Collision Avoidance, Cones of Uncertainty, and Appropriate CPA

[jackdale]

Originally Posted by Seaworthy Lass

In all this discussion I have not heard what people consider is the most reliable tool for estimating if a risk of collision exists. Or to put is another way, what they are most comfortable depending on.

Personally, I rely on a hand bearing compass. It is always around my neck during watches. I can keep track of 2-3 boats fairly easily from memory. Any more, then a pen, paper and often headtorch is needed.

What do others rely on the most?

SWL

I mentioned the need for a hand bearing compass in an earlier post. I also have one available in the cockpit. I ensure that my Iris 50 gets exposed to light to ensure that it fluoresces, a necessity at night. During the day I use my old trusty mini Morin, which is not longer suitable for night use.

Rule 7 requires the use of a HBC to determine the risk of collision.

(d) In determining if risk of collision exists the following considerations shall be among those taken into account.

(i) such risk shall be deemed to exist it the compass bearing of an approaching vessel does not appreciably change.

(ii) such risk may sometimes exist even when an appreciable bearing change is evident, particularly when approaching a very large vessel or a tow or when approaching a vessel at close range.

[TJ D]

Originally Posted by Seaworthy Lass

In all this discussion I have not heard what people consider is the most reliable tool for estimating if a risk of collision exists. Or to put is another way, what they are most comfortable depending on.

Personally, I rely on a hand bearing compass. It is always around my neck during watches. I can keep track of 2-3 boats fairly easily from memory. Any more, then a pen, paper and often headtorch is needed.

What do others rely on the most?

SWL

I used to be HBC and radar (stabilized, north up-always), but our AIS has taken over the lion's share of this work now.
We have the Vesper Vision, which has a great vector display (repeated on the plotter), and it's proven to be really great for us.

Of course, we don't try to cut 180' from anyone. The HBC still is within reach, and gets used for non-ais traffic, but for guys who are broadcasting, especially class A. We've found the AIS to be unbeatable.

I view AIS as having made the biggest contribution to maritime safety since GPS came on the scene. It's just important to remember not to rely exclusively on it.

[Paul Elliott]

Originally Posted by Seaworthy Lass

IWhat do others rely on the most?

AIS and hand bearing compass, and radar if it's foggy. But I've never had a really close crossing situation, and no crossings that required any more than standard practice.

The worst encounter we've had was returning to San Francisco from Hawaii, as we were approaching the Golden Gate Bridge. We had passed Pt. Bonita and were under power, between the edge of the shipping lanes and the rocky cliffs. A big incoming freighter steamed past us, with perhaps 100 ft of clearance (we couldn't get any closer to shore), and his wake really tossed us around. There was no damage, but it was rougher than anything we had encountered in the previous seventeen day passage. The ship was going fast, we were going slow, and the ship was completely in control.

[Dockhead]

Originally Posted by Seaworthy Lass

In all this discussion I have not heard what people consider is the most reliable tool for estimating if a risk of collision exists. Or to put is another way, what they are most comfortable depending on.

Personally, I rely on a hand bearing compass. It is always around my neck during watches. I can keep track of 2-3 boats fairly easily from memory. Any more, then a pen, paper and often headtorch is needed.

What do others rely on the most?

SWL

Good question.

I always have a HBC in the cockpit. It's a great tool, but it is far less accurate than AIS and it takes some time to form a picture, since you need to take a succession of bearings.

So I tend not to use the HBC much for this purpose unless I'm dealing with a target not transmitting AIS and I don't have any particular time pressure. Sometimes I do it just to keep my skills up.

Otherwise, I tend to use the Mark 1 Eyeball and a stanchion to instantly recognize ships which are passing far ahead or behind and so not requiring deeper analysis. Then I don't need to get up to look at the plotter screen. Anything which needs deeper analysis, I get up and look at AIS first. This gives an instant calculation of CPA and TCPA and will then tell you, with regard to many targets which look dodgy by bare eyes, that they are actually safe. And if they are not safe, then you can start your collision avoidance procedure.

For targets which are not transmitting AIS, I'll use radar. My particular radar set does carp MARPA, so I generally don't bother with that. Normally I will set the EBL and watch that. If the target is walking down the EBL, then I assume a collision course and start collision avoidance procedures. This is a lot less accurate than AIS, but it's good enough if you take action early enough. For more accuracy, I might do an actual radar plot, but that's rarely necessary. You don't need really accurate data if you take action early enough and alter course or speed to pass well behind. You do need accurate data if you're trying to thread through traffic or otherwise need to pass fairly close.

