Panbo

NMEA 2000 AIS, not yet right!

... written for Panbo by Ben Ellison and posted on Jun 12, 2009
Simrad_AI50_testing_cPanbo.JPG

Yesterday I fired up this sample Simrad AI50 Class B AIS transponder and found it to be pretty much as self-contained and impressive as I'd hoped.  I attached one of my boat's VHF antennas, deployed the AI50's included GPS antenna, gave the unit a little 12v juice (just 8 watts at 100% screen brightness), and, voila, Gizmo was transmitting its position and plotting other AIS targets, including another Class B I had set up as "Panbo.com Lab".  A full AI50 entry will follow, but first I'll report on its SimNet/N2K output.  I was excited about how easily NMEA 2000 could feed the AI50's target and GPS info to all devices on the network, but nervous about that how well 2000 currently handles the data (nobody has yet tried it much). Both feelings were justified...

It was trivial to tee a SimNet-to-DeviceNet patch cable from the AI50 into Gizmo's N2K backbone, but the results were quite disappointing.   As you can see from the above photo, the AI50 understands the pesky "Class B CS Static Data Report" -- i.e. Message #24, added to the AIS standard after its initial release (though discussed here over two years ago)  -- but that info did not get to the Garmin 5212 that's currently on Gizmo's fly bridge.  In other words, the Garmin saw the Class B "Lab" target fine via N2K, but never received the name, call sign, etc. message it was sending out every six minutes.  And I don't think that's Garmin's fault, because when I had N2K Analyzer monitor the AI50's data output (below), I never saw a Class B Static Data PGN.  (Another bummer: the Maretron USB100 does not translate N2K AIS messages into 0183 AIS messages.)  Plus, when you select the AIS category in NMEA's nifty new online PGN database, it's not listed there either (check Brian Lane's full AIS message list for comparison).  Could it be that NMEA hasn't caught up yet, and therefore no N2K AIS transponder or receiver (they're coming) can yet deliver Class B static data?  Sheesh!
   And there are other issues.  Note below that the AI50 is only sending out two basic GPS PGNs, no satellite status or other secondary info.  Which is similar to what Class B's do via NMEA 0183 (the subject of recent discussion), and has mixed results.  For instance, the Garmin offers the AI50 as an alternate GPS source, but if selected then alarms you that it's "Lost Satellite Reception" (even though the AI50 is receiving GPS fine).  Doh!  An AIS engineer recently explained to me that Class B transponders only output limited GPS info because more data might interfere with a high volume of AIS target data trying to fit itself into a 36,400 bps NMEA 0183 pipe.  But NMEA 2000 doesn't have this bandwidth limitation, and thus N2K Class B output should include all GPS info so that any device on the network can use it for primary or backup positioning. (Or, better yet, let the user/installer decide.)
   There may also be a problem with a vessel's own N2K Class B data...

Simrard_AI50_NMEA_2000_output_cPanbo.JPGIn NMEA 0183 format your own vessel info is segregated from regular !AIVDM target messages as an !AIVDO message. But in 2000 format, your own vessel seems to be treated as just another PGN 12039, as Gizmo is in the screen above (that's my own shiny new FCC MMSI).  This would explain why the Garmin was plotting Gizmo's own transponder as a dangerous target a few yards away, seen below (the two highly accurate GPS's involved were separated about as shown).  This may also explain why the Lowrance HDS, also able to do AIS over N2K (I think), has a place to enter your own MMSI (I'll try it for sure, but that sounds like an easy way for an MFD to identify its own vessel as such).
   But there is some good news in these two images.  Note how the Simrad AI50 automatically picked up Gizmo's heading from the N2K network and included it in the boat's AIS message. That's why the Class B target below is one of the first I've ever seen that is pointed in the right direction even though the boat is not underway (tied up at a good watering hole, actually), and therefore lacks valid COG.  In fact, both the AI50 and the Garmin are both getting heading info from the Airmar PB200, with little effort on my part, which is a good example of NMEA 2000 doing what it can do so well.  N2K could also do a great job of networking AIS data, but it seems like the NMEA and the manufacturers haven't yet gotten the details right. 

