iKommunicate from Digital Yacht, Signal K gets kickstarted

... written for Panbo by Ben Ellison and posted on Oct 26, 2015

iKommunicate Signal K proof of concept aPanbo.jpgSorry for the fuzzy photo, but it makes sense as the iKommunicate Signal K Gateway isn't a real product yet, and the real product may look quite different from the "Proof of Concept" prototype above if it actually materializes early next year. Why so vague? What's happening here is a techy chicken and egg thing. Digital Yacht was understandably hesitant about developing a gateway for Signal K apps and services that hardly exist yet, but the Signal K universal boat data concept -- open source, entirely free, and created by unfunded volunteers -- needs a commercial level NMEA gateway so that third party developers can easily show us what the potential apps and services can do. That's my understanding of why DY launched the iKommunicate Kickstarter campaign and also why I'm hoping that fellow boat geeks and related companies will join me in backing the project and thus helping Signal K become real...

iKommunicate_Kickstarter_page_cPanbo.jpgActually, it's very likely that the iKommunicate Gateway crowdfunding campaign will reach its $20,000 goal, as it's already more than 25% there after just three days and little publicity. In fact, I think that Digital Yacht set the goal low and the campaign period long because the company is already committed to making the gateway happen. The idea of using Kickstarter is not just to hedge gateway sales against the uncertain commercial future of Signal K, but also to build a community of enthusiastic kibbitzers toward that future. And note that backers of product projects like this aren't exactly philanthropic; most of us are buying something that hasn't been fully developed yet so we can get early delivery at a discount, plus a little say in how it's developed or at least the right to claim "early investor" status (if and when that's something to be proud of ;-).

iKommunicate is the ninth Kickstarter project I've backed -- and I just signed on to the interesting Hello Ocean campaign on Indiegogo -- so I know something about how effective and enjoyable this new form of fundraising (and/or sales) can be. For instance, I enthusiastically followed the original Pebble watch Kickstarter as it raised nearly 11 million dollars from almost 70,000 backers, and then I became an evangelist for the actual resulting Pebble that had been so well developed and manufactured with all that backing. And while Pebble certainly did not need Kickstarter to develop the new Pebble Time, I was one of the many repeat backers to join the Kickstarter fun and now I'm much enjoying the second watch I got early and discounted.

iKommunicate_Gateway_system_diagram_aPanbo.jpgThere are crowdfunded product projects that produce crappy results or never deliver at all, but the danger with iKommunicate is more that Signal K doesn't work well and/or isn't used by many developers to create useful apps and services. That's why DY plans to make the iKommunicate Gateway capable of outputting NMEA 2000 and 0183 data through your boat router in the same way that boat data hotspots like the Navico GoFree, the Vesper XB8000, the Brookhouse iMux and others do. This means that early iKommunicate users should be able to get basic boat data like GPS, Wind, Depth, and AIS to apps like iNavX, iRegatta, and others (like I did last summer with the TimeZero V2 app) regardless of SK.


But I think that Signal K will really happen - and perhaps spectacularly. Besides from the iKommunicate Kickstarter and the Signal K site, you can learn more about the concept from my May entry about the friendly new NMEA SK relationship and from Bill Bishop's recent musings about how SK actually fit in at the NMEA Conference. I heard concerns about Signal K data security routines that need definition, and the number of N2K PGN's that still need to be translated into SK syntax by volunteers who have day jobs, but it's terrific that this ambitious open source idea is now in the same room with the major and minor makers of marine electronics and software. And at least some of them are interested, as I'll relate below.

First I want to describe my hurried though eye-opening experience with a Signal K demo given by Digital Yacht CEO Nick Heyes (who was a NMEA new guy in 2008). What you're seeing in the photo above is dummy vessel Maverick's NMEA 2000 data being gatewayed by the prototype iKommunicate and WiFi'd to an app which is turning the SK info into an instrument and mapping display. While you could say, "So what?" because this is already fairly easy to do, the missing image would show the iKommunicate SK screen where the boat user can enter a boat name like "Maverick" and a raft of other information that currently has no place to live in NMEA data standards. So Signal K adds a rich data layer to regular NMEA 0183 and 2000 (which it also translates better), and all of it can be available to app developers as you see fit.