You do also need accurate data if you are stand-on and can't quite tell whether the other vessel has made an adequate maneuver or not. This is where AIS really comes into its own, because the typical 1 mile CPA for North Sea or Channel is hard to spot -- can be a quite slowly changing bearing. You might need to do a plot to be sure, and if you don't have time or resources or don't know how to do it, that means that you might have to make unnecessary maneuvers to be sure of a safe crossing. But EBL and radar might be enough -- it's faster and better than HBC.

And if you have a radar set with decent MARPA (Furuno?), you could use that in lieu of a hand plot.

[micah719]

I had to look up some of the radar acronyms, so if anyone else needs them....

Article Details

Would it be a good idea to have a plexiglass sheet, a grease pencil and a rag handy at the helm to jot down info? You could have a bearing grid marked on the back, and a clock/kitchen timer, calculator and handy formulae to help calculations. One could get fancy and make one of these:

Submarine Attack Course Finder Mark I Model 3 Manual

No batteries required!

[jackdale]

Originally Posted by Dockhead

You do also need accurate data if you are stand-on and can't quite tell whether the other vessel has made an adequate maneuver or not. This is where AIS really comes into its own, because the typical 1 mile CPA for North Sea or Channel is hard to spot -- can be a quite slowly changing bearing.

I had a situation in 2015 in which the AIS triangle and COG vector line were showing me on a collision course with another vessel. Looking at the vessel through binoculars it was clear by the lights that he was crossing my bow at a very safe distance. This was verified by a VHF conversation. We never did determine why the AIS presented itself in this manner.

[jackdale]

Originally Posted by Paul Elliott

AIS and hand bearing compass, and radar if it's foggy. But I've never had a really close crossing situation, and no crossings that required any more than standard practice.

Radar (orientated heading up is my preference) is great. Setting the EBL on a target is quick and easy and a somewhat less fiddly that trying to acquire a target on MARPA.

[Dockhead]

Originally Posted by jackdale

I had a situation in 2015 in which the AIS triangle and COG vector line were showing me on a collision course with another vessel. Looking at the vessel through binoculars it was clear by the lights that he was crossing my bow at a very safe distance. This was verified by a VHF conversation. We never did determine why the AIS presented itself in this manner.

A textbook example of why you should never rely on any one source of data. Imagine if it had been the opposite -- AIS told you you were passing with a mile CPA but you were actually on a collision course.

AIS is very reliable -- I've never seen such a case like that, but where life and limb are at stake, one should ALWAYS verify with your eyeballs and other sources of data.

[Paul Elliott]

Back around 2003 I was sailing towards the US west coast and had a westbound freighter approaching with an AIS position and track one mile south of ours. I called up to warn the on-watch crew, then stuck my head out the hatch and found the ship, one mile *north* of us.

I did some research when I got home, and discovered that there were ship AIS installations where a GPS position was fed into a chartplotter, where the GPS data was adjusted to match the chart datum. This "corrected" position was then sent to the AIS transponder, in some cases putting the ship's AIS position off by as much as a couple of miles. The USCG had put out a bulletin about the problem, and while I've not seen this problem since, there could still be some problem installations out there.

[Uricanejack]

Originally Posted by Seaworthy Lass

In all this discussion I have not heard what people consider is the most reliable tool for estimating if a risk of collision exists. Or to put is another way, what they are most comfortable depending on.

Personally, I rely on a hand bearing compass. It is always around my neck during watches. I can keep track of 2-3 boats fairly easily from memory. Any more, then a pen, paper and often headtorch is needed.

What do others rely on the most?

