Radar NMEA messages

[SolentBlue]

Recently somebody mentioned that you needed more information about nmea radar sentences. I extracted this from my Raymarine E80 which I hope you find useful. It might be a great enhancement to overlay radar data on the chart.

The first sequence is the Radar System Data which covers:
- Radar Range Setting
- Radar Cursor Position (relative to centre/ship)
- EBL/VRM 1
- EBL/VRM 2

The "N" is nautical miles scale (I believe others are K and S but the E80 doesnt do them)


$RARSD,,,,,,,,,0.000,000.2,0.125,N,N*50

Cursor at vessel pos
Range 1/8 mile

$RARSD,,,,,,,,,0.145,244.5,0.125,N,N*55

Cursor moved now 0.145 NM relative bearing 244.5 degrees


$RARSD,0.883,099.3,0.200,047.6,,,,,0.145,245.2,0.1 25,N,N*54

EBL1 = 47.6 Stbd
VRM1 = 0.2000 NM

$RARSD,0.150,233.5,0.177,086.4,,,,,0.150,233.5,0.1 25,N,N*59

Origin of EBL1, VRM1 moved 0.15 nm bearing 233.5 from ship
EBL1 is 86.4 degrees (relative) and VRM is 0.177 NM


$RARSD,0.251,243.9,0.177,087.5,0.376,131.3,0.167,0 11.5,0.376,131.3,0.25 ,N,N*5,N,N*5

VRM2 offset 0.376 nm 131.3 degrees relative
VRM2 range 0.167 nm
EBL2 10.6 degrees stb

VRM1 offset 0.251 nm 243.9 relative

$RATTM,01,0.438,018.6,T,0.00,180.0,T,0.438,0.0,N,T GT 01,Q,,180540.43,M*50

Target 01, range 0.438 nm direct 018.6 TRUE

$RATTM,01,0.438,018.9,T,0.99,103.5,T,0.438,0.0,N,T GT 01,T,,180600.47,M*57

Spd target 0.99 kts
Heading target 103.5T

Distance to CPA 0.438

I hope this helps, let me know if you need any more.

[nohal]

SolentBlue...
Many thanks for the info. I have the initial implementation done already and the info collected so far is at https://github.com/nohal/OpenCPN/wik...implementation - your E80 fits in there pretty well. If you have a chance and time, could you please have a look at the stream when new target is acquired, a tracked target is lost and deactivated?
Being curious I would also appreciate a recording of NMEA stream (including own ship GPS data) while the radar is tracking a static and a moving target from a moving vessel. None of it is critical though and I'm pretty sure there will be a lot of this info available when the next beta is out with this code integrated...

Thanks again

Pavel

[SolentBlue]

OK I will try and get that next weekend

[Coquina]

Back in the Day - 1994 or so - I was a Raytheon dealer and they had a function where you could put the cursor on a radar target and have that same cursor appear on the chart plotter. This could be a useful function for OpenCPN.

[SolentBlue]

Hopefully attached you will find about 6-7 mins of nmea stream taken whilst I captured multiple radar targets, some moving and some stationary bouys. I was sailing throughout, and in the stream is heading data from a fast compass and plenty of AIS targets. Some of the radar targets are also AIS targets. One or two radar targets were subsequently "lost".

Please note my radar works in true motion; and all this is "real" not simulated; so the compass is off about 1.5 degrees, hence radar targets done exactly line up with AIS targets. Also there was a strong tide, sometimes stationary targets appear to move a little !

I think this covered the use cases you wanted - do let me know if you would like any further data.

OK - file didn't attach because it is .txt and 385k - how should I get file to you ?

[LeaseOnLife]

Originally Posted by SolentBlue

OK - file didn't attach because it is .txt and 385k - how should I get file to you ?

Change the extension from .txt to .txt.pdf and attach again. pdf files can be some 593kbytes, while txt files can only be 1 kbyte on this forum.

Dirk

[SolentBlue]

Thanks for advice. NMEA stream as documented above

[nohal]

SolentBlue...
Got it, many thanks for your help.

Pavel

[nohal]

Very educative... What's puzzling are the speeds reported in the stream - unlike the rest clearly not true as expected but relative to own vessel. Now let's find out how as for the static targets they look just "weird"...

Thanks once again

Pavel

[nohal]

SolentBlue...
One more question: Is your radar fed with all the navigation data (especially speed) or just with the compass heading? From the values it produces it looks as if it was convinced your own speed is actually 0 when it computes the target speed...
Knowing that would help me to understand if we have to recalculate the speed it provides in some way or do our own calculation based on target's positions in the subsequent reports which BTW works pretty well and we have to do it when TLL is received anyway.

Thanks

Pavel

[SolentBlue]

The radar gets all the data:

Heading + DSC + AIS (incl GPS 2) via MUX into NMEA IN
Speed/Log/Depth/Wind/GPS 1 via SEATALK

Radar uses my speed/heading to compute CPA.

Also, I remembered TGT 01 was a lost target (red diamond symbol) and generated an alarm, but I noticed it just disappeared from the NMEA stream. You might need to watch that because once the radar has lost TGT 01, it will reuse that name for a subsequent acquisition.

Bear in mind, even when a target is acquired for maybe a minute it's speed and course are not reliable; it takes the radar some time to plot going, longer than just a target being acquired.

[nohal]

SolentBlue...
Thanks for the info. What's puzzling is that the target moving at 21 knots northwards has the radar SOG about our SOG less than it's actual speed reported over AIS (I would understand that if the radar actually didn't know we are also moving, as the difference is +/- the same as our SOG and so is our and the target's COG)
But the buoys, while reporting "random" COG (which is understandable as we look at them with a radar and they are more or less stationary while we are moving and weather happens), are pretty constant in being reported moving at more or less our SOG, regardless if we are approaching, passing by or leaving...
Anyway, I will do some calculations and choose the best we can get from it.

Pavel

[SolentBlue]

Pavel

Thanks for including this target tracking functionality in the latest beta.

I noticed a couple of items about the implementation:

When I cancel a target on the E80, it says with OpenCPN for a long time before it is dropped -- I guess there is no message in the NMEA stream that says target cancelled/lost so you just timeout when you dont get an update.

If I acquire a target, and drop it. Then acquire a new target (even several minutes later), the E80 might reallocate the NAME. It always uses "TGT 01" if that slot is available.

OpenCPN sees this as the SAME target, and joins the tracks together. You can see this in the picture, where the magenta line into 2 of the radar targets shows a long straight line which is the connection to the previous target of the same name.

Can you timeout quicker (or allow a user setting for timeout) ?
Can you provide a manual way to say cancel target on OpenCPN ?
Can you check the lat/long and time between updates and conclude it must be a different target ?

Also on the picture you can see slow moving targets have wobbly tracks. No doubt partly due to my radar and returns. Nevertheless, can you increase the damping time for slower targets to eliminate the wobble ?

Thanks
John

[nohal]

John...
Thanks for the comments. The ARPA targets except getting it's data are now treated exactly as if they were AIS (including timeouts). From my collection of NMEA samples, Raymarine radars seem to be the less predictable especially when it comes to target deletion - they simply stop sending the data, without any good bye sentence...
I will implement a few more settings there to tune the timeouts to the user's needs and also probably use, Spock forgive, the "warping" detection as you suggest to distinguish between targets with same name.
How do you consider the target COG/SOG? Acceptable? With the sample data available I actually get much better results calculating it myself based on the changing target position than relying on what the radar reports, but definitely it may be just in that particular case (probably another candidate for user setting)...

Pavel

[rgleason]

Fascinating discussion. Many things I know little about. Thank you both.