ActiveCaptain_Locations_w_Signal_K_cPanbo.jpgThe ActiveCaptain app Locations is one of the first to support Signal K and a good example of why. Locations has been around for a while, and in order to fully identify nearby AC users or special friends it needs both profile info from eBoatCards and current location data preferably from somewhere else. I use a link between eBoatCards and my test inReach Explorer, which AC makes easy, but many eBoatCarders just input their location manually, which is why my friends on the S/V Makana aren't really moored in Camden today like they were last June. Now Locations 1.2 can get GPS info from "0183 info over IP" WiFi sources like the ones I mentioned above, but better yet, I'll bet users with an SK connection will soon be able to share info like what anchorage they're headed to or even the route. What's more, Signal K means that other developers can create something like Locations without having to build something like eBoatCards first.

At any rate, AC's Jeff Siegel was enthusiastically present at the NMEA Conference, and I suspect that iKommunicate will receive some AC love soon. Also present and grinning was Luis Soltero of RedPort and GMN, who has embraced Signal K too. He envisions how the richer, more app friendly SK protocol can help make boat data more useful over the narrow band satellite communications he specializes in. I've hacked up his diagram to fit the space below, but here's a tangible example: Picture an AIS man overboard device alarming not just the MFD system on your boat but also auto sending a report ashore using a modest device like an Iridium Go!

Not much is concrete yet, but it's encouraging to see serious marine developers like Digital Yacht, Siegel and Soltero at least tentatively get behind the already impressive Signal K crew. That's why I'm backing iKommunicate and maybe you should consider it?




Can't wait to use it on my sunlight readable Idevice

Posted by: Hendrik at October 26, 2015 5:31 PM | Reply

Cool, the iKommunicate Kickstarter campaign just passed the 50% mark. But note that it won't end when the funding goal is achieved. In fact, still left are 59 of the least expensive, earliest shipping gateways.

PS. Hendrik, I'm with you in appreciating dedicated marine multifunction displays, but I do think that Signal K may encourage more 2nd tier high bright displays because it can much more than just repeating NMEA 2K and 0183 data.

Posted by: Ben at October 26, 2015 8:20 PM | Reply

I am in, this is a no brainer.

Posted by: Dan Corcoran (b393capt) at October 27, 2015 12:05 AM | Reply

Great, just got a new boat so I hope to be able to test it next sailing season!

Posted by: Yme Bosma at October 27, 2015 4:14 AM | Reply

Posted by: Bill Bishop at October 27, 2015 6:16 AM | Reply

I'm in and hopefully this will have some legs ... finally an open marine data exchange format.

If SK does gain traction, I hope a certification program is intended.

Posted by: Sparky at October 27, 2015 4:35 PM | Reply

Sounds like the iKommunicate Gateway allows you to keep the “old antiquated” NMEA 2000 network and still move ahead with Signal K apps. Do I have that right? I just spent a lot of money upgrading and replacing all the mismatched drop cables and “T’s” with new high quality equipment on my network.

I’m looking forward to hearing more about Signal K.

Posted by: Richard C at October 27, 2015 8:46 PM | Reply

Holy Cow! ActiveCaptain showed the Signal K love...


...and now the iKommunicate project has exceeded its goal 3X with nearly 300 backers. There could well be over 500 before the Kickstarter campaign ends, and that means there will be a significant SK test group when the iKommunicate gateways start shipping next year.

I imagine that this success is very encouraging and motivating for the volunteer Signal K development team, the Digital Yacht team, and the various developers like AC who are intend on making SK do great things for us. All good!

