Autonomous marine travel

[psneeld]

Hmmm.... sounds like the same old discussions of VHF radio, then DSC, then Radar for small boats, then autopilot, then GPS, then computer nav, then AIS, (probably more I have forgotten).... the difference to me...is I could understand the basics all those and all those are probably necessary for autonomous navigators/helmsmen. Through the years, I could delete or add any of those and still be legal and safe.

Autonomous nav...I guess it is at least rudimentary AI based and probably very sophisticated AI for those vessels already making historic trips. Have again, only the basic understanding of this new tech, and how it is about replacing me and my training and experience to a point. Those other things added to boating pleasure/safety because I could not do most of what they do. Sure for AP I can hand steer, and sure I can estimate CPAs like AIS but they still make it easier. Seems like it as most people swear by them.

I just wonder how many and how quickly many will be converts to the tech like I described above and get it, maybe not use it all the time, but will treat it just like some of the things I mentioned above.

 

[SteveK]

Hippo. "while watching the tube"
That comment made and understood certainly dates us.

 

[Jklotz]

g.
It’s not if but when. Accept that it’s inevitable.
Brits don’t fire people they make them redundant. You’re redundant. You just don’t know it yet.

Ha! Maybe we should change the old saying to "Everything on a boat is already broken AND you are redundant, you just don't know it yet".

 

[delirious duo]

guys with wood boats and oil lamps with sails say same things about power boaters that’s mentioned about machine piloted boats. Perspective bet if it lowered full coverage insurance when my budget jumps from 50k boats to 500k if the admiral goes full time.

 

[DDW]

The natural termination of this arc of lifestyle is well parodied in the movie Wall-E. You can simulate it now by putting on some VR goggles, leaning back in a recliner, and going on a VR boat cruise.

 

[psneeld]

Some people love the trip, some love the destination.....some love both.

For some, VR is the only way they will see it...but many here on TF have been, there done that and are still going new places doing new things.

VR never replaces that.... for those that the destinations are more to their liking and enjoying the trip without being the captain (IOW, letting Otto Pilot and First officer "Computer" do all the driving) this tech sounds right up their alley someday.

Some still love playing with sextants...but I bet few bet their lives on them if they have GPS aboard.

 

[Snapdragon III]

I Feel like I already have an autonomous system for my boat. It's just crappy and dangerous, so I have to watch over it. I would love to have a way better, safer one. I try and preplan all my routes then put the autopilot in Nav mode and let it follow the route. I feel that it adds greatly to safety as you do all the planning while you are actually paying attention, and are way less likely to be spacing out on hour 6 of your trip and hit a rock. I would love to have a system that watched for logs and would alarm, or dodge them. Would be so nice for the trip up to Alaska when you have many 10+ hour days, and thousands of logs to dodge.

As for autonomous cars, I just got Tesla's latest 12.3 Full self driving software update a few days ago, and it is astonishingly good. I have driven it about 300 miles of mixed highway, and city so far, and have not had to intervene a single time for a safety issue. I have taken over a few times because the Nav was making stupid route choices. I feel like it is making about 10% of the mistakes that the previous software version I had last week was making. Self driving cars are coming fast.

 

[Mambo42]

Interesting to read the points of views of everyone, great discussion.
Are we indeed redundant ? And if so, is that a good or bad thing ?

With automation and things like AI we can run into the trap of complacency, thinking we don't need to know certain things anymore, after all the computer knows it better. While in certain cases that will be a correct assumption, in my opinion it will only apply to certain conditions and two of them are a perfect technical condition of the vessel as well as a perfect weather forecast.
What we will get to is replacement of the skipper by basically a systems monitor. When is that observer allowed to interfere and when not ?

For commercial shipping I can understand it might be interesting. Decisions like what sank El Faro would not happen anymore and ship collisions could become something of the past.

However, I have witnessed and seen the 'progress' in the airline industry. It went to the point where the system was so complicated that it required extensive training, something many airlines were unwilling to provide. But it went further, also training of pilots became less and less, which has resulted in several very avoidable accidents with great loss of life.
As an example I can point out Air France 447, which crashed because the pilot flying had limited systems knowledge, limited training and had no cognitive abilities. In other words, it was a completely avoidable accident, there was no need for all those people to die. Better training would have prevented the accident.
And there have been more crashes that were totally avoidable.