SWL

How about we forget all about the 1 degree alteration and the 20 knot ship.
You asked about how we determine risk of collision. According to the colregs there are only two acceptable means. A visual compass bearing. Or by systematic RADAR Plotting. Even the use a full ARPA does not meet the requirements of the colregs. Some might argue the rules are a little bit out of date. They have been revised a few times since ARPA was developed in the late 70s or early 80s.
The point is the ARPA is just an aid to supplement not replace the visual lookout and compass bearing. Same goes for AIS in fact its not intended to be used for collision avoidance.
Systematic RADAR plotting is required any time a compass bearing is not possible. ARPA may supplement Systematic RADAR plotting but according to the colregs does not replace RADAR plotting.
Do you have to do this in practice? Up to you. So long as you are not involved in a close quarters situation or a collision.
What I actually do. I always have a HBC with me. In practice I rarely need to use it. I do have a RADAR on my boat. I almost never turn it on. It's at the chart table. I'm in the cockpit. I don't have AIS.
The first part collision avoidance is a good lookout. By sight and hearing ect. My Radars not on but there you go I will have to answer for that.
Your eyes are very accurate. Not at measuring distance. Other things. Particularly the corner of my eye. I often miss seeing something I am looking right at. Look away the corner of my eye picks it up. Early detection is important. Keep a good look out. Particularly behind the sail.
Where is it relative to me?. Ahead, Astern, Port, Starboard how many points(11.25 deg) from bow, beam astern.
When I see something. I like to use my binos. It starts as a spec, It gets bigger. Its coming closer. or I loos sight its getting further away. Simple and quite accurate. (provided its good vis)
I haven't pulled out my compass yet.
Is it still in the same part of the horizon relative to me or not? The binos look again. Can I see it getting bigger, Can I see its aspect.
Aspect is important. Am I looking at the bow or the stern? Which side of the other vessel can I see. Port or Starboard.
Head on? Crossing? Overtaking?
Green to Green, Red to Red all safe to go ahead tends to work. In Daylight Starboard to Starboard Port to Port.
What type of vessel is it? Again this is important. Power Driven? Sailing? Fishing? Possible restricted? NUC.
Personally I like to use the points system, If first saw it at 4 points of my bow. Now It's about 5 points? good, Or Now it's about three points good. Still 4 points maybe I should pull my HBC out.
Or just my opinion wrong aspect and the relative position on my horizon hasn't changed by at least a point. It going to be uncomfortably close.
The stantion, Is not accurate enough to give more than a rough Idea. Land behind an approaching vessel works like a transit.
Now I have a good idea if risk of collision possibly exits or I have eliminated risk of collision.
If I am give way. I will probably take action. The sooner I take action the more comfortable I am. The earlier I take action, Its more effective and the sooner I can return.
If I am stand on I have my compass out, I take a series of bearings. How much should the bearing change? Depends? on the angle and speed of approach?
I like to use 3 and 6 minutes. Why? 6 minutes is 1/10hr. Helps me make a rough guess on time and speed.
So for by far the majority of situations I am not really compliant with the colregs. I am usually only using my HBC when stand on and I am concerned. If I am give way and I am concerned I just give way.
Someone asked when should you give way.
My answer. As soon as after you have determined risk of collision exists as you can. Unless there is a very good reason to wait. I.E. avoiding another vessel. which is a complication not contemplated by open water

[El Pinguino]

Originally Posted by Uricanejack

...............
[SIZE=3][FONT=Calibri] I like to use 3 and 6 minutes. Why? 6 minutes is 1/10hr. Helps me make a rough Someone asked when should you give way.
My answer. As soon as after you have determined risk of collision exists as you can. Unless there is a very good reason to wait. I.E. avoiding another vessel. which is a complication not contemplated by open water

An excellent post.

And the sooner you alter the smaller the alteration you have to make. In the simple 'ship on steady bearing on starboard bow' situation I alter enough to show them my red and then keep aiming for their stern or close to it while still showing them red and pass as close astern of them as I consider prudent in the situation. That results in less deviation from track and more efficient passage making.

I have AIS... its an 'early warning device'... nothing more.... and at least 80% of the 'ships' I encounter don't have AIS anyway.

AIS infallable ? I suggest people read this ... https://www.panbo.com/archives/2017/...c_and_srt.html

My AMEC has spat the dummy twice in the last 2 years... that I am aware of....

[Seaworthy Lass]

Originally Posted by Uricanejack

How about we forget all about the 1 degree alteration and the 20 knot ship.
You asked about how we determine risk of collision. According to the colregs there are only two acceptable means. A visual compass bearing. Or by systematic RADAR Plotting. Even the use a full ARPA does not meet the requirements of the colregs. Some might argue the rules are a little bit out of date. They have been revised a few times since ARPA was developed in the late 70s or early 80s.
The point is the ARPA is just an aid to supplement not replace the visual lookout and compass bearing. Same goes for AIS in fact its not intended to be used for collision avoidance.
Systematic RADAR plotting is required any time a compass bearing is not possible. ARPA may supplement Systematic RADAR plotting but according to the colregs does not replace RADAR plotting.
Do you have to do this in practice? Up to you. So long as you are not involved in a close quarters situation or a collision.
What I actually do. I always have a HBC with me. In practice I rarely need to use it. I do have a RADAR on my boat. I almost never turn it on. It's at the chart table. I'm in the cockpit. I don't have AIS.
The first part collision avoidance is a good lookout. By sight and hearing ect. My Radars not on but there you go I will have to answer for that.
Your eyes are very accurate. Not at measuring distance. Other things. Particularly the corner of my eye. I often miss seeing something I am looking right at. Look away the corner of my eye picks it up. Early detection is important. Keep a good look out. Particularly behind the sail.
Where is it relative to me?. Ahead, Astern, Port, Starboard how many points(11.25 deg) from bow, beam astern.
When I see something. I like to use my binos. It starts as a spec, It gets bigger. Its coming closer. or I loos sight its getting further away. Simple and quite accurate. (provided its good vis)
I haven't pulled out my compass yet.
Is it still in the same part of the horizon relative to me or not? The binos look again. Can I see it getting bigger, Can I see its aspect.
Aspect is important. Am I looking at the bow or the stern? Which side of the other vessel can I see. Port or Starboard.
Head on? Crossing? Overtaking?
Green to Green, Red to Red all safe to go ahead tends to work. In Daylight Starboard to Starboard Port to Port.
What type of vessel is it? Again this is important. Power Driven? Sailing? Fishing? Possible restricted? NUC.
Personally I like to use the points system, If first saw it at 4 points of my bow. Now It's about 5 points? good, Or Now it's about three points good. Still 4 points maybe I should pull my HBC out.
Or just my opinion wrong aspect and the relative position on my horizon hasn't changed by at least a point. It going to be uncomfortably close.
The stantion, Is not accurate enough to give more than a rough Idea. Land behind an approaching vessel works like a transit.
Now I have a good idea if risk of collision possibly exits or I have eliminated risk of collision.
If I am give way. I will probably take action. The sooner I take action the more comfortable I am. The earlier I take action, Its more effective and the sooner I can return.
If I am stand on I have my compass out, I take a series of bearings. How much should the bearing change? Depends? on the angle and speed of approach?
I like to use 3 and 6 minutes. Why? 6 minutes is 1/10hr. Helps me make a rough guess on time and speed.
So for by far the majority of situations I am not really compliant with the colregs. I am usually only using my HBC when stand on and I am concerned. If I am give way and I am concerned I just give way.
Someone asked when should you give way.
My answer. As soon as after you have determined risk of collision exists as you can. Unless there is a very good reason to wait. I.E. avoiding another vessel. which is a complication not contemplated by open water

