Fatal Flaw in Standard Horizon Matrix AIS.

[Dockhead]

Originally Posted by tanglewood

Just FYI, class A and B are different in their use of internal vs external GNSS data. Class B is required to use the internal source, however A is required to use the ship’s GNSS data, with internal data strictly as a backup. I don’t think it’s relevant to the issue at hand, but still worth noting.

Wow, great information! Thanks for the correction.

[SailingPNW]

You might want to look and see if the issue is a GPS issue with the SH. While it has an internal GPS, it won't ever have a clear view of the sky in most cases so could get lost more easily, this is one of the reasons AIS transmitters usually require an external antenna to transmit.

Assuming your using an external antenna for the openCPN, can you configure the SH to use it instead of the internal antenna?

[donradcliffe]

Originally Posted by goboatingnow

You really can’t compare these as you need ships heading comparison to ensure both receivers are dealing with the same info.

I can compare them and have, especially in the second case. The picture of the SH radio does not show ships heading because it does not fit on the screen, but heading can be shown if you turn the tuning knob. The SOG from both screen shots is essentially the same.

I usually believe the SH numbers, but these were so wildly off that there must be a bug in their CPA calculation, and I felt the need to document them. Don't attack the messenger.

[requiem]

Originally Posted by Dockhead

Heading is not indeed used in any AIS system to "plot future positions". The algorithm for determining CPA and TCPA uses COG and SOG exclusively as inputs.

Sometimes displayed course vectors can be a bit bendy. For example, I'm pretty sure I wouldn't want to be crossing the buoyed fairway to the south in the near future.

[s/v Jedi]

Originally Posted by Dockhead

Sorry, you are mistaken here.

Heading is not indeed used in any AIS system to "plot future positions". The algorithm for determining CPA and TCPA uses COG and SOG exclusively as inputs.

HDG is transmitted by Class A sets as a separate data type, like ROT, but neither HDG nor ROT is used in any of the calculations.

Some Chinese guys a few years ago complained that AIS would work better if more data than just COG and SOG were used as inputs for the CPA and TCPA calculations. See here: https://www.researchgate.net/publica...ion_Prediction

Maybe they're right. But their suggestion has not been implemented.

I think you may be right, but I’m not talking so much about the AIS systems, as they just send/receive data, but about the plotting. I don’t know software used by a VHF manufacturer or a plotter or multi function display. I guess I can look into what OpenCPN does but nobody else publishes their routines.
The reason you’re probably right is to make it easy with one algorithm for class A and class B. But I remember using German software many years ago that show “no-go zones” in red on a kind of radar plot window and they acted very differently for class A targets. I really think they used ROT as least, because it showed corkscrew like projections for targets that were turning.

But just data in a table like CPA and TCPA yes that’s probably only from satellite position.

[Captain Bill]

The OP's matrix AIS/GPS is a receive only unit. I'm not sure that the standards for Class B transponders apply at all. SH sells or at least used to sell a model that used an external GPS signal called the Matrix AIS.

[jamhass]

I've still not read which display was correct?

The OP suggested SH was the culprit, but no real evidence noted.

I will say that I owned one of the earliest Matrix AIS radios and ended up returning it to SH for a software update -- because it wasn't showing certain Class-A AIS targets such as tugs in tow. It was promptly and effectively fixed. I got a couple of free SH Matrix AIS T-shirts as an extra reward for bringing the issue to their attention.

[El Pinguino]

Originally Posted by jamhass

I've still not read which display was correct?

The OP suggested SH was the culprit, but no real evidence noted.

I will say that I owned one of the earliest Matrix AIS radios and ended up returning it to SH for a software update -- because it wasn't showing certain Class-A AIS targets such as tugs in tow. It was promptly and effectively fixed. I got a couple of free SH Matrix AIS T-shirts as an extra reward for bringing the issue to their attention.

See post #7

[wingssail]

I think we can determine that the SH is incorrect.

Using info from the screenshots post #7 we can establish the positions of both vessels:

Pan Glorious 38 deg 02.2 N 135 deg 39.03 W

Ownship 36deg 47.16 N 133 deg 01.55 W (based on bearing and distance from Pan Glor)

The two system reported completely different scenarios:

OpenCPN shows CPA of 1nm in 63 minutes. Pan Glor would have traveled 9.6 nm on it's heading of 215. For a CPA of 1 nm then Ownship would have to traveled on a course of 100deg at 8.4 knots. To me this is entirely reasonable.