Garmin_5_own_vessel_class_b_target_cPanbo.JPG

Comments

Doh! I wrote this entry early this morning, but forgot to hit the 'publish' button.

But today I did get a chance to try the Simrad AI50 with the Lowrance HDS-10, via NMEA 2000. Sure enough, my own transponder showed up as a dangerous, and very nearby (;-) target. But when I entered the right MMSI on the HDS, the ghost target went away. So Lowrance has figured out this little N2K/AIS oddity, and hopefully Garmin and others will follow.

The HDS is also willing to use the AI50 as a GPS sources. Rack up some NMEA 2000 points for Lowrance, and have a nice weekend!

Posted by: Ben at June 12, 2009 5:52 PM | Reply

Ben you have the identical problem I have with my NAVICO 300L. If I disable the transmitter I am fine but I didn't purchase a receiver! I just bought a NEMA2000 Maretron tester to start looking my network over. I have all the PN's the Navico transfers. It will be nice when I can use iot in the transponder mode instead of the receive only mode with my Garmin 4212 and 4210 network. I am also missing the static data from Class B users like the ships name etc. On a positive note I did install a Lowearance NEMA2000 VHF radio and am happy with it's ability to interface with my Garmin 2000 network. I am waiting for the Class B AIS N2K fix!
Regards Bill Lentz
Wireless One

Posted by: Bill Lentz at June 12, 2009 11:02 PM | Reply

Ben one more comment when I use the RS232 and my laptop it does not think my vessel is a dangerous target. It is only when it is interfaced to the Garmins.
Bill Lentz

Posted by: Bill Lentz at June 12, 2009 11:16 PM | Reply

Bill, I suggest that you try using the transponder's NMEA 0183 output instead of N2K. I think that if you install it to one of the Garmin's 0183 ports it will get to the other Garmin MFD via Ethernet. It's a backwards move in terms of easy, tidy, powerful installations, but it will very likely work better until Garmin, Navico, and NMEA get their N2K/AIS acts together.

Posted by: Ben at June 13, 2009 7:13 AM | Reply

Interesting. My ACR class B sends 0183 data to the nav PC running Fugawi, and I see the AVIDO messages but Fugawi doesn't plot my boat as an AIS target. I do see all data from both A and B units, though.

Michael

Posted by: Michael Porter at June 13, 2009 9:20 AM | Reply

It's hard to believe that marine electronics manufacturers really want to make the transition over to N2K. At every step of new product development they seem to shoot themselves in the foot. I understand that most participants on this blog are very skilled and professional and can find a work around to solve a lot of the dumb design problems presented in each new product, however, I look at how the end user will deal with these issues. No one wants to pay $3000 for a MFD and $700 for class B, AIS, install an N2k backbone and then have to jury rig a solution so they don't show up in the middle of route I95 or set off collision alarms. NMEA 0183 is a mess to work with and why N2k isn't embraced, perfected and universally adapted in every new product is a mystery to me. When AIS class B was developed N2k was already known to be the format of the future so why did a single unit show up without it? One more gripe, any manufacturer who starts to cannibalize N2K with proprietary PGN's is not even a consideration for purchase on my boat. I'm personally tired of buying product that can't fully communicate with what I already have installed, (N2K to N2K).

I'm willing to make the jump to N2K as soon as the equipment manufacturers get it right and I'm sure you will agree that when this happens it will be like opening the flood gate for their business. Good riddance to NMEA 0183 and the rats nest of wire.

Posted by: Richard C at June 13, 2009 9:50 AM | Reply

The vendors are almost to the point of eliminating a major advantage of MFD's over PC's. (Gripe x 3 ... compliment to follow if you keep reading)

Gone seems to be the promise that innovation will be delivered in MFD's if we just wait the extra time needed (years) for innovation to appear in marine products along with the rigourously testing of software/hardware that is promissed and needed when putting out a marine product that won't require the microsoft like patches (once a month).

