Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

[Exile]

Originally Posted by ramblinrod

OK enough with the strawmen already.

1. I stated earlier in this thread that commercial vessels should be excepted to have a higher level of competence aboard due to the size, weight, speed and corresponding risk to other boaters, and due to the number of crew and training.

2. I stated earlier in this thread and have always held that Colregs apply to all boaters equally.

3. I never said this report held anyone to different standards (though commercial vessels are held to different standards than rec vessels.)

4. What I did say is I feel the ship should be held to higher percentage of blame, because those boys should be well enough trained and skilled not to run down a sailboat despite a faulty trilight and no radar reflector.

In other words, Colregs apply to both, additional regulations apply to commercial vessels, and a higher standard of competence should be expected on a big, fully crewed ship vs a little short-handed sailboat.

And any professional seaman that doesn't get that has a major problem IMHO.

You left out where you also said, or strongly suggested perhaps, that in a crossing scenario between a recreational vessel (you) and a commercial ship, the recreational vessel can assume the ship will take appropriate avoidance action because, unlike recreational vessels, the ship is required to have radar, AIS, and professional crew training, etc., and is therefore held to higher standards. Hopefully the two recently discussed cases of recreational vessels being run down by big ships will disabuse you of making such assumptions in the future.

If I misunderstood or mischaracterized what you said or intended to say, then it's my mistake and I would happily welcome corrections. It certainly isn't intentional and therefore worthy of your ubiquitous "strawman" characterization. On the other hand, if you are changing facts to suit your argument, then that is the very defn. of a "strawman" tactic that does nothing to advance an otherwise interesting discussion (at times).

[Dockhead]

Originally Posted by ramblinrod

. . ..

4. What I did say is I feel the ship should be held to higher percentage of blame, because those boys should be well enough trained and skilled not to run down a sailboat despite a faulty trilight and no radar reflector.

In other words, Colregs apply to both, additional regulations apply to commercial vessels, and a higher standard of competence should be expected on a big, fully crewed ship vs a little short-handed sailboat..

The law does not say that. Read any accident case concerning a collision between a rec vessel and a ship, including the one cited earlier in this thread, and you will see that the same standards are applied. If you get into a collision and you make a mistake, you are not held to a lower standard, and blame is apportioned accordingly.

Originally Posted by ramblinrod

. And any professional seaman that doesn't get that has a major problem IMHO.

There's nothing to "get", since the proposition itself is false. But FWIW -- even if it were true, I don't think it's only the pro seaman who has a "major problem". If there's a collision, on the contrary, it is the rec sailor who is more likely to be killed. It would be cold comfort that he pro was more to blame just because he's a pro, even if it were true.


I think the moral of the story is that if you want to mix it up with ships in busy waters, you need to have decent skills, not so much lower than what you would need on a ship. You just have no business being out there, if you think you can just bumble along and rely on ships to do all the heavy lifting.

Discussions like this make me think again that we really ought to be required to have licenses, after proving that we are competent in the several essential skill sets.


And if you talk to commercial mariners, their #1 complaint about us (and #2 and #3, actually) is precisely this lack of knowledge and skill. Most of us rely on seat of the pants and bare eyeballs, and maneuver spontaneously, according to a subjective feeling of safety or danger. We are not aware of potential collisions far enough away. When we take action, it is erratic and incorrectly calculated. We often mistake -- because we are relying on bare eyes and gut instinct -- a safe crossing, which they have set up carefully, for a dangerous situation, and we maneuver incorrectly and create a risk where none existed. They want us to know the Rules and follow them, applying a normal level of skill, like what they use themselves.

If you want to be a casual sailor and you don't want to work up the necessary skills to participate in real traffic, then you need to stay away from areas where you may get into a risk of collision situation with ships. In principle, this a realistic approach, too -- many sailors, maybe most, sail in waters where you can just stay out of the channels and fairways. But if you do get into a crossing situation in open water, it is not right to think that you can just rely on the ship to deal with it. You shouldn't be there in the first place without the skills and knowledge to do it right.

[micah719]

Originally Posted by Dockhead

Discussions like this make me think again that we really ought to be required to have licenses, after proving that we are competent in the several essential skill sets.

The danger here is the assumption of license = competence. You have it the right way around, but many won't. Also, once Government gets its hands on it, they'll succeed in the usual expensive fashion...

[evm1024]

Originally Posted by Dockhead

The law does not say that. Read any accident case concerning a collision between a rec vessel and a ship, including the one cited earlier in this thread, and you will see that the same standards are applied. If you get into a collision and you make a mistake, you are not held to a lower standard, and blame is apportioned accordingly.