PS Richard, SK is definitely designed to work with NMEA 2000 (and 0183). In fact, I believe it will make your new network more valuable by making all your NMEA data easier to access with mobile and Internet technology plus allowing you and/or apps to add data not available in NMEA standards.

Posted by: Ben at October 28, 2015 11:36 AM | Reply

Reading through the use cases, a suggestion for sailing ... How about an application that takes boat information combined with boat performance information, allows the user to specify which sails they are current using, and pipe back into the boat a speed in knots for target boat speed that could be brought up on an instrument display that may only know how to display boat speed ... but by selecting a boat speed source, meet the needs of communicate to the boat owner they are falling short or coming close to the potential of their boat as they make each sail trim adjustment. This would be a good example of the 2-way capability of signal K.

Posted by: Dan Corcoran (b393capt) at October 29, 2015 12:44 PM | Reply

That's a cool idea for an app you have there, Dan! Why don't you suggest this on https://groups.google.com/forum/#!forum/signalk

Posted by: Kees at October 30, 2015 4:04 AM | Reply

I would love to see an auto-polars iPad/android app that would display the polar graph and plot the boat position on it. This could integrate the theoretical polars for your boat class, real "this boat" polars as well as current boat performance. Since a huge amount of data can be logged in 4GB, and even more in the SD card, there is no reason to discard instrument data for the entire sailing season...

Posted by: Jason Taylor at October 30, 2015 10:26 AM | Reply

I imagine the unit will use their new packaging and look like the AIT2000 (http://www.digitalyachtamerica.com/index.php/en/products/ais-systems/ais-transponders/product/32-ait2000-class-b-transponder)... at least I hope it does, this is a very nice case.

I am looking forward to testing this unit.

Jason, you could always write the app that you want... or at least suggest it in detail on the above mentioned link by Kees.

Posted by: Chris Ellingsen at October 30, 2015 12:09 PM | Reply

400 backers and climbing. Already far exceeded the goal for funding. Impressive.

This seems a pretty strong statement that the user base wants an open data exchange format.

Posted by: Sparky at November 6, 2015 9:53 AM | Reply

Lots more iKommunicate gateways still available at $70 off retail price, but Kickstarter campaign ends on Dec. 15


Posted by: Ben at December 12, 2015 11:03 AM | Reply

Congratulations to DigitalYacht for a successful Kickstarter! I'm looking forward to trying out my device!

Posted by: Graham Collins at December 16, 2015 9:23 AM | Reply

Very cool! A smart (and funny) guy named Adam Hyde has started a blog about Signal K and already he can't help but write about other neat marine tech:


Posted by: Ben at February 19, 2016 8:17 AM | Reply

vyacht released new firmware at the end of January implementing a Signal K server on the vyacht router so when iKommunictae is delivered we'll have at least two commercial Signal K options. 2016 might be a very good year for Signal K.

Posted by: Keith at February 24, 2016 5:11 AM | Reply

Hi Keith,

I appreciate your enthusiasm but I just heard about vYacht and I feel a little skeptical about the claims being made.

"The {vYacht} NMEA 2000 Gateway will translate all known NMEA 2000 sentences to NMEA 0183" jumps out at me, for instance. There are many known N2K PGN's that have no corresponding NMEA 0183 message so that statement seems doubtful.

Moreover, I'm not sure that any developer should be making an N2K gateway unless they are at least a member of NMEA and hopefully willing and able to get the gateway certified. vYacht doesn't seem to even be a member. (I guess it's ethically OK to use N2K if you reverse engineer it, but that's a whole lot of work that I doubt vYacht did on its own.)

Also, I first heard of vYacht from a Panbo reader complaining about a lack of technical support:


I'd be happy to learn different, but as it stands I don't see how vYacht compares to iKommunicate, where a known team of NMEA experienced developers are working with the Signal K team to make all of NMEA 2000 accessible via SK coming out of a certified gateway.