When it comes to ships I can understand that, if all the ships can communicate with each other, they should and will be able to avoid each other under all circumstances. It only becomes a problem when technical glitches prevent normal operation.
Just this week it was published in the news that Russia is apparently busy with jamming the GPS signal in areas in Poland, Lithuania, Estland and even Finland. For a well trained pilot that is not a problem, you just switch over to a different type of navigation and off you go.
When it comes to ships however, loss of GPS signal (or worse, such as distortion of the signal, so that wrong positions are received) or loss of means of communicating the result could be much worse if we would start relying too much on automation. The airline industry has shown where the problems lie.......the operators lose the ability to analyze the problem, take control, take the proper actions and land the aircraft safely.

If this same scenario plays out in the shipping industry we could be in for a very unpleasant surprise. The level of maritime education goes down and human error will become more and more the cause of accidents in very simple technical problems.
So to answer the question: 'will humans become redundant in shipping' ? Most likely it will happen, but I don't think it is a good development.
Already at this moment the biggest cause of maritime accidents is lack of training, i.o.w. system knowledge.
Make the systems more complicated and we will see more accidents, not less.

What will it mean for us, the pleasure boater ? That all depends on what we are willing to accept. I can understand that the more expensive yachts (charter industry) may implement these new technologies and perhaps full automation. For the smaller boats this may become simply too expensive.

 

[RedRascal]

I'm curious, on this, as we get log's in So Cal after the rains, nothing like you guy's do up there, but, what if you are running at night, which I often do, to the fishing grounds, there is no seeing logs at night, even with Flir, it would be hard to pick out. I'm wondering if these new systems would work in conjunction with a Flir systgem to pick up these objects at night and change course accordingly.

In he PNW we generally get two flavors of trees in the water. Those that fall in the water with branches and those that have no branches because they have been in the water forever or have gotten loose from a logging operation. Tugs move logs in crazy big rafts around here and sometimes logs will break free. The trees with branches are the easy ones to spot. For night time running I slow down from 7 knots to 5. Also I feel a little better with a lot of moon light or our long summer nights where we still have light up to 10:00pm. I usually don't run past midnight. I also have a ditch bag and target being able to get into the dinghy in 2 minutes or less. Logs have sunk boats up here and the bad ones that I heard about force you to get into something else that floats in 90 seconds to 2 minutes. I think FLIR would pick up most of the logs but there are always a few devils out there just beneath or at the surface. Some times they will break up the wave pattern enough to indicate danger but sometimes not.

 

[fgarriso]

RULE 5

Look-out

Every vessel shall at all times maintain a proper look-out by sight and hearing as well as by all available means appropriate in the prevailing circumstances and conditions so as to make a full appraisal of the situation and of the risk of collision.

 

[psneeld]

From what I have read, autonomous travel has electronic lookout that in some ways can be superior.

Maybe not yet, but ultimately....superior vision, night vision, FLIR, better acuity, no fatigue, 360 degree full time, no distractions, more sensitive hearing for sound signals, instantaneous use of all available means (sight, sound, radar, AIS...maybe more)...... all might be better when that tech comes together fully.

 

[Hippocampus]