SH Matrix shows CPA of 13.3nm in 62 minutes. Pan Glor would have traveled 9.6 nm on it's heading of 215. For a CPA of 13.3 nm then Ownship would have to traveled on a course of 221deg at 5.9 knots. To me this is also reasonable, EXCEPT, the SH display shows Ownshio on a heading of due EAST.

The SH display cannot be correct unless it is getting a instantaneous SOG for ownship of 90deg,and in that case the CPA and TCPA should have been quite different.

I don't know what was happening in the SH radio, but I think that the display was wrong and depending on it would have put the vessels in closer proximity than the radio projected.

It would be very helpful if OP could give us his course and speed.

[s/v Jedi]

I have a newer SH with NMEA2000 and have found many bugs so I would not be surprised to find bugs like that.

I also reported them but SH simply denies; for legal reasons I assume.

I remember many years ago that I reported AIS bugs to SRT (China) for one of their first AIS transponders. They contacted me within hours and sent me a complete development kit incl. chip programmer and together with their software guys I got the bugs out and they took off now being one of the major AIS players in the market.

SH is not on a good path, very un-Yeasu imho.

[guitarrich]

Originally Posted by tanglewood

Looks to me like a display bug where it's displaying RNG in place of CPA. If would be interesting to see if it alarms based on the correct CPA, or the faulty one begin displayed.


What firmware revision do you have?

Yes, I noticed that in the first example, and there it is again in the second. Chase down the firmware. Looks like some QA folks have been sleeping on the job!!

[wingssail]

Originally Posted by guitarrich

Yes, I noticed that in the first example, and there it is again in the second. Chase down the firmware. Looks like some QA folks have been sleeping on the job!!

No, it's more than that. Looking closely at the display you see the OP's vessel heading EAST, away from the path of Pam Glor.

[CaptJPMcGuire]

This has been a great conversation. I have been out of serious boating for a year or more but do have a kayak. I plot my CPA and R&B by eye --used to do a similar thing with a hand bearing compass and a plotting sheet with pencil and dividers. Then I got Radar, AIS, GPS and KVH Compass all backboned & networked with NMEA 2000 ---forgot the fun it once was --glad to be back in the kayak🥺

I am being a bit turs with these comments but seriously guys and gals look out the window and enjoy the beauty of the ocean. Cruising should not become a video game.

Jim

[requiem]

Originally Posted by wingssail

No, it's more than that. Looking closely at the display you see the OP's vessel heading EAST, away from the path of Pam Glor.

How did you get the heading numbers for the two vessels? From the SH display, I estimate both Pan Gloris and OP on a course of ~115T ±10°. From the OCPN display, Pan Gloris reports a course of 102M, or about 115T.

[wingssail]

Originally Posted by requiem

How did you get the heading numbers for the two vessels? From the SH display, I estimate both Pan Gloris and OP on a course of ~115T ±10°. From the OCPN display, Pan Gloris reports a course of 102M, or about 115T.

Well I was looking at the second set of screen shots, in post #7.

OpenCPN shows a Heading for Pan Glor of 215deg. SH shows (the little progress vector) of roughly the same direction. The numbers are not present but the small screen circle had north at the bottom and the small progress vector to the right of 180, so maybe 215. We have the lon/lat of Pan Glor and the distance and bearing from ownship. From that we can deduce ownship's position.

So if we assume that the heading of Pan Glor is 215 and its speed is 9.4kts, we can advance its position an hour later along that same course.

For using OpenCPN's calulation, to arrive at a CPA of 1nm an hour later (according to the Opencpn display), the new position of OP's vessel can be assumed and it's heading and course as well.

On the other hand, the SH calculation of CPA of 13.3 miles 1 hour later, and the SH showing ownship heading of EAST means that they were not converging. Finding where ownship could be an hour later with a distance of 13.3 miles, results in a a heading of roughly 221deg and probable speed of 5.9kts. However the SH screen shot shows ownship heading of east. I can't find any solution which works for a heading of 90deg. So I conclude the SH display is incorrect,. Again, OP could tell us his position, speed and heading which would clarify things.

Perhaps there could be other interpretations but that's what I see. I tried to draw it that way.