AIS vessel accuracy looks to be getting USCG attn

[ronstory]

I have been windows shopping for an AIS transponder for a year and current have a Simrad RS35 with AIS receive only connected to my Raymarine chartplotter. I often look at MillTech marine for info since they provide a lot of info their website.

I just ran across a blog on their site that has some suprising stats from the CG on the accuracy of the info in the AIS fleet.

https://www.milltechmarine.com/US-Co...9.html?draft=1

Paraphased from the CG Navigation center:

"A snapshot from a single day of data collection from southern California showed half of all transponders weren't transmitting the correct information, which included both Class A and Class B transponders. For certain vessel categories (fishing vessels) the error percentage is as high as 79% which indicates the Coast Guard has a long list to work through before reaching 100% accuracy."

I found it irritating that we need to fill out a form and only "authorized" centers can program the AIS Transponder info. Looks like it's a needed irritation.

BTW, I have no affiliation with Milltech Marine, other that for window shopping... so far.

[Paul Elliott]

Fortunately, the accuracy discussed in the Milltech Marine piece apparently isn't the critical stuff (position, course, and speed, contained in the dynamic AIS messages). The inaccurate information comes from the static messages: vessel dimensions, type, nav-status, callsign, destination, ETA, etc. These are the human-entered values, and yes, they are often not correctly set.

While is may be disconcerting to see a ship allegedly "At Anchor" speeding towards you at 15 kts, most of us have learned to take any of the human-provided data with a big grain of salt.

Even an erroneous MMSI (which appears in *every* AIS message, not just the static ones) will not affect our basic collision-avoidance piloting. A bogus MMSI will likely prevent DSC calling on the VHF radio (unless the radio is similarly mis-programmed).

So I can understand the USCG wanting to see all this data properly entered. So do I, but I also understand that any human-entered data is potentially wrong, and should not be relied on when safety is at risk.

[jackdale]

I once encountered a "moored" vessel in the mid South Atlantic. What was confusing was the triangle orientation indicated a collision course, while the AIS information indicated crossing ahead at right angles. That is also what the vessel lights indicated. A radio call confirmed a crossing. I forgot to tell the other vessel about the erroneous status. My bad.

[Paul Elliott]

And don't forget that even the critical dynamic data can be dangerously in error, due to gear malfunction or GPS problems. At sea I once had a freighter pass me roughly one mile to my north, while it's AIS position put it about one mile to my south.

I believe they had a problem with the way their chartplotter was delivering the lat/lon to their Class-A AIS. Apparently the plotter changed the lat/lon internally to compensate for the chart datum, and then sent this "corrected" position to the AIS transponder. The USCG put out a bulletin about this issue, and I believe that the problem has largely been corrected.

But you never know! I love my AIS, but I still use my eyeballs and radar as well.

[CaptTom]

There are lots of errors out there.

The ones I notice most on Class A targets are status information like the aforementioned "at anchor" vessel steaming along. Last port and destination are often clearly incorrect, too.

On class B targets it's very common to see the length entered incorrectly. Usually you'll see something like 30 meters when it's clearly more like 30 feet.

[TheOffice]

My neighbor's Catalina 44 was 267 feet long on AIS until I told him about the error.

[bobgarrett]

And then there are sailing yachts identified in AIS as pleasure vessels.
On a rally we noticed one of our companions having more issues with other vessels at night - then we noticed he was identified as pleasure so commercial vessels did not recognise he was the stand-on vessel.

[S/V Alchemy]

Originally Posted by Paul Elliott

And don't forget that even the critical dynamic data can be dangerously in error, due to gear malfunction or GPS problems. At sea I once had a freighter pass me roughly one mile to my north, while it's AIS position put it about one mile to my south.

I believe they had a problem with the way their chartplotter was delivering the lat/lon to their Class-A AIS. Apparently the plotter changed the lat/lon internally to compensate for the chart datum, and then sent this "corrected" position to the AIS transponder. The USCG put out a bulletin about this issue, and I believe that the problem has largely been corrected.

But you never know! I love my AIS, but I still use my eyeballs and radar as well.

I love my GPS, but I still bring my sextant, as well!

[S/V Alchemy]

Originally Posted by TheOffice

My neighbor's Catalina 44 was 267 feet long on AIS until I told him about the error.

I noticed my sailboat making five knots in a 12 knot breeze once on GPS. Then it made 55 knots for a few seconds and repositioned me a quarter-mile to the west. While some research later on suggested that the three satellites it was referencing were perhaps too tightly clustered overhead (or were all low on the horizon), it was a reminder that nothing's perfect, and to keep a watch. I often do coastal pilotage with a sextant reading off heights of radio towers and other prominent daymarks, or with a hand compass doing running fixes. Keeps those parts of the brain active that do not generally wake up reading a plotter.

[roland stockham]

Guilty, I don't always enter my destination port (don't always know it!). Occasionally I have forgotten to reset from anchor when getting underway and done it after clearing harbour.

[CaptTom]

Originally Posted by roland stockham

Guilty, I don't always enter my destination port (don't always know it!). Occasionally I have forgotten to reset from anchor when getting underway and done it after clearing harbour.

In the overall scheme of things, that's no more than a chuckle for the rest of us, at your expense. Sort of like leaving the fenders out underway. We laugh because we've all done it at least once!

[smac999]

99% of people looking at AIS targets only care where they are on the map, direction, and speed. that info is from staring at the tringle moving on the screen. no need to click or look any further. only if you need a boat name incase they have to call it. those things likely aren't wrong.

the only people looking at boat status, boat type, destination etc is marine traffic controllers. and maybe coast guard / customs people. joe blow in his boat isn't going to carea bout that stuff. so it doesn't really even matter if it's wrong.

I mean if a boat is heading towards you. do you really click into multiple menus on the plotter to see if it's listed as anchored or moving? it's obviously moving.

[smac999]

Originally Posted by bobgarrett

And then there are sailing yachts identified in AIS as pleasure vessels.
On a rally we noticed one of our companions having more issues with other vessels at night - then we noticed he was identified as pleasure so commercial vessels did not recognise he was the stand-on vessel.

do you change it back and forth when you go from sail to motor?...

[Jim Cate]

Originally Posted by roland stockham

Guilty, I don't always enter my destination port (don't always know it!). Occasionally I have forgotten to reset from anchor when getting underway and done it after clearing harbour.

Roland, are you saying that you have a class A AIS on your 35 foot yacht?

Pretty unusual!

Jim

[bobgarrett]

Originally Posted by smac999

do you change it back and forth when you go from sail to motor?...

Most class B sets will not allow this to be changed once it is installed. In fact in this particular example we could not find out how to change the setting and the owner had to get the installer back to correct it.

Remember, this denotes vessel TYPE not what the vessel is currently doing. This is set out here on page 11 http://www.imo.org/en/OurWork/Safety...uments/227.pdf

So, there is the issue that a vessel identified as a "Sailing Vessel" but is actually motoring.