As to PS’s point. There’s a fundamental difference between all aids to navigation be it radio directional finders, loran, GPS, weather modeling , predictive AP and autonomous navigation/piloting. In the first case you are being aided. In the second you are turning over the nav/piloting experience to a machine.
I see this fundamental difference in other settings. Many neurosurgery OR suites have integral MRI and robotic surgery. The tech allows the operator to get much more of the tumor or other target without untoward injury to eloquent areas of brain. However the other choice is stereotactic radiosurgery . You put in the diagnostic imaging (generated by a computer system). Then a program decides how to map out the beams to deliver cidal doses to all areas of the target. The skull isn’t opened and comorbidities lessen. The entire procedure is computer driven. Fundamentally there is no human operator. It is an entirely different gestalt. The advanced skill set required of a manual neurosurgeon are no longer required. He/she is no longer holding a knife. Yes there’s a huge skill set required to develop and maintain such a system. But it has no resemblance to what occurred in the past. Is it better? Unquestionably yes.
Unlike such a system which is entirely goal directed getting from point A (tumor/AVM etc.) to point B( no neoplasm, vascular abnormality) recreational boating generates pleasure from the mastery and application of a skill set and mode of observation. Those of you who hunt would understand. I walk through the woods and “see” the presence of deer,coyotes, bear and such. I can “read” what happened that day and the last. My wife is blind to it. I’m out on the water and “see” weather to come in the clouds or change in the wind, currents, fish under the water and the depth.or “seeing” land in the clouds or smell or wind quality long before it’s over the horizon. My wife is blind to them. Driving I know which cardinal point I’m heading. Wife needs to be told. In all spheres of experience the experience is fundamentally changed by what knowledge you bring to it. Yes when I need to go from point A to B I turn the driving over to driver +. When I’m cruising in the truck or the boat I don’t if time allows. Part of the experience is lost. I lament our experience of the world will be fundamentally changed and degraded as time goes on and those skills aren’t required so not learned.

 

[rgano]

Three questions and three times no.

 

[Pascall]

I will help many out of the dream.
On the major rivers between, for example, Rotterdam and deep into Germany, several cargo ships are already sailing fully automatically, and that on one of the busiest rivers in the world.
I have a friend who sails up and down the Rhine between Rotterdam/Duisburg twice a week.
A year ago, he switched to fully automatic sailing and is now more of a supervisor.
This very experienced captain says the following; With all my experience, I can't match the computer that controls the ship, it's really incredible!
By automating, it is expected that fewer accidents will occur because 99% is a human error.
Every day, bridges are hit because they do not lower the wheelhouse in time or the load is too high for the bridge, this is now excluded by automation.
The new development is the legislation, because it is now lagging behind the facts, and a lot of work is being done on that.



Greeting

Pascal.

 

[SeaDogAK]

There are some changes that I’m not sure I am able to adapt to. I guess there’s a reason we aren’t meant to live forever.

 

[RT Firefly]

Greetings,
Mr. P. I can see this being a major advantage when transiting a known, set route on a regular basis BUT I think it's a different set of circumstances when cruising in potentially "new" areas or areas known for changing hazards (logs/shifting sand bars etc.) aside from other marine traffic.

 

[Jklotz]

Greetings,
Mr. P. I can see this being a major advantage when transiting a known, set route on a regular basis BUT I think it's a different set of circumstances when cruising in potentially "new" areas or areas known for changing hazards (logs/shifting sand bars etc.) aside from other marine traffic.

I'm sure they will incorporate some sort of forward seeing sensor to avoid obstacles. My car has that now. Shurley the technology exists I would think. I do see your point though, commercial vessels would benefit greatly whereas a pleasure craft may use it a lot less frequently.

 

[psneeld]

Back to the OP's questions....

Do you ever see yourself traveling like that? Yes...there are times that I want to do something else rather that be the helmsman/lookout. Probably not the whole trip but during periods of it such as food prep, long head calls, fishing, engine checks, etc....my usual crew I KNOW doesn't know the rules well, sometimes vision is in question, attention span is never perfect....etc...etc...

Do you think it will make boating safer? Compared to many skippers....ABSOLUTELY...even a lot of commercial guys. Sure not as many as on here but overall...yes.

If it were available today, at a reasonable price, is it something you'd consider installing on your boat? yes for the reasons I stated in one. Some of those reasons can be augmented with yesterday's tech like radar with guard zones and on autopilot with AIS as a bonus...but if the AI incorporated in the new tech is WAY more knowledgeable at maritime decision making and vessel control...then I would prefer that at a "reasonable price"

For those that like to play with sextants and plan trips and drive the boat hands on...ya know... the same ones that feel thrusters over ship handling (couldn't resist)....that's fine.

