Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

[ramblinrod]

Originally Posted by Paul Elliott

Rod, this is not a useful comparison. I occasionally race, and according to the RRS (Racing Rules of Sailing) sometimes a miss *is* *not* as good as a mile. If you fail to give room at the mark and so impede a sailboat from it's "proper course" you have fouled that boat. A collision is not required.

When racing all the participants have agreed to abide by the RRS. We are all sailing boats of a similar size (+/- 50 ft or so), at similar speeds (+/- 5 kts), and close crossings are expected. Colllisions are seldom life-threatening.

That 600-ft ship has not agreed to follow the RRS. It is sailing according to COLREGS. It can't maneuver like a racing sailboat, and it is usually traveling *much* faster than you are. It isn't out there to have fun, but to safely carry passengers and cargo.

The relevant point is, when crossing in close quarters, one needs to keep their eyes on the vessel and away from the electronics.

If one can consistently duck the transom of a racing sailboat, perhaps going twice their speed, within a hairs breadth, they can certainly duck the transom of a ship at 4 times their speed, within 180 ft.

And even more to the point, the 1/4 nm CPA from the lead boat transom Dockhead proposed would be much more likely to get the sailboat skipper dead.

Case closed.

[transmitterdan]

Originally Posted by ramblinrod

If one can consistently duck the transom of a racing sailboat, perhaps going twice their speed, within a hairs breadth, they can certainly duck the transom of a ship at 4 times their speed, within 180 ft.

No, you cannot do this. You could show that my foregoing statement is wrong by doing a plot with both vessel's position, heading and speed every 5 seconds for 100 seconds (20 plots). Assume the big ship is "on rails" if you want but it won't change the outcome. You can never achieve the 180' distance you espouse.

[ramblinrod]

Originally Posted by Dockhead

I have explained in great detail why it is completely different, when the speed and size of the vessels are different, and especially when you have a large difference of speed like this.

You are applying a completely different set of parameters which cannot be applied to crossing with fast ships at sea. You can eyeball a crossing pretty well when two slow vessels are approaching each other at a decent angle. You cannot do it when you're dealing with a high speed, large vessel, approaching you from nearly your beam.

You are imagining mixing up with fast ships at sea like you jostle with race boats around the cans -- it doesn't work like that at all! It's a pity we don't have a simulator -- you would see it immediately. It can be shown with math -- starting with angles which are visible and invisible to the human eye at what distance, from which we can get what rate of bearing change is visible and what is not perceptible, then how accurately distances can be judged. From that we can show that you can't, for example, tell the difference between passing three cables ahead from passing three cables behind, from a mile out, at these speeds -- something I know from experience, but can also be demonstrated by math. This would be a really useful exercise, actually. I am working this week and don't have time for this, but maybe our math whizz Juho would like to have a crack at it.

I don't need a simulator.

I have crossed ships travelling greater than my speed many times.

For the given scenario, the 1/4 nm CPA from the stern of the lead ship your proposed is wrong.

Possibly DEAD WRONG!

A much closer CPA is required to maintain a "safe distance" ahead of the following ship.

There is no way out of this. No twisting of words, no plethora of strawman arguments, no lawyer tricks, no amount of deflection of attention from this error, will change this.

You were wrong.

Case Closed.

[evm1024]

Originally Posted by ramblinrod

I don't need a simulator.

I have crossed ships travelling greater than my speed many times.

For the given scenario, the 1/4 nm CPA from the stern of the lead ship your proposed is wrong.

Possibly DEAD WRONG!

A much closer CPA is required to maintain a "safe distance" ahead of the following ship.

There is no way out of this. No twisting of words, no plethora of strawman arguments, no lawyer tricks, no amount of deflection of attention from this error, will change this.

You were wrong.

Case Closed.

Case open.

Perhaps I mis understood you. Please clarify for me.

So you state that you have crossed between a line of ships 50 ships long going 4 times your speed many times?

Or did you say that you crossed between 2 or more ships in a line going 4 times your speed many times?

Or did you say that you crossed between 2 or more ships in a line going 4 times your speed at least once?

Or did you mean to say that you crossed behind a single ship going 4 times your speed many times?

Or did you mean to say that you crossed behind a single ship going faster than you at lease once?

[ramblinrod]

Originally Posted by Dockhead

I have explained in great detail why it is completely different, when the speed and size of the vessels are different, and especially when you have a large difference of speed like this.

You are applying a completely different set of parameters which cannot be applied to crossing with fast ships at sea. You can eyeball a crossing pretty well when two slow vessels are approaching each other at a decent angle. You cannot do it when you're dealing with a high speed, large vessel, approaching you from nearly your beam.