There's nothing to "get", since the proposition itself is false. But FWIW -- even if it were true, I don't think it's only the pro seaman who has a "major problem". If there's a collision, on the contrary, it is the rec sailor who is more likely to be killed. It would be cold comfort that he pro was more to blame just because he's a pro, even if it were true.

SNIP (cut out some real wisdom)

I'm wondering what payout was done as a result of the collision. Given that there were no deaths, injuries of significance or significant costs (at least in terms of megabucks) I would not expect it to go to trial. It may have been much less costly for the shipping company to pay for repairs to the sailboat. Both in terms of actual money and in bad press.

With that said it is clear from the conclusions in the report that the proximate cause of the collision was:

1(a): The yacht GRUNTER, in contravention of Rule 25, did not exhibit "sidelights" as defined in Rule 21, namely "a red light on the port
side".

Additionally to 1(a) is finding 5 which I consider a root of the proximate cause.

GRUNTER could be considered unseaworthy in terms of Section 207 of
the Navigation Act 1912 in that the lack of a port sidelight rendered
it unfit to encounter an ordinary peril of the voyage, namely
collision risk.

Whether this was negligence or gross negligence is up for debate (given that there was no trial that I know of). If Gunther went to sea knowing that their lights were faulty then I would tend to consider Gunther's conduct gross negligence.

N.B. The improper display of navigational lights was the first stated cause of the collision. If you go to sea with your deck level nav lights and tricolor on at the same time then you will be displaying improper navigational lights. You may quibble whether too few, too many, wrong color light etc. But that is all it, is a quibble. Food for thought.

[evm1024]

Originally Posted by Dockhead

Discussions like this make me think again that we really ought to be required to have licenses, after proving that we are competent in the several essential skill sets.

I hear what you are saying - But you should come drive in my city. Everyone (well most of) passed a test and has a drivers license.

[Kelkara]

Originally Posted by Uricanejack

I presume you have used a head up unstabilised radar display as the template for the example.

The display is a stable north-up radar ... just used as a way of graphically showing how the situation unfolds. Not implying that radar is the method of observation.

Included a Yaw effect on the target vessel of
10 deg? so the target appears to have spread aprox 10 deg.

The display is supposed to show the uncertainty in predictions of the future positions of the ship from the perspective of the sailboat. Not the actual course of the ship. With more information gathered by continued observation of the ship the uncertainty of its course and speed will be reduced ... but by then it will be closer, and there is less time left to make your decision.

This would represent a level of uncertainty trying to predict the CPA, TCPA, Aspect, true course, and speed.of the approaching vessel in restricted visibility. The scenario was for vessels in sight of each other.

No my example is not for restricted visibility, the premise is "a ship is sighted 5 miles away ... what is the risk of collision, and should I make a manoeuvre?" The display shows how the situation could unfold from this initial sighting alone.

Much of the uncertainty could be eliminated by visual observation of the approaching vessel. Particularly a compass bearing. Or just eyeballing it against a station or background.

For a sailboat using just a hand compass and eyeballs, this is actually all the info we have: a bearing (with an associated uncertainty) and a very crude estimate of distance (large uncertainty). Estimating the speed and heading of the ship from this info alone is really not easy unless the bearing is changing significantly, in which case there is no risk of collision and we don't care (so maybe I should have used a much larger uncertainty for this).

Typically commercial ships have fairly accurate course and speed. With little variation of either unless intended. While it is not possible to accurately predict the navigation plan or route of another vessel. It may alter course for some unknown reason.
Their course and speed is usually very predictable.

They know their course and speed with a high degree of confidence ... but using just a compass and eyeball how well can the sailboat estimate it? It is your understanding of their speed and course, not theirs, that will determine when/what action you should take.

The uncertainty comes form the observers own vessel not being able to hold a very accurate course and or speed.

This is another source of uncertainty.

Most of this uncertainty can be eliminated by observation by other means as well as radar. I good visibility.

I was actually thinking about the info I have on my own boat boat. My radar, which has no heading sensor, gives me a reasonable estimate of distance, from which I can estimate speed if I watch for long enough, but with no heading sensor it is useless for bearings. And my eyeballs and compass are good for bearings but poor for estimating distance and speed. So by combining my radar and compass observations I can see what is going on, but both still contain uncertainties.

Using radar alone ie in restricted visibility.
You need to increase your margin of error.
In other words act sooner and more decisively.

The questions I was trying to answer for myself are "how quickly after sighting a ship can I determine whether a safe crossing condition exists?" and "How long should I wait before taking action if I'm uncertain to the safety of the situation?" and "how drastic an action do I have to take if I make my decision to turn late?"