I too take pride in all those seaman-like skills and now that I am out of big boating...I don't think I will ever know if I would get it. I do know there are times I with I had autonomous driving for my RV... not all the time but construction zones with zero clearance on both sides and city loop roads packed with tractor trailers and chuckleheads zipping in and out of lanes...you bet I would pay BIG bucks for it. Those two things keep me more white knuckled than all the years I flew and drove assistance boats.

Not so sure others would feel the same with boating if they wanted to cruise with limited experience. Who knows, the system maybe could be programmed to be like an instructor captain....just teaching what to do but taking over or give stern advice like a training captain would do if things went beyond the owner's capabilities. Heck, it may add years to cruising for those over in the aging thread that worry they are just not sharp enough anymore....but if Gizmo was there to double check your work....not a bad first mate (probably wouldn't be in the liquor locker either)

 

[Hippocampus]

To PS Pascal has it right it’s inevitable and actively being done. I’ve stated that multiple times. To date little has been said about what’s lost. I’ve pointed out multiple times in multiple recreational activities what is lost. One can imagine a future where you will have difficulty obtaining insurance or pay a higher premium if you aren’t automated. There’s no question that for cars, trucks, boats and ships the overwhelming majority of adverse outcomes are due to bad human decisions. Insurance in large measure decide premiums on the basis of risk so it’s inevitable even without legislation. Still there’s a definite pleasure utilizing old school methods. Having done Bermuda races using celestial, a recording barometer, observational skills to predict weather, a wind vane instead of a AP gives an organic experience not matched by even semi autonomous sailing. As one sheds responsibilities one disengages from the activity and unfortunately something is lost.
No argument less accidents will occur, less mentally and physically skilled will be safe, less intellectual prep will be required. All to the good but the experience will be degraded.
Will I use this tech. Of course I will. I’m no Luddite. But will give thought to to running dark and old school for the joy of it.

 

[psneeld]

As a PS to my post...what may be lost for a few may be gained by many if it IS adapted to be first mate and instructor as opposed to ONLY autonomous marine travel which generally means no one aboard.

So I guess the whole thread means will true "autonomous marine travel" take the place of a captain...from ours to cruise ships in the capacity where we are strictly passengers or does it look to the "tech" which will accompany us on cruises where a simple switch turning it on/off or a set of filters that lets us still function as part of a human/robo team?

It would be silly to think that most on this forum don't enjoy cruising for what it is....using skills.

Boating however to many boat owners is not "cruising" whether day tips or long term. They fish, water sport, transport, etc..... so letting a robo captain manage the boat may be the greatest tech of all.

So obviously different views are mostly answering different perceptions of the tech and how it will be used.

 

[dannc]

I can see automation being able to cross large bodies of water are well surveyed coastal areas and rivers easily.

However, I know areas where the inlet is constantly changing and the chart is questionable. One has to look at the buoys AND read the water. There can be very little space between the shallow water and the buoy. I suppose the automation "vision" can detect which is port side vs starboard via the shapes. I have been in places where the channel is snaking around and figuring out which buoy is which is interesting. These situations would be challenging to program.

How does the automation work in the vast areas of the world where surveys are from the day's of sail, if surveyed at all? Hopefully, the program just fails to provide a route or pops up a big warning message multiple times.

The various phones we have used over the years for navigation have caused us to travel to "interesting" areas. Some not really safe to be in, and once, took us to the front gate of a major army base. We were not alone, plenty of other vehicles were being turned around, and the MPs were NOT amused. Even our latest and greatest phone has sent me on routes that make no sense or are not the best route and this is simple road navigation that does not take into account draft, air draft, tide, currents, and weather that vessel automation would need to take into account.

I can see auto routing be useful as a suggestion but one would really need to look at the route closely. I read about coastal areas in Ireland and many of the areas will have places one should not enter, or pass through, without local knowledge. The chart looks "ok" but.... For example, one entrance has two rock islands/out crops named Adam and Eve. The local information is to avoid Adam and hug Eve. The charts are such that one wonders if one should pass starboard or to port. How is the automation going to know to avoid Adam and hug Eve and one which side?

Even with local knowledge, fishing trawlers using the port have sunk after hitting rocks.