Posted by: Ben in reply to Keith at February 24, 2016 7:29 AM | Reply


For clarity, I'm not personally connected with vyacht but was interested in the product which seems to have positive reviews from people on cruisersforum and the ybw forums: Like others I've been following Signal K development and talking to some of the people developing products about their offerings and more specifically their views on Signal K. In the first link above you'll see that I asked vyacht about NMEA certification and you're right: public / reverse engineered documentation has been used, but then that's the case for the majority of mobile marine data apps available today as well as a few hardware multiplexers (though not obviously those from actisense/digital yacht/shipmodul who are all NMEA members).

"Grotty Yachty"'s problem which he also posted on cruisersforum is the only negative I've seen about vyacht, and that's not actually a fault with the product: it's an unrelated networking issue.

It'd be interesting to hear the experiences of other pan readers who've purchased a vyacht.

Posted by: Keith at February 24, 2016 8:56 AM | Reply

Thanks, Keith. I admit to not having read your interview with "Bernd from vYacht" thoroughly but now I have. Nicely done!

I think that Bernd is confused about proprietary standards, conflating open standards (available to any company) with open and free standards (that are generally financed by public institutions and/or very large corporations). NMEA 0183 and 2000 are open standards, but not free, and the same is true of many similar industrial standards, especially CANbus based.

He's right that a fair number of NMEA 2000 devices from larger marine electronics companies are not yet certified (which is odd), but those companies all contribute to the NMEA standards-making process with volunteered employee time on committees, corporate membership fees, advertising in the Marine Electronics Journal, etc.

And, incidentally, don't the margins on retail marine electronics have to include those sorts of expenses, as well as the high cost of technical support? I still think that vYacht and iKommunicate are apples and oranges.

Posted by: Ben in reply to Keith at February 24, 2016 9:35 AM | Reply

Thanks Ben

I don't think there's confusion about "open" vs. "free" there. Neither NMEA-0183 nor N2K are "open": NMEA-0183 has been widely publicised, to a large extent by gpsd author and open source advocate Eric Raymond, but it's still a "closed" protocol and its use in OpenCPN, the reference Signal K servers and many commercial marine apps where the authors haven't purchased the official standard is still regarded (as I understand it: correct me if I'm wrong) as dubious. "free" is of course a nuanced word and for those who get passionate about the subject (as I'm sure you've seen happen), "free-as-in-speech" is the important angle rather than "free-as-in-beer".

Personally I see the value of a controlled whole-stack standard and certification process where safety and reliability are primary concerns. I also see how open standards stimulate innovation and give people the freedom to tailor their environment to their needs. It's nice that we now have choices with Signal K, with options from the DIY of the reference servers to the NMEA-certified iKommunicate, options which can be chosen between according to technical ability, depth of pockets, software freedom concerns or the security of backup from an established and well-known company. It's a shame we don't have to-the-transducer open networking options yet but it'll be interesting to see if anything falls out of the current IoT bunfights that will be relevant in our marine world.

Posted by: Keith at February 24, 2016 11:44 AM | Reply

Well, at least we agree that using an open standard without also honoring attached documentation or licensing fees is a dubious practice ;-)

Seriously, I don't think it takes much research to determine that proprietary, open, open and free, and open source are all distinctly different ways of establishing standards and that there are many valid examples of each. CANbus and its many higher level implementations is a great place to start:


Seatalk is a proprietary standard (though "closed" seems to be the term you're using now and I'm not sure Raymarine cares any more). NMEA 0183 is an open standard with documentation fees (but not licensing fees), NMEA 2000 is the same plus certification requirements, and Signal K is open source. I believe that USB and Bluetooth are examples of open and free standards we use around boats.

Of course open and free is great when possible, and open source seems the ultimate in innovative possibilities, but I also care a lot about accurate terminology, which in this case has nothing to do with free speech or beer.