Is this Innovation? If innovation it is coming too slow, and of recent, without evidence of rigourous testing.

I think this comment / fustration is well deserved given that adopters of NMEA2000 seem to still be on the bleeding edge.

2008/2009 seemed to be the year of NMEA2000, maybe that claim will have to wait until 2010 ?

Either way ... more and more I am convinced that NMEA-0183 is well past retirement age, and welcome / embrace / and value Ben's assistance and reporting that gives us visability and reduces fustration as we make the best of this major upgrade to how our boats systems communicate with each other.

Thank you Ben !!

Posted by: Dan Corcoran (b393capt) at June 13, 2009 2:39 PM | Reply

I agree with Richard. After talking with many engineers at IBEX I decided (easily) to go with the clunky 0183 instead of 2K in outfitting Barbara.

I had then (2005-06) no confidence that any manufacturers were committed to making _any_ new standard work, but that they felt they had to make some minimal gesture, which is to say 0183.

The manufacturers were/are much more interested in locking consumers into proprietary protocols, and evidently this is not changing.

Michael

Posted by: Michael Porter at June 13, 2009 4:10 PM | Reply

Thanks, Dan, but I only partially agree with your gripes. I think the manufacturers have come up with numerous innovative MFDs, though never as quickly or as bug free as we'd like. But then again product updating has gotten easier and easier. Plus we should not forget that marine electronics is a relatively tiny industry with a wickedly fragmented customer base.

Posted by: Ben at June 13, 2009 4:51 PM | Reply

Michael, I probably would have recommended at least a rudimentary N2K network for Barbara, and still would. I strongly believe in its potential, and in designing electronics systems for future growth...and in many ways N2K is working well right now.

But it's always been possible, and sometimes necessary, to mix N2K and 0183. For instance, there weren't any N2K AIS receivers or transponders at first, and there still aren't many.

Still, it's shameful that the NMEA has apparently not yet adopted the Class B static data message to N2K, though the message was created years ago, and I admit to some denial on the subject. I heard about this last February, as discussed here:

http://www.panbo.com/archives/2009/05/jrc_jhs-770s780d_direct_call_by_ais.html

But I had a hard time believing it until I actually saw it. I don't see it as a Machiavellian manufacturer's plot, though. Think about it. Big players like Navico and Raymarine are selling Class B transponders that output NMEA 2000 and their newer displays -- as well as Garmin's and maybe others -- are able to display it. But they can't see the static data coming from other Class B transponders! That's looks more like NMEA incompetence than a Machiavellian plot.

Posted by: Ben at June 14, 2009 8:01 AM | Reply

Well, Ben, there is a nice bit in "Winnie the Poo" where Eeyore says "bounced or hooshed, it's all the same at the bottom of the river." By which I mean that I never was a great fan of networked instruments in the first place and have no great faith in the manufacturers or NMEA to make it work.

What I have works for me -- the GPS and fluxgate talk to the radar, GPS, Wind (PB-100) and AIS talk to the PC, GPS talks to the radios, and that is it. I may someday let the GPS talk to the autopilot.

I'm a dinosaur, but my boat does not go on the rocks because the GPS told it to.

Posted by: Michael Porter at June 14, 2009 10:16 AM | Reply

Machiavellian plots and huge government conspiricies require greater cohesion, strength of purpose, intelligence and determination than either respective population can provide.

Posted by: Sandy Daugherty at June 14, 2009 11:42 AM | Reply

Do you have any firsthand experience with the Raymarine 500 unit yet, displaying on their E series displays? I am tempted, but wonder if many issues still exist with the E series NK2 bus. In other words, does this combination actually work? Several recent things from Raymarine don't....

One worry being that the 500 still outputs 0183, so like their X30 pilot may have to be hooked up that way to an E series, in spite of the NK2 connection at both ends. I would ask Raymarine tech support, but recently their answers have had a random correlation to the truth.