You are imagining mixing up with fast ships at sea like you jostle with race boats around the cans -- it doesn't work like that at all! It's a pity we don't have a simulator -- you would see it immediately. It can be shown with math -- starting with angles which are visible and invisible to the human eye at what distance, from which we can get what rate of bearing change is visible and what is not perceptible, then how accurately distances can be judged. From that we can show that you can't, for example, tell the difference between passing three cables ahead from passing three cables behind, from a mile out, at these speeds -- something I know from experience, but can also be demonstrated by math. This would be a really useful exercise, actually. I am working this week and don't have time for this, but maybe our math whizz Juho would like to have a crack at it.

I don't need a simulator.

I have crossed ships travelling greater than my speed many times.

I have ducked boats (and passed ahead of boats) travelling at a greater speed than I, many times.

For the given scenario, the 1/4 nm CPA from the stern of the lead ship you proposed is wrong.

Possibly DEAD WRONG!

A much closer CPA is required to maintain a "safe distance" ahead of the following ship.

I provided the calcs in post # 591, that prove this beyond any shadow of a doubt.

There is no way out of this. No twisting, no plethora of strawman arguments, no amount of deflection of attention from this error, will change this.

You were wrong.

Case Closed.

[ramblinrod]

Originally Posted by evm1024

Case open.

Perhaps I mis understood you. Please clarify for me.

So you state that you have crossed between a line of ships 50 ships long going 4 times your speed many times?

Or did you say that you crossed between 2 or more ships in a line going 4 times your speed many times?

Or did you say that you crossed between 2 or more ships in a line going 4 times your speed at least once?

Or did you mean to say that you crossed behind a single ship going 4 times your speed many times?

Or did you mean to say that you crossed behind a single ship going faster than you at lease once?

Argumentative your honour.

I believe my post was clear and concise.

Case still closed.

[evm1024]

Originally Posted by ramblinrod

I don't need a simulator.

I have crossed ships travelling greater than my speed many times.

For the given scenario, the 1/4 nm CPA from the stern of the lead ship your proposed is wrong.

Possibly DEAD WRONG!

A much closer CPA is required to maintain a "safe distance" ahead of the following ship.

There is no way out of this. No twisting of words, no plethora of strawman arguments, no lawyer tricks, no amount of deflection of attention from this error, will change this.

You were wrong.

Case Closed.

Others are not twisting words....

Others are not using strawmen and other fallacies....

Others are not using lawyer tricks - they only point out the LAW....

Others are not deflecting - shall we talk about racing?...

It appears that you are very right - nothing will change your mind....

You appear very invested in being right no matter what the facts....

[evm1024]

Originally Posted by ramblinrod

Argumentative your honour.

I believe my post was clear and concise.

Case still closed.

You did not answer my question. And it is a good question. You state that your solution is based on your experience. You state your experience.

But to me your statement of experience is ambiguous so I ask for clarification.

Please clarify your experience in crossing ships.

[Paul Elliott]

Of course it's *possible* to pass behind a fast ship with an arbitrarily tight CPA. You can plot this out on paper, choosing the speeds and angles to shave the crossing as close as you like.

But that's on paper.

In the real world, the uncertainties in course and speed, and the imprecision of eyeball piloting make this unsafe, and likely to fail in a very bad way. Thus the "Cone of Uncertainty". Hell, at 150 ft behind the ship the prop-wash and turbulence would probably spin my boat in circles. That's way too close.

[Dockhead]

Originally Posted by evm1024

You did not answer my question. And it is a good question. You state that your solution is based on your experience. You state your experience.

But to me your statement of experience is ambiguous so I ask for clarification.

Please clarify your experience in crossing ships.

What we have here is a total clash of visualization. It's clear now, after the racing example.

Rod has an instinctive feel (I'm sure he's an excellent racer) for how to dodge around other boats in a race. He is absolutely sure -- not a doubt in his mind -- that avoiding ships is the same thing, just scaled up. "If I can dodge a race boat by inches I can sure dodge a big ship by 180 feet", which is utterly fallacious, but this is what the visualization is based on, and there is no way to change anyone's mind without a simulator.

Why? Because all the geometry has been explained in great detail, but you have to be willing to take it on board and make the leap to a different visualization. If you insist on sticking to the way you imagine it, then there's nothing anyone can say.

And I have to admit that I wouldn't understand any of this myself, if I hadn't had a concrete epiphanal experience some years ago which taught me the hard way. A crossing with a fast ship in the Channel where he appeared somewhere near my beam, and I had no idea which way to turn. I just didn't know what to do. You couldn't see it, no matter how sharp your eye was. I was amazed, because I had previously thought that surely you could always see at least whether you're passing ahead or behind, if not precisely by how much. This was years ago, before AIS. That made me question my previous internal model of how these crossings work and forced me to learn to do it the pro way.

If Juho will do some math to show what you can and what you cannot see -- that would be terrific. But otherwise I think we won't really get anywhere further without a simulator.