Posted by: Ben in reply to Keith at February 24, 2016 3:32 PM | Reply

The term "open standard" is defined differently by different bodies and many of those definitions, including the legal definitions in some countries would not include standards like NMEA-0183 which are not available to read without payment of a fee.
I believe that the NMEA deem use of their standards by those that have purchased them in open source software to be unacceptable as it constitutes "reproduction", but I have been unable to get a reply on this from their legal department (any up-to-date info on that appreciated)

Other definitions of "open standard" do include standards available to all on payment of a fee and with restrictions on use.

This is in no way a criticism of the NMEA's business model or an attempt to make any value judgements of what kind of model is "better", I'm simply arguing that the term "open" is subject to cultural variation. NMEA protocols might be classed as "open standards" in the electronics engineering world, but you might not find many software engineers who would describe them as such. We're in interesting times where those two worlds are coming together...

Posted by: Keith at February 24, 2016 5:00 PM | Reply

The problem with lumping together "Open" with "Open and Free" seems obvious to me. It also lumps together "Proprietary" -- a standard that no one can use unless allowed to by a certain company or institution -- with "Open" to anyone though including reasonable fees needed to maintain and grow the standard.

I don't think I'm biased about this either, except toward Free and Open Source as previously stated, but I suspect that fudging what seem like meaningful and useful categories by some bodies and nations smacks of motive.

Some of the Panbo readership and I have been at this rodeo before...


...and it may be time for me to move on. The good news is that NMEA fees have gone down since then and, maybe more important, the gap between NMEA standards and the open source world seems about to be properly bridged with Signal K. Maybe you should ask the SK team what they think about the ethical issues?

Posted by: Ben in reply to Keith at February 24, 2016 6:01 PM | Reply

My iKommunicate Signal K Gateway just arrived at my door and is looking good!

Delivery was later than originally planned -- common for Kickstarter projects! -- but it also seems like communications got a little spotty. At any rate, Digital Yacht now has a Twitter account for iKommunicate:


and also a forum:


Personally I won't be able to test it much in the next few weeks (Gizmo launch is horribly delayed) but here's hoping we hear from others.

As T Kurki just noted in the Panbo Forum where SignalK sort of got its start:

"Well, here we are, a few years later. Now that Digital Yacht's IKommunicate is shipping we have an off the shelf, NMEA 2000 and 0183 certified solution for getting all the NMEA data onto an IP network...

So let the applications be built!"


Posted by: Ben at June 2, 2016 12:21 PM | Reply

Also received mine. I was hoping to get it set up for the Annapolis Bermuda race, but there were higher priorities.
I don't like that the LEDs and cables are on opposite ends of the box.It needs to be mounted so you can access the cables and see the LEDs.
No OpenCPN support for SignalK in the foreseeable future and limited time for an ethernet cable run. Being an early adopter is not always a good thing.

Posted by: Joel at June 2, 2016 4:00 PM | Reply

I'm a little late to the discussion, but to me, N2K is a proprietary standard in every way. Even using Ben's definition in the following supports this notion:

Ben said "The problem with lumping together "Open" with "Open and Free" seems obvious to me. It also lumps together "Proprietary" -- a standard that no one can use unless allowed to by a certain company or institution -- with "Open" to anyone though including reasonable fees needed to maintain and grow the standard."

N2K is "owned" by a single entity, NMEA

N2K is confidential and not accessible to anyone who does not hold a license.

Anyone wanting to see the standard has to purchase a license. But then the licensee of N2K is obliged to hold the standard and all associated documentation confidential.

That's about as proprietary as proprietary gets. Heck, even patents, which are all owned and licensed, are published for anyone to read.

NMEA is non-discriminatory about who they will license to, but there is nothing open about contents of the standard. It's not like any of the IP protocols, all of which are published and accessible to anyone without a fee. And all the 802 standards are similarly accessible to anyone, though one needs to purchase the book(s).

Posted by: twistedtree at June 2, 2016 8:39 PM | Reply

Hi Joel,

The external LEDs and cables are at the same end of the iKommunicate enclosure, so I was a little surprised at your comment. Sorry we did not get it to you in time for the A to B Race, but hopefully you will be able to use it for the rest of the summer.

