Isn’t that boat ALREADY towing something heavy 
I have used small drogues/tires when towing boats that would sheer all over the place on the towline. It settles them right down.
there are several projects working on it now, Wartsila doing one in Singapore.
Its seems to be to reduce the bridge workload.
Might just end up like autonomous cars and that wont really work until all the cars are talking to each other?
Ships have a head start as they broadcast AIS data and already have ARPA
AIS data especially being completely reliable all of the time.
1 Like
Of course! It’s digital.
I’ll show myself out.
1 Like
After reading about work going on in autonomous ship steering in high seas, it seems to me that one thing that is missing (or I am not finding) is a way of collecting all of the negative impacts on a ship from the sea it is encountering RIGHT NOW. That is, wherever the ship is on the slope of a big wave there are lots of things to consider:
- Angle along the beam of the ship (maybe called “pitch”) which can be pulling things loose
- Angle across the beam (maybe “roll”) moving things and affecting pumps
- Rate of movement of tilt and roll, as (in parametric rolling), g-forces and hard to stop
- Ship superstructure strain along the beam, like bending at the top of a large wave
- Impact strain, like on the bow as a wave impacts
- Force on the rudder, as this may limit steering and course keeping
- Current thrust, as the screws could be pulled out of the water, loosing thrust and control
- Total weight of water taken aboard, either bow or stern. May run off fast, but it does not help!
- Height in the water at several points, like if the nose is getting buried in the next wave
Are there big ones I am forgetting?
Once we have a way of a mechanical system to understand all of these and have a sense of “how serious is this moment NOW”, then it seems like the designer can start asking the question “you know how you are now, but given the choices you have made, can you see where you will be in 30 seconds?!”. This requires everything above PLUS situational awareness of how the sea is moving and how the vessel is moving and how those relate.
I think you would need both to see an autonomous ship making choices like going 2 points off heading relative to the waves and adjusting speed, because that is a maneuver that is reacting to the most recent collision (“that was a bad one”) and tweaks the collision angle and speed so the next one is less (“here is a smoother ride”). But it seems to me a machine FIRST needs to be able to say “that was a bad one”, and that takes a lot of sensing of what is happening to the ship NOW, that I don’t see being considered…
Thoughts?
Ship’s motions are roll, pitch, yaw and heave, sway, surge.
The other factor to consider with regards to encountering heavy weather aside from max allowable motions is the increased probability of failure over time.
For instance heavy pitching could lead to pounding or slamming, various items that would normally not be expected to fail will fail.
For example a fuel line connection that is inspected every 1000 hours may be caught if it loosened up under normal operations but might fail after 24 hrs of heavy pounding.
Does every ship have max allowable motions (amplitude or amplitude and velocity?) in all six degrees of freedom - roll, pitch, yaw and heave, sway, surge?
Is this information published with the ship’s technical specifications or something?
Thanks!
There is a limit to how far a ship can roll before it capsizes but in practice some reserve is required;
Thanks! These videos are really good! (Our defense dollars at work!)
One thing I did not bring up on my list of “status report” items is what are the eventual FAILURE MODES that are trying to be avoided. You already brought up two, but others have been mentioned:
- Capsizing (about the worst thing that can happen)
- External structural damage (taking on water, nose blown off of the ship, mast/radar down etc.)
- Internal structural damage (pumps fail, fuel lines rupture…)
- Cargo damage (stuff rolling around, probably not super important for autonomous navy ship, unless it is ammo!)
- Off Course (maybe not an emergency, but fuel wasted, possibly risking not having enough to get back, if blown far enough off course)
Others?
2 Likes
Bending (hog and sag). Over time a ship riding large waves (or improperly loaded/ballasted) can hog and sag so that it eventually weakens the ship and may lead to cracking, splitting or crumpling of horizontal plating or longitudinal piping. Continued failure leads to flooding, broken pipes or a ship splitting in half.
1 Like
Synchronous rolling must be avoided as well as parametric rolling. Wavelength/period and roll/pitch frequency should be monitored on a continuous basis. What helps considerably countering these phenomena is having stabilizers installed as well as anti-roll tanks. Anti-roll tanks are also advantageous for the gunnery department.
There also is the Arrow software tool for detecting and avoiding synchronous or parametric resonance.
Its all very simple, if you can feel it in the seat of your pants then a sensor can too.
Ships will have more sensors but once you collate all that data a computer will be able to make much better decisions than a captain with or without the same data.
Just imagine the office will be able to recreate the current environment the ship is in their test tank live and give the best feedback back to the vessel. ( vessels computer as there wont be any crew onboard)
Yes! I have been looking at how to model the relationship between ocean wave wavelength and the ship length. As you diagrammed, for a given boat and a given sea, there are specific angles and situations that need to be avoided. We are trying to work that into a kind of “heat map” of advisable and inadvisable options for the autonomous steering system, but it is very path and sea specific, so gets complicated!
MAW:
I figure there will be sensors to measure drafts of the ship? Many modern ships already have bow, midship (port and starboard) and stern draft sensors built in so the drafts underway can be known. Ours show the average draft for each sensor. I figure instead of showing a human the average a computer can watch the rise and fall of the bow and stern draft sensors and avoid a situation where the peaks or troughs occur simultaneously.
Or you could use fancy range finders (laser, radar, sonic, 3D vision) to look at the waves. No commercial ship would do that. $$$
All very well until they fail, which they will.
First ran into this about 20 years ago. A polar diagram based on wave height / period and vessel rolling period. The mariners that were using it said that it worked good for forecasting ship’s rolling before a course change.
1 Like
you can already get wave radar, its used by heavy lift DP vessels for working out the operating environment window.
https://www.offshoreenergytoday.com/amarcon-gets-octopus-order-from-pan-ocean/
1 Like
I see where that could work on a small vessel. On big ships, dragging used aircraft tires strung on chains would only help if the tires were still attached to airplanes. 
2 Likes