It is interesting to read how someone else operates.

A few comments:

Use of peripheral vision is critical at night. Lights that are invisible when you look directly at them may pop into view when you shift your gaze just a little to the side, making them visible out of the "corner of your eye", as you put it. The reason for this is that different retinal receptors that function better at low light levels are engaged. This is not the case during the day.

Unlike you, I pull the HBC out early. The reasons are several. It is more reliable than simply using the stanchion method, it provides practice, and it gives me something to do during watches. It takes a bit of skill to take reliable measurements and this skill benefits from being reinforced.

The stanchion method is a good first indictation if a risk exists, but it only works reasonably if you can keep your eye in a fixed position. On our previous boat I would put my nose or the back of my head on the centre point of the dodger and try and average the position relative to some point on the boat, not necessarily a stanchion. Again, this is a skill that takes some practice on a boat that is rolling and pitching.

I also scan the horizon slowly and carefully every 5 minutes. I don't read during watches. Careful lookout helps keep me focussed and awake.

One other point about lookout: standing when scanning the horizon is also beneficial. Elevation affects how far away we can spot targets.

SWL

[StuM]

Originally Posted by Seaworthy Lass

In all this discussion I have not heard what people consider is the most reliable tool for estimating if a risk of collision exists. Or to put is another way, what they are most comfortable depending on.

Personally, I rely on a hand bearing compass. It is always around my neck during watches. I can keep track of 2-3 boats fairly easily from memory. Any more, then a pen, paper and often headtorch is needed.

What do others rely on the most?

SWL

I have three devices:
1. a hand bearing compass (Plastimo Iris 50 which, like Jackdale, I keep in daylight to charge the fluorescence),
2. binoculars with a built in digital compass
3. an Android phone app called Digital Field Compass Plus which lets me sight a target and store three bearings.with single taps.

By night it's the HBC ( no light in the binoculars). By day, it's generally the binoculars.

[Lodesman]

Originally Posted by Uricanejack

Systematic RADAR plotting is required any time a compass bearing is not possible. ARPA may supplement Systematic RADAR plotting but according to the colregs does not replace RADAR plotting.

I believe rule 7 states "radar plotting or equivalent systematic observations..." IMO, ARPA is radar plotting, and quite frankly surpasses manual plots in many facets. That said, I could plot and calculate around course changes faster than ARPA, but there's no way on earth I could plot 200 contacts. Do you have some reference to support your point of view?

[evm1024]

Originally Posted by Seaworthy Lass

SNIP (the sound of siccsors)


The stanchion method is a good first indictation if a risk exists, but it only works reasonably if you can keep your eye in a fixed position. On our previous boat I would put my nose or the back of my head on the centre point of the dodger and try and average the position relative to some point on the boat, not necessarily a stanchion. Again, this is a skill that takes some practice on a boat that is rolling and pitching.

SNIP

Just as a point of reference:

The stanchion method is referenced to the boat (which does change its heading), The HBC is referenced to magnetic North which is constant tor the time scale we are talking about.

Of course everything you said is true but it is important to understand the uncertainties in out reference points.