pirates?
:thumb:
Thanks for bringing us back to reality;:hide: but that never stopped a thread from reaching astronomical heights of ludicrousness.:eek:
 
Marin- awesome testament to the new high tech. But the question remains, how does that handle the stuck ballast valve?
 
Marin- awesome testament to the new high tech. But the question remains, how does that handle the stuck ballast valve?

Ballast valves stick today. That doesn't mean they will in the future.

It's human to think of the future in terms of what we know today. After all, the present (and the past) is all we know. But the future, particularly the technological future, is always vastly different than what we know today.

I don't have an answer for you regarding ballast valves because I have no clue what kind of ballast systems will be developed for ships in the future and neither does anyone else right now. The only sure bet is that whatever they are, they will be a whole lot more efficient, effective and reliable than what we have now.

Try to wrap your head around this scenario: When Johnny Roman soldier guy decided to buy a new chariot and went to the high-end chariot dealer over on the Piazza Navona, what he saw there were the were the coolest new chariots on the planet. That, to him, represented the absolute ultimate in transportation.

Now what if another guy had come up and said, "You know, Johnny, someday people will be driving around in carts powered by lightning and they will go more than 33 leagues in an hour. And I have a hunch they will be called Tesla."

Johnny would have looked at him and said, "That's ridiculous. How are you going to get a horse to go that fast?"

The stuck ballast valve question and Johnny's question are basically the same.
 
Last edited:
Disasters are averted on many occasions when something goes wrong because experienced people who can rapidly assess many factors are involved.

Disasters occur because two or three things go wrong at the same time, and even the most experienced people have little chance to stop them.

Computers cannot be programmed for the unforeseen.

How about catastrophic gyro failures? We had one in Douglas Channel with a pilot aboard and it still bounced off the beach;

"On Sept. 25, 2009, two hours out of Kitimat in the Douglas Channel, the freighter Petersfield suffered complete gyroscopic failure, lost steering, took a turn to starboard and struck a rocky outcrop across from Grant Point."
 

Attachments

  • petersfield.jpg
    petersfield.jpg
    20.1 KB · Views: 113
Last edited:
I think we are not ready or capable of autonomus ships with todays understanding of the issues...things would be riskier.

But in the future, which will arrive sooner than us old farts expect, a much more highly automated ship is feasible, that reduces crew to a fraction of todays needs.

The crew makeup will be different for sure. They will have to be more nerdy than salty, but the salty part is muy importante. As we all know, the seas are an unforgiving mother.

1200 BMWs a day is easy on land, where the enviornment is totally under control compared to the Atlantic or the North Sea. It is the variability, the unpredicability of the sea that makes this problema hard to codify.

Progress like this comes from those who say "why not." I used to do a lot of that, and it helped my customers (and me) get to where we wanted to go, which is now sitting at anchor in my trawler responding to this thread while anchored in Puget Sound.

When I started out as a mechanical engineer, I could not imagine that I would be doing this when I "retired." And that is how ships will become more automated in the future, because someone will imagine something and say why not.
 
1200 BMWs a day is easy on land, where the environment is totally under control compared to the Atlantic or the North Sea. It is the variability, the unpredictability of the sea that makes this problem hard to codify.

Absolutely. It is also, in my opinion, one of the three major challenges to fully automated air transportation, the other two being airspace management and control and passenger acceptance. The airplane itself is the easy part.

However..... not all that many decades ago when a hurricane came thumping in it was anyone's guess what was going to happen. Nobody knew how strong the winds would be, how much rain, what the storm surge would be, what direction the damn thing was going to go. (For a true story about hurricane prediction way back when, read the terrific book Isaac's Storm by Eric Larson. It's the story about the hurricane which in September, 1900 totally destroyed the town of Galveston, Texas and killed 6,000 people.)

Today, we can predict all these things. Sure, the predictions aren't always spot on as Hurricane Juaquin just demonstrated to us. But compared to not knowing anything, the current prediction capabilities are pretty impressive.

And, like everything else, they will just keep getting better as our knowledge base increases and our technological capabilities keep getting better.

So I, and more and more other people in my industry at least, have no problem envisioning the day when weather will not be the unpredictable bugaboo that it still is today.

As predictions get faster and more accurate, that becomes data that can be fed to increasingly "smart" computer/artificial intelligence systems that can have the ability to guide aircraft or vessels to avoid the kinds of situations and problems we don't have the ability to accurately and consistently avoid today. Or worse, situations and problems where today our only choice is to rely on amazingly fallible human judgement, hunches, opinions, guesses, you name it, to avoid.

Nobody's claiming this is all going to happen tomorrow. But there is no reason, outside of people's inability to see beyond today, that it can't happen. Fortunately for mankind, since time began the naysayers always die off leaving the next generations to move ahead.

Of course, each generation eventually becomes naysayers in turn, but as time goes on the naysayers won't be denying fully automated air and sea transportation anymore--- that will have become the norm--- but will denying the possibility of time travel or some such thing.:)
 
Last edited:
Computers cannot be programmed for the unforeseen.


The ones we have today can't, you are correct.

But the moment you can envision something that might happen, it's no longer unforeseen. And once it's no longer unforeseen, you can foresee it and therefore program to avoid it.

I can easily see how the incident you described could be a scenario entered into a data base that controlled a vessel. I don't have a clue how such a system would actually work but I have no problem envisioning that it would work once the necessary software, hardware, data base, etc. were developed.

This sort of thing is already being done to a degree in aviation with scenarios being envisioned and then the airplane's flight management system programmed to avert them or react correctly to them, in some cases over-riding what the flight crew might mistakenly try to do or forget to do.
 
Last edited:
But the moment you can envision something that might happen, it's no longer unforeseen. And once it's no longer unforeseen, you can foresee it and therefore program to avoid it.

Think about what you said there...then think about the infinitely intertwining eventualities you would have to think about to slap that computer program together. (Take your time.)

Then, there's black swan events...
 
That's called model predictive control (MPC). Using multivariate inputs to measure and compare to the desired result envelope, then create outputs to not only achieve the result, but to predict future I/O requirements to optimize the system, swing less and ultimately make the operating envelope smaller.

I think the key thing that many are missing here is that technology is not something that replaces humanity, but is, in essence, an extension of humanity. So technology is likely to inherit many if our human shortcomings.

That we are predetermining the requirement for autonomous ships may in itself be a fallacious goal.

I mean, if we are so technologically advanced in the future, why would we want to move goods across an unstable interface of two fluids, one fluid corrosive, at a relatively low speed (aka shipping) anyways? I'm sure there is a better solution, if this is even a problem then...
 
Last edited:
Think about what you said there...then think about the infinitely intertwining eventualities you would have to think about to slap that computer program together. (Take your time.)

Then, there's black swan events...


You mean like predicting the 7th inning of the Jays-Rangers game today?
 

Latest posts

Back
Top Bottom