I was trying to set it up for a 0.5nm CPA, which most people here are suggesting is too close for comfort ... just to see how comfortable I'd be watching a ship approach on that course.

Good illustration.

Thanks, It was just a quick and dirty calculation, it wasn't obvious to me how to best incorporate all the uncertainty and display it, but I think this kind of shows some of it.

ps. If anyone wants to play with it or modify it I've attached the xls file. It's not exactly optimised programming so it runs a bit slow if you give it uncertainty on everything.

[ramblinrod]

Originally Posted by Dockhead

But if you do get into a crossing situation in open water, it is not right to think that you can just rely on the ship to deal with it. You shouldn't be there in the first place without the skills and knowledge to do it right.

Grandstanding and strawman together, woohoo. We're setting records here.

In over 1000 posts, please show me once where anyone suggested for one second, that one should "rely solely on the ship to deal with it" (other than one tongue in cheek comment by another poster that one could just set the autopilot, uncap a cold one, and the ships would move out of the way).

I believed that was a joke, didn't you?

Do you seriously think someone would post that and mean it?

Do you seriously believe that there is anyone in this thread who believes this?

[Dockhead]

Originally Posted by micah719

The danger here is the assumption of license = competence. You have it the right way around, but many won't. Also, once Government gets its hands on it, they'll succeed in the usual expensive fashion...

Well, you've got a point there for sure.

I have always liked the fact that we are little regulated, and that what happens to us out there is almost entirely in our own hands, a matter of individual responsibility.

But when so many of us think you just don't need to know anything, I start to wonder about my libertarian principles . . .

[Dockhead]

Originally Posted by ramblinrod

Grandstanding and strawman together, woohoo. We're setting records here.

In over 1000 posts, please show me once where anyone suggested for one second, that one should "rely solely on the ship to deal with it" (other than one tongue in cheek comment by another poster that one could just set the autopilot, uncap a cold one, and the ships would move out of the way).

I believed that was a joke, didn't you?

Do you seriously think someone would post that and mean it?

Do you seriously believe that there is anyone in this thread who believes this?

I think it's possible to take this down a notch and discuss it rationally.

I would also suggest that you might also stop calling any argument you don't like a "strawman". The accusation has been used so many times that it has been tuned out long ago by, I daresay, everyone reading this.

Wise men may disagree on this, and unlike you, I am not calling anyone a fool.

But my argument was addressed to the proposition, stated many times already, and repeated here, that we shouldn't be held to the same standards of competency. I will try not to repeat myself too much, but the law certainly does hold you to the same standards. That's one. And two, it is entirely possible for crews of small vessels to do collision avoidance competently. Many of us don't, but I know many rec sailors who do it to an entirely acceptable standard. It takes some effort, but you can't go to sea safely in a small boat (or have fun) without learning a few different skills to a fairly high level.

And lastly, another thing which I have said before, but which is worth repeating, to put all this in context -- different waters, require different skill levels and different techniques. If you're sailing a 26 foot boat in a lake, with no radar, so rarely or never sailing anywhere overnight or in trafficked open water, then you just don't need to understand a lot of stuff, which is essential for those of us sailing in other kinds of waters, and you may do just fine by eyeballing and dodging, using common sense to stay out of the fairways. But nobody should universalize his own experience, and insist that it applies to everyone. Dodging race boats around the cans is totally different from keeping out from under the bows of fast moving ships while crossing the English Channel, requiring totally difference concepts, skills, and techniques. Both require alertness and sharp visual watches, but that is about the only thing in common with those two situations.

[Dockhead]

Originally Posted by evm1024

I'm wondering what payout was done as a result of the collision. Given that there were no deaths, injuries of significance or significant costs (at least in terms of megabucks) I would not expect it to go to trial. It may have been much less costly for the shipping company to pay for repairs to the sailboat. Both in terms of actual money and in bad press.

With that said it is clear from the conclusions in the report that the proximate cause of the collision was:

1(a): The yacht GRUNTER, in contravention of Rule 25, did not exhibit "sidelights" as defined in Rule 21, namely "a red light on the port
side".

Additionally to 1(a) is finding 5 which I consider a root of the proximate cause.

GRUNTER could be considered unseaworthy in terms of Section 207 of
the Navigation Act 1912 in that the lack of a port sidelight rendered
it unfit to encounter an ordinary peril of the voyage, namely
collision risk.

Whether this was negligence or gross negligence is up for debate (given that there was no trial that I know of). If Gunther went to sea knowing that their lights were faulty then I would tend to consider Gunther's conduct gross negligence.