Posted by: Anonymous at June 14, 2009 12:14 PM | Reply

Raymarine is sending over a C140W and an AIS500 soon, but in the meantime I think the situation is this:

* The AIS500 uses a proprietary PGN to send Class B static info over SeaTalkNG/N2K (though Raymarine told me in Feb. that it would switch to a standard message as soon as available, which NMEA told me would be "very soon").

* But I doubt that the current E-Series can see any AIS info over N2K, and guess that's coming with the delayed major update that's supposed to fix a number of Ray's N2K issues. Let's hope.

Posted by: Ben at June 14, 2009 12:31 PM | Reply

Fair enough, Michael, and an Eeyore reference trumps all others. But I think we can agree that a GPS can only tell an autopilot to go on the rocks if the GPS operator tells it to.

PS:
http://www.engadget.com/2009/06/13/gps-coordinates-lead-demolition-crew-to-destroy-wrong-house/

Posted by: Ben at June 14, 2009 12:35 PM | Reply

Right Ben, it is true that my GPS does not usually tell me to run onto rocks. But I think the QE2's did (mate on watch didn't know how to use it) and I have several times had to duck out of the way of boats running along the line their course plotter told them to follow, without regard for the situation.

Of course one can be an idiot without GPS too, and the GPS-plotter combination has really revolutionized piloting.

Posted by: Michael Porter at June 15, 2009 7:06 AM | Reply

This is my first posting on Panbo, so please bear with me.
I am an ex Navico engineer in fact I am the engineer who was in charge of the AI50.

I have been reading through your comments and must admit that a rye smile came to my face at some of them.

I can assure you all that there is no conspiracy between the manufacturers, if anything they are falling over themselves to ensure that their equipment can communicate with the competitors. After all, how else are they to get customers to buy their products if they don’t talk to anything else (other than if you wanted to kit your vessel out with just one manufacturer)?

Unfortunately, the frustrations you feel over the use of NMEA2k, is indeed, felt by those same people that make the products.

I first started to look into AIS products back in early 2004, back then there was a NMEA2k specification for AIS but only at a draft level. I pointed out and discussed some concerns I had with the structure at the time, including the fact that there is no differentiation between VDO and VDM sentences.

I even proposed that the NMEA2k AIS sentences should adopt a similar structure to that of NMEA0183 AIVDM & AIVDO. As this would significantly improve the required network bandwidth for NMEA2k AIS (NMEA2k requires a lot of overhead).

Much to my dismay none of my suggestions were considered and when I came to begin the AI50 project, in 2006, the NMEA2k sentences had been released but pretty much as per the original draft. Then to put the cherry on the top, they had been released without any consideration of Class B static data.

Once again, I had many discussions with NMEA2000 concerning this, including the Navico committee members (all of the big marine names have people who sit on the NMEA2k committee) trying to cajole them to my righteous cause.

Unfortunately NMEA2k tend to be very slow to react and when it became apparent that these sentences where not going to be approved in time for product release we decided to release with a proprietary sentence that would at least enable other Navico product to show this data; with the view that as soon as the approved sentences were released a simple software update (easily performed by any user) utilising the SD card could be performed.

That was back in 2007 and although I no longer work for Navico still have my contacts within the organisation and to my knowledge the new sentences still have yet to be released (although I have seen them in draft form).

I realise that this posting will not do anything to abate your frustrations at NMEA2k but please try to remember that you are not alone in your woe, just give a small thought to the poor sods that have to deal with NMEA2k to bring you all of the nice products you want to have on your boat :)

Posted by: Wayne at June 15, 2009 9:34 AM | Reply

Thanks for posting, Wayne. NMEA has not answered my query yet, but I'm getting the strong impression that the N2K Class B static data PGN does not yet exist. So there is nothing Navico/Simrad could have done to make the AI50 work better over N2K.

Frankly, I'm dumbstruck that the NMEA could take so long to create an important PGN. It seems that due to its slowness every boater willing to try AIS over NMEA 2000 will only get complete Class B target data if they happen to have the same brand transponder and plotter (thanks to the proprietary message work around). And when NMEA finally completes the PGN, the transponder and all networked plotters and charting programs will have to be updated to understand it.