[ramblinrod]

Originally Posted by transmitterdan

No, you cannot do this. You could show that my foregoing statement is wrong by doing a plot with both vessel's position, heading and speed every 5 seconds for 100 seconds (20 plots). Assume the big ship is "on rails" if you want but it won't change the outcome. You can never achieve the 180' distance you espouse.

Of course it is possible to achieve a 180 ft CPA from the ships port stern transom in this scenario.

No plots needed at all.

One could set up a 0.1" CPA (may be a little turbulence ;-), a 180 ft CPA, a 1/4nm CPA, a - 1.0 ft CPA (smack the lead vessel) or a 3 cable CPA (smack the following vessel).

Interestingly enough, Dockhead's solution was a 2-3 cable CPA, (I picked the value in the middle of this range for calc purposes at 1/4 nm). If it was actually 3 cables, he would most definitely be smacked by the following vessel, no question about it.

That's the geometry and it is undeniable.

[Dockhead]

Originally Posted by ramblinrod

Of course it is possible to achieve a 180 ft CPA from the ships port stern transom in this scenario.

No plots needed at all.

One could set up a 0.1" CPA (may be a little turbulence ;-), a 180 ft CPA, a 1/4nm CPA, a - 1.0 ft CPA (smack the lead vessel) or a 3 cable CPA (smack the following vessel).

Interestingly enough, Dockhead's solution was a 2-3 cable CPA, (I picked the value in the middle of this range for calc purposes at 1/4 nm). If it was actually 3 cables, he would most definitely be smacked by the following vessel, no question about it.

That's the geometry and it is undeniable.

Rod, how do you "set up" a 0.1" CPA? Or a 180 foot CPA? You totally ignore the limitations of the information you have to work with, and this is a fatal mistake.

Do you really think you can eyeball THIS:





That's you, sailing at full speed directly towards the side of the ship, in the exact place which will take you 180 feet from his quarter. You have just barely missed getting run down under his bow, with CPA from his bow of less than a cable.

Now you're going to eyeball and correct any error in course?

If you can't visualize this, then there is nothing we can do with you.


You go on and on about MY solution -- which will give you about 3 cables from the bow of the following ship. But your solution requires a <1 cable CPA with the bow of the ship which is ahead. Your visualization totally fails to take account of the danger of getting run down by the ship AHEAD -- because you assume you are just sailing around somewhere near his side, like in a sailboat race. It seems to me that even though we drew it for you, you still don't understand that as you approach this crossing, the ship AHEAD will be running right at you, within 2 points of your beam. It's obvious that you don't know what that looks like. It's not pretty.

[ramblinrod]

Originally Posted by Dockhead

What we have here is a total clash of visualization. It's clear now, after the racing example.

Rod has an instinctive feel (I'm sure he's an excellent racer) for how to dodge around other boats in a race. He is absolutely sure -- not a doubt in his mind -- that avoiding ships is the same thing, just scaled up. "If I can dodge a race boat by inches I can sure dodge a big ship by 180 feet", which is utterly fallacious, but this is what the visualization is based on, and there is no way to change anyone's mind without a simulator.

Why? Because all the geometry has been explained in great detail, but you have to be willing to take it on board and make the leap to a different visualization. If you insist on sticking to the way you imagine it, then there's nothing anyone can say.

And I have to admit that I wouldn't understand any of this myself, if I hadn't had a concrete epiphanal experience some years ago which taught me the hard way. A crossing with a fast ship in the Channel where he appeared somewhere near my beam, and I had no idea which way to turn. I just didn't know what to do. You couldn't see it, no matter how sharp your eye was. I was amazed, because I had previously thought that surely you could always see at least whether you're passing ahead or behind, if not precisely by how much. This was years ago, before AIS. That made me question my previous internal model of how these crossings work and forced me to learn to do it the pro way.

If Juho will do some math to show what you can and what you cannot see -- that would be terrific. But otherwise I think we won't really get anywhere further without a simulator.

Nope. No clash of visualization. No simulator needed.

Of course the 180 ft CPA is going to put me in a closer relative position to the ship as it approaches. That is a given. No argument.

Whether one has set up a 180 ft CPA (as I proposed) or a 2-3 cable CPA (as you proposed) there will be an acute angle between the sailboat and the ship, slightly less acute for your proposal, but still acute.

But your proposal, of possibly a 3 cable CPA from the lead ship, will most definitely get you hit. No question about. The set-up is flawed from the get go even if executed perfectly.

You were wrong, twist in the wind all you wish, but the calcs I provided in post # 591 prove it beyond a shadow of a doubt.

[ramblinrod]

Originally Posted by Dockhead

Rod, how do you "set up" a 0.1" CPA? Or a 180 foot CPA? You totally ignore the limitations of the information you have to work with, and this is a fatal mistake.