N.B. The improper display of navigational lights was the first stated cause of the collision. If you go to sea with your deck level nav lights and tricolor on at the same time then you will be displaying improper navigational lights. You may quibble whether too few, too many, wrong color light etc. But that is all it, is a quibble. Food for thought.

It's a very good thing, that no one was killed in that case.

I thought the judges did a very good job with that one. There were a number of different causes of the accident, as there always are, but the primary cause was the inability of the ship's bridge to see the yacht's aspect based on the nav lights displayed. Judging aspect from nav lights is the go-to technique for making a maneuvering decision, in the absence of good radar data and when it's too dark to see the other vessel directly. The ship couldn't see the right aspect because the wrong lights were displayed, and made the wrong move. Could they have figured it out if they had been keeping a better watch, observing from further away? Maybe -- but wrong information leading to wrong decision and then a collision is a pretty straightforward chain of causation.

As Uricanejack said, correctly in my opinion, legal liability, payouts, etc., are not really the main point -- the main point is safety. But this case should show pretty dramatically how safety can be compromised by showing wrong nav lights. Show the right nav lights, so that ships relying on those nav lights to make maneuvering decisions, get the best possible information. This should be pretty basic and obvious, isn't it?

You should also have a proper radar reflector, and best of all, transmit AIS. But let's not neglect the basics -- the pros are well trained in interpreting nav lights, and will often, maybe more often than we do, rely on them for understand how to maneuver. Radar reflectors don't always really work, and recreational AIS, especially the Mark 1 Class B, is not always reliable. I think pretty often nav lights are all they have to know how to deal with us. They are important.

[Jim Cate]

and recreational AIS, especially the Mark 1 Class B, is not always reliable.

Upon what are you basing this statement, DH? I was not aware of such a flaw.

Jim

[Paul Elliott]

Originally Posted by Jim Cate

Upon what are you basing this statement, DH? I was not aware of such a flaw.

Jim

Slower Class-B update rate? But that has no effect on the Class-B-carrying vessel to analyze the AIS broadcasts that the Class-A vessel is transmitting. A ship will see fewer reports from the Class-B vessel, but if this is a typical sailboat the slower updates don't matter much.

[evm1024]

Originally Posted by Jim Cate

Upon what are you basing this statement, DH? I was not aware of such a flaw.

Jim

I'll let Dockhead speak for himself.

Keep in mind that Class B AIS transmits speed and heading data every 30 seconds (every 3 minutes if going under 2 kts).

This presents some interesting positional/track errors (uncertainties if you will - not flaws).

It may be that for an extended period your class B AIS transmission slot syncs with the boat yaw thus giving a significantly inaccurate projected course on the ECDIS (chart plotter).

On my own plotter (raymarine and openCPN) the projected line tends to dance around. Most of the folks who crew for me tend to look at that projection and the CPA prediction.

I show them that in addition to the projection they need to look at the actual track breadcrumbs as a way to validate the projection.

[StuM]

Originally Posted by Uricanejack

You pointed out yourself its normal practice for these types of investigations not to find fault or blame. MAIB NTSB TSB follow this so people will speak freely.

A criminal case only answers one question Guilty or Not Guilty.

A civil case is about money. fear of liability may influence cooperate action. Even so a civil case is about proving liability not determining how to prevent a future occurance

Admiralty courts (federal district courts in the US) which adjudicate over maritiiime law and collisions at sea assign percentage liability (fault) to the parties involved. You may like to research "Proportional Damage Rule"
.

[Dockhead]

Originally Posted by evm1024

I'll let Dockhead speak for himself.

Keep in mind that Class B AIS transmits speed and heading data every 30 seconds (every 3 minutes if going under 2 kts).

This presents some interesting positional/track errors (uncertainties if you will - not flaws).

It may be that for an extended period your class B AIS transmission slot syncs with the boat yaw thus giving a significantly inaccurate projected course on the ECDIS (chart plotter).

On my own plotter (raymarine and openCPN) the projected line tends to dance around. Most of the folks who crew for me tend to look at that projection and the CPA prediction.

I show them that in addition to the projection they need to look at the actual track breadcrumbs as a way to validate the projection.

Yes, this. Low power, slow updates, dodgy installation, especially antenna and feedline, but also possible flaky splitter - AIS transponders on rec boats often are not effective, and it is possible to get into trouble assuming you can be seen at some given range, without testing it.

Remember that most ships want to detect potential problems at at least 10 miles out - it's really worthwhile checking to be sure you're visible from such ranges. They might even miss you altogether if you only pop up a couple miles off, if they are focusing on a longer horizon.

Also AIS transponder is no substitute for an effective radar reflector.