And this situation certainly isn't because Class B transponders were rushed to market. How did it happen?

For the sake of clarity, let me add that the Simrad AI50 plots Class A and B targets fine on its own, and I'm sure it does so via its NMEA 0183 output too. In fact, it's turning out to be a very nicely designed Class B transponder, and I think Wayne (who I don't know) deserves to be proud of it.

Posted by: Ben at June 15, 2009 10:29 AM | Reply

Thank you Wayne!

It helps to understand the real source of the problem.

I had thought about volunteering to join the N2K committee as a way to speed up the cycle of innovation (specifically in the area of enabling more capabilty between VHF and MFD, and configuration of everything) and be more involved in the innovation process rather than just whining about it.

Keeping in mind if I joined such a committee probably my vote wouldn't count (I am not a manufacturer, nor any other way involved in marine electronics other than using them) but presenting ideas right at the source would possibly enable me to influence things a little ... but, from what you wrote ... sounds like I would have zero impact if your efforts to influence N2K had so little effect for such a clear cut need.

Ben, have you ever considered joining? I bet you could make a difference? I would hate to be the manufacturer that didn't take notice of your advice, and then had their product reviewed here !

Posted by: Dan Corcoran (b393capt) at June 15, 2009 1:07 PM | Reply

I just had a long chat with Steve Spitzer, Technical Director of the NMEA. He assures me that the Class B static data PGN will be published next week, along with four other AIS related messages. Which means that all the manufacturers involved will then be able to rewrite their code to do more thorough and more standardized AIS plotting over N2K.

Steve acknowledges that the process has been slow, but when he starts listing all the different parties who need to review, comment, and sign off on proposed PGNs, it's hard not to feel some sympathy. And the current business climate is not making things any easier!

But Steve did say that the Intelligent Gateway initiative, now more desciptively renamed the Third Party Gateway, is well along and should be done by the end of summer. I think that will be a big deal, when small developers can get access to N2K data on a PC, and even write code that broadcasts PGNs (safely), all at modest cost (purportedly).

I think we'll see big improvements in the charting programs, including those made by the marine hardware manufacturers, plus a slew of small niche programs, like for analyzing sail performance or comparing wind sensors (;-) However, comprehensive PGN sets to handle Alarms & Faults and Distributed Power are still very much in draft mode, with much work left to do.

Dan, we should be careful not to confuse the PGN creation process with managing that process. You or I might be helpful in some cases for designing useful network messages, but I for one would have little idea how to manage the whole cumbersome process. I just want it go faster.

Posted by: Ben at June 15, 2009 2:27 PM | Reply

Panbo has had some serious server problems this morning, but I think they're fixed. However, if you have trouble commenting, or notice something funky, please don't hesitate to contact me: ben.ellison at panbo.com.

Posted by: Ben at June 16, 2009 9:15 AM | Reply

A technical question for the group about Class B AIS units. Do any of the Class B units accept ship's heading input for the AIS message 18/19 heading fields? That is from a fluxgate not the COG provided by the internal GPS. In reading through the Simrad AI50 manual it does not appear to.

Posted by: Richard D at June 16, 2009 2:53 PM | Reply

Richard the AI50 is definitely using and broadcasting heading info. You can see Gizmo's PB200 derived heading displayed on the AI50 screen at top and then included in the N2K AIS message parsed out in the middle screen shot. And reason the Gizmo target and boat icon are perfectly parallel on the bottom Garmin screen is that they're both getting heading from the same source.

I think that non N2K Class B transponders can receive heading via NMEA 0183 input, but I've never messed with it. I didn't mess with it in this case, either; it just appeared.

Note that in the course of these illustrations Gizmo did a 180 as I shifted her from my float to a berth at a very nearby bar ;-)

Posted by: Ben at June 16, 2009 3:21 PM | Reply

Ben, Thanks, and I reread your earlier posts and realize you had already answered my question. The wrinkle is that RED has a full compliment of 0183 instrumentation and the AI150 does not appear to accept heading via that channel.