What happens with the automation when the data is is using is wrong or non existent? Much of navigation is SMOP(Simple Matter Of Programming) but the programming has to be perfect and it has to operate on perfect data. Eventually the programming will get really good but the data? Which weather forecast should the program use? Which chart?

 

[Hippocampus]

I can see automation being able to cross large bodies of water are well surveyed coastal areas and rivers easily.

However, I know areas where the inlet is constantly changing and the chart is questionable. One has to look at the buoys AND read the water. There can be very little space between the shallow water and the buoy. I suppose the automation "vision" can detect which is port side vs starboard via the shapes. I have been in places where the channel is snaking around and figuring out which buoy is which is interesting. These situations would be challenging to program.

How does the automation work in the vast areas of the world where surveys are from the day's of sail, if surveyed at all? Hopefully, the program just fails to provide a route or pops up a big warning message multiple times.

The various phones we have used over the years for navigation have caused us to travel to "interesting" areas. Some not really safe to be in, and once, took us to the front gate of a major army base. We were not alone, plenty of other vehicles were being turned around, and the MPs were NOT amused. Even our latest and greatest phone has sent me on routes that make no sense or are not the best route and this is simple road navigation that does not take into account draft, air draft, tide, currents, and weather that vessel automation would need to take into account.

I can see auto routing be useful as a suggestion but one would really need to look at the route closely. I read about coastal areas in Ireland and many of the areas will have places one should not enter, or pass through, without local knowledge. The chart looks "ok" but.... For example, one entrance has two rock islands/out crops named Adam and Eve. The local information is to avoid Adam and hug Eve. The charts are such that one wonders if one should pass starboard or to port. How is the automation going to know to avoid Adam and hug Eve and one which side?

Even with local knowledge, fishing trawlers using the port have sunk after hitting rocks.

What happens with the automation when the data is is using is wrong or non existent? Much of navigation is SMOP(Simple Matter Of Programming) but the programming has to be perfect and it has to operate on perfect data. Eventually the programming will get really good but the data? Which weather forecast should the program use? Which chart?

Need to think about current AI capabilities. Most “smart” cars can and do deal with a changing environment. Between multiple radars and sonars they currently are aware of and avoid pedestrians and bicycle riders. Add in forward looking sonar, side scanning and FLIR current autonomous vehicles on land and on or in the water can do a better job than you at present. That remains especially true when the chart is not representative of the physical reality. Have had opportunity to talk with an attack sub master and a senior tank officer. Wife’s nephew commanded a battalion of unmanned attack drones. Have come to accept that in multiple modalities air, land and sea your concerns have been dealt with. This, of course, is not cost effective for us but the tech is already here.these systems are not dependent upon outside input nor fixed data sets but rather are collecting data in real time and responding to that. You need to remember GIGO. Sensor tech has advanced to a remarkable degree.

 

[psneeld]

Even Garmin navigators for RVs make the recommendation whether a road is suitible or not for your RV based on many parameters, it flat out tells you of things that are show stoppers like low clearance, weight, etc. So vessel routing may follow suit.

Ships of certain size have been plying certain routs that may have limited survey info but reality says they are generally safe. New tech lets anyones sounder send depth info to a master database....that will help too. So my guess is that auto routing will used only certified data or "historic" safe
routes to plan a hard route, if you ask for a destination without those parameters it may not use a route of even best data unless you override it.

 

[Jklotz]

I'm sure you guys have heard about the container ship taking out a major bridge in Baltimore last night, yes? I'm guessing if it were autonomous, it probably wouldn't have taken out a bridge. Robots don't get tired at 1:30am. Just saying....

https://www.cnn.com/2024/03/26/us/baltimore-bridge-collapse-what-we-know/index.html

The article said the ship lost power during it's entry. I wonder if the robot could have prevented it or reacted more quickly?

 

[Mambo42]

I'm sure you guys have heard about the container ship taking out a major bridge in Baltimore last night, yes? I'm guessing if it were autonomous, it probably wouldn't have taken out a bridge. Robots don't get tired at 1:30am. Just saying....

https://www.cnn.com/2024/03/26/us/baltimore-bridge-collapse-what-we-know/index.html

The article said the ship lost power during it's entry. I wonder if the robot could have prevented it or reacted more quickly?