Do you really think you can eyeball THIS:


Attachment 158253


That's you, sailing at full speed directly towards the side of the ship, in the exact place which will take you 180 feet from his quarter. You have just barely missed getting run down under his bow, with CPA from his bow of less than a cable.

Now you're going to eyeball and correct any error in course?

If you can't visualize this, then there is nothing we can do with you.


You go on and on about MY solution -- which will give you about 3 cables from the bow of the following ship. But your solution requires a <1 cable CPA with the bow of the ship which is ahead. Your visualization totally fails to take account of the danger of getting run down by the ship AHEAD -- because you assume you are just sailing around somewhere near his side, like in a sailboat race. It seems to me that even though we drew it for you, you still don't understand that as you approach this crossing, the ship AHEAD will be running right at you, within 2 points of your beam. It's obvious that you don't know what that looks like. It's not pretty.

Irrelevant your honour, the opposition is claiming they can pass the stern at 3 cables and be safe. (Which they can't they'll be run down by the following ship as proven in the calcs in my post # 591.)

But yes, your plot shows that I will be 10 boat lengths (mine) away, when abeam of the ship.

And yes, I can make that 180 ft CPA if I wished. (Personally I wouldn't do it, unless there was some life or death emergency to cross the convoy now, but I could if I had to, no question, as long the ships hold course and speed).

With the 3 cable CPA you proposed, it most likely be death, definitely a collision no doubt about it, if the ships held course and speed.

[evm1024]

Originally Posted by Dockhead

What we have here is a total clash of visualization. It's clear now, after the racing example.

Rod has an instinctive feel (I'm sure he's an excellent racer) for how to dodge around other boats in a race. He is absolutely sure -- not a doubt in his mind -- that avoiding ships is the same thing, just scaled up. "If I can dodge a race boat by inches I can sure dodge a big ship by 180 feet", which is utterly fallacious, but this is what the visualization is based on, and there is no way to change anyone's mind without a simulator.

Why? Because all the geometry has been explained in great detail, but you have to be willing to take it on board and make the leap to a different visualization. If you insist on sticking to the way you imagine it, then there's nothing anyone can say.

And I have to admit that I wouldn't understand any of this myself, if I hadn't had a concrete epiphanal experience some years ago which taught me the hard way. A crossing with a fast ship in the Channel where he appeared somewhere near my beam, and I had no idea which way to turn. I just didn't know what to do. You couldn't see it, no matter how sharp your eye was. I was amazed, because I had previously thought that surely you could always see at least whether you're passing ahead or behind, if not precisely by how much. This was years ago, before AIS. That made me question my previous internal model of how these crossings work and forced me to learn to do it the pro way.

If Juho will do some math to show what you can and what you cannot see -- that would be terrific. But otherwise I think we won't really get anywhere further without a simulator.

I think you are right.

I too have had one of those moments where the ship was closing fast and it looked huge 1/4 mile away. Was it going straight and going to run me over or was it going to turn to port and head for Victoria Harbor? I had no indications and even with the USCG intervening on the VHF we could not get an answer from the ship.

That was just one ship.

Rod is very heavily invested in being right. It appears to me that he must prove that crossing behind a ship with a 180' CPA is nothing (if you have big enough balls?) and the right way to do it in the BS example we are given.

You are right that it appears that he cannot visualize what a line of ships doing 20 kts 1 nm apart would look like.

10 miles out you are not even going to see them. Heck they form a line over 66 nm long.

At 5 miles you might see them from the deck of a sailboat. The would look like a string of pearls with each pearl (ship) about 3.6 degrees apart from each other directly ahead of you. Let the convergent lines take care of themselves.

At 2 miles of course you would see them. They would be about 32 degrees apart and covering that 32 degrees in 3m 45s.

At 1 mile they would be 51 degrees apart when directly ahead and covering that distance in the same 3m 45s. Just flying by....

Let me repeat that Flying By

At 1 mile out you will not be crossing under the stern of the ship directly ahead of you. It will be long gone. You will not be passing under the stern of the next ship (2nd) or the next (3rd) or the next (4th)....

You will be lining up to pass under the stern of the 5th ship. You are a little too late for the 4th ship and a bunch too early for the 5th. So you have to fudge your speed or angle to a little faster to make your 180 CPA on the 4th ship or slow a bit to make the 5th.

Is there any doubt folks how this crossing plays out?

We are debating a bogus scenario designed to bolster the arrogance and ego of someone who does not have the experience to understand what the scenario is like and cannot visualize the reality of the scenario.

In the meantime we as a group are getting something out of this, at least for those who can visualize the scenario. All the while we are missing out on a better understanding of cones of uncertainty and how they apply to night time or low visibility crossings.

- You are in a maze of twisty little passages all alike -