Posted by: Richard D at June 16, 2009 4:48 PM | Reply

Ben, I am fairly new to marine electronics my specialty is in wireless network integration for large buildings, tunnels and campus environments to name a few. With that being said I have found the manufactures to be helpful as well as fairly responsive. I am completely sold on the idea and integration of N2K on my vessel. I have replaced almost every device available with N2K products. I purchased my Garmin 4212 prior to the release of the N2K kit which included the N2K GPS. When I decided to replace my 2nd 3210 with a 4210 that is when I started to build the N2K network the 1st replacement or I should say backup was the older GPS 0183 antenna. My 0183 network as well as my Garmin proprietary network is shrinking as the N2K grows. This week I am replacing the Garmin GDS-22 and transducer with a N2K transducer packaged and sold by Garmin. This gets rid of another large box and Garmin's proprietary network device. I have been quiet pleased with the number and ease of upgrading all my Garmin components using the SD card and their website. Next I imagine I may replace my KVH1000 position sensor with a N2K heading sensor. In the mean time with my Navico300L AIS B thinking my boats a dangerous target I just shut the transmitter off when we are tied up or at a marina and it receives AIS static data very well from other ships and the annoying alarm doesn't go off constantly. At least at this point both Garmin and Navico are aware of the problem and I expect a software patch may already be in the works. I am patient and know this will soon be worked out. I am not normally near the shipping lanes so using the Navico in receive only mode works and if I feel we are in the fog or cruising at night I enable the transmitter and just clear the alarm periodically. It will be interesting to see how the new Garmin VHF200 stacks up against the Lowrance LVR-880 I installed at the helm (replaced a Raymarine 240). The Garmin VHF200 will replace a Uniden 625 in the com center in my salon. To the engineer above who worked on the AIS150 I say thank you and it is always nice to hear from an insider.
Regards Bill Lentz
Wireless One
Tuckerton, N.J.

Posted by: Billlentz at June 18, 2009 9:06 PM | Reply

I just learned that the NMEA has published the new Class B AIS PGN messages. In fact they already did so on May 8, 2009.

Read all about it here: http://standards.nmea.org/corrigenda/nmea-2000/nmea-2000-corrigendum-1-2009.pdf/view

Posted by: Kees at June 30, 2009 10:45 AM | Reply

Airmar released it's N2K to usb gateway about a month ago.

Posted by: Sandy Daugherty at July 2, 2009 8:30 AM | Reply

I'm trying to get a sample of the Airmar N2K/USB Gateway, but suspect that the Actisense model I'm trying is very similar. I don't know that for sure, but if you open up an Airmar Combiner box -- which is essentially an NMEA 0183 multiplexer with a USB port -- you'll see "Actisense" printed on the circuit board. Plus check out the screenshot at the bottom of the Tacktick entry:

http://www.panbo.com/archives/2009/06/wireless_n2k_wind_more_tacktick_style.html

Posted by: Ben at July 2, 2009 9:26 AM | Reply

Dear All,
I am looking for a software that can generate some AIS data to emulate heavy traffic environment for another software that i am writing for Port Management. Is there any such software that can generate some AIS data based on the required inputs of an arbitrary ship (like start point, course speed etc. )fed by me initially. I am looking for this software to generate only the very basic position reports so that i can validate my other software/
See if any of you could help. please email me the required link..

Posted by: SP at September 8, 2009 12:45 PM | Reply

Check out Mats Kagstrom's AIS Simulator, not inexpensive but quite powerful:

http://www.panbo.com/archives/2008/02/kagstroms_ais_simulator_and_good_b_news.html

Posted by: Ben at September 8, 2009 3:38 PM | Reply

So, does anybody know if this issue is solved with the Simrad AI50, ie. does it now send the N2K PGNs for class-B targets?

tnx,
Nick.

Posted by: Nick at February 8, 2011 11:51 PM | Reply

Leave a comment