Robots need power and that was what this ship was lacking. We have not heard from the NTSB what caused this accident, but am pretty sure having no power would have not been solvable for autonomous steering either.

 

[Hippocampus]

We are currently at war with Russia and China in the cyber space. I’m more concerned about a concerted hack or a EMF burst than a localized power failure in a single vehicle or vessel. Baltimore is tragic. Perhaps we’ll learn how to decrease the odds of recurrence of similar events. But Baltimore was a single event. Still as we shift to autonomous believe redundancies to create hardening in case of failure need be in place. At present boats, ships and airplanes “fly by wire”. My last sailboat had direct mechanical linkage for steering, a separate arm for the AP, and an emergency tiller that could be inserted. It was backed up by an independent wind vane that could be easily converted to a steering rudder (Hydrovane). My current boat has hydraulics. The AP operates through the hydraulics. There’s no emergency tiller. A single point failure and there’s no steering. Look at current full displacement boats from around 18 meters on up. Many have commercial grade fly by wire steering. Deemed very reliable and has an extensive track record of safety. But as Baltimore points out not foolproof. As we transition we create new potential vulnerabilities. To that point I agree with you.

 

[Invincible Summer]

I think that those who are interested in autonomous anything need to ask themselves a simple question. Have you ever, in your entire life, encountered a piece of electrical equipment that you would bet your life on?


Most people, if responding honestly, would answer no. Some who I have posed that question to have suggested things like a pacemaker. OK, I suppose ... but, it's a pacemaker or die ... so, not much of an option.


I'm almost 60. I've never owned a piece of electrical equipment that I would bet my life on ... I don't care how many redundant systems are in place.


Anyway, my .02

 

[SteveK]

I think that those who are interested in autonomous anything need to ask themselves a simple question. Have you ever, in your entire life, encountered a piece of electrical equipment that you would bet your life on?


Most people, if responding honestly, would answer no. Some who I have posed that question to have suggested things like a pacemaker. OK, I suppose ... but, it's a pacemaker or die ... so, not much of an option.


I'm almost 60. I've never owned a piece of electrical equipment that I would bet my life on ... I don't care how many redundant systems are in place.


Anyway, my .02

Did you mean to say electronic, because we have electrical equipment that saves lives.

 

[Portage_Bay]

Like it or not, and I don't, electricity and electronics already have my life in their hands. Modern cars for one example are drive by wire, no hard linkage to steering, brakes and engine. Autonomy is not about the connections, it's about the decision making. Our 2 yr old Subaru already makes some of the decisions. Reducing the throttle or hitting the brakes if need be without our intervention. Doing it's best to keep us in the lane. Even when I'm driving the old mostly mechanical pickup my life is in the hands of the "smart" components of the vehicles around me. Autonomy is coming and it's coming fast. I'm uncomfortable with it. But there isn't anything I can do about it.

I think that those who are interested in autonomous anything need to ask themselves a simple question. Have you ever, in your entire life, encountered a piece of electrical equipment that you would bet your life on?


Most people, if responding honestly, would answer no. Some who I have posed that question to have suggested things like a pacemaker. OK, I suppose ... but, it's a pacemaker or die ... so, not much of an option.


I'm almost 60. I've never owned a piece of electrical equipment that I would bet my life on ... I don't care how many redundant systems are in place.


Anyway, my .02

 

[psneeld]

I think that those who are interested in autonomous anything need to ask themselves a simple question. Have you ever, in your entire life, encountered a piece of electrical equipment that you would bet your life on?


Most people, if responding honestly, would answer no. Some who I have posed that question to have suggested things like a pacemaker. OK, I suppose ... but, it's a pacemaker or die ... so, not much of an option.


I'm almost 60. I've never owned a piece of electrical equipment that I would bet my life on ... I don't care how many redundant systems are in place.


Anyway, my .02

Air travel. Train Travel. Hospital equipment. A good blender.