I have been in contact with some of the OpenCPN developers and I know that they have been doing some initial implementation work and are keen to support Signal K in the next major release. Now that iKommunicate is released and Signal K is "almost" at Version 1.0.0 release, hopefully this will be the catalyst for them to finish their work.

If there were no early adopters, there would be no products, so please keep adopting ;-)

Posted by: Paul_DY in reply to Joel at June 3, 2016 6:16 AM | Reply

Wow, that's some twisted logic, Mr. Twistedtree! By your analysis, ABYC, CE, SAE, AIS, GMDSS and many other standards are all "as proprietary as proprietary gets." That will come as a big surprise to many boat, marine electronics, and engine manufacturers.

And a published patent showing them intellectual property they can NOT use is somehow more "open"? How in the world would you characterize the Mercury Smartcraft engine control and monitoring standard?

And, finally, do you realize that you're making your twisted comment on a thread discussing SignalK, an open and free standard that's somehow worked with NMEA to make N2K available developers quite like the IP standards that were financed by large corporations, universities, and governments?

Posted by: Ben in reply to twistedtree at June 3, 2016 8:02 AM | Reply

Hi Joel,

The finished iKommunicate box design is quite different than the "Proof of Concept" design shown in the entry above, but I think you misunderstood it. That's a wire color identity list for the NMEA 0183/Power cable at one end, not LEDs:


Meanwhile the working end has four data comms LEDs, the Ethernet port, and two 2.4 foot pigtail cables. One is for the NMEA 2000 port and the other holds those NMEA 1083/power wires. Here's a photo, though the cables aren't shown:


Personally, I'd prefer having the LEDs on one end and the cables/ports on the other, but in practise I don't think it will be important either way.

Posted by: Ben in reply to Joel at June 3, 2016 8:24 AM | Reply

Wow, here we go again with different views of what's "twisted" thinking.

As one of N2K's proprietary characteristics, I noted that N2K is owned by a single entity (NMEA) and only accessible to people who have licensed it. Back in Feb you characterized "proprietary" as "....'Proprietary' -- a standard that no one can use unless allowed to by a certain company or institution...". So we seem to agree that this aspect of N2K is proprietary. This is very different from ABYC, CE, SAE, IEC, NEC, UL, IETF, ISO, etc. None of those organizations license their standards. They are published and anyone is free to practice it, and the contents of the standard are not a secret held by licensees. Anyone can purchase the publications with no strings attached, practice their content.

And patents require a license to use, just as N2K requires a license to gain access to the standard. So in practice you can't implement an N2K product without a license. But a patent is published for all to see, not just licensees. In contrast, N2K is not published for all to see, and anyone who acquires a license to gain access also agrees to maintain secrecy of the contents.

So I stand by what I said and don't see what's twisted about it.

Posted by: twistedtree in reply to Ben at June 3, 2016 8:41 AM | Reply

All those standards are owned by a single entity, the standards making body, and all those standards documents are copyrighted, which is a license. NMEA claims that N2K is an open standard by which they mean that any person or company can buy it and use it. Twistedtree, do know of any person or company who has not been allowed to buy and use NMEA 2000?

And why couldn't you answer my question about Mercury Smartcraft, which really is a proprietary marine standard? There's nothing wrong with that, but do you think that you can buy and use it for control and monitoring of a new engine brand?

But please continue this discussion in a more appropriate entry like:


This entry is about the iKommunicate SignalK Gateway, which just materialized and deserves a conversation free of this bickering. Thanks in advance for moving.

Posted by: Ben in reply to twistedtree at June 3, 2016 9:23 AM | Reply

Twistedtree, this is an exercise in semantics that has been well thrashed here and many other places for a long time. Few things in life are free, and organizations like ABYC, NMEA and the other similar standards organizations you listed and despite the large amount of volunteer help they get all need income from some source, and there are lots of ways to do it. They need phone lines, office space, employees, printing and so on. Even Signal K has need for some modest income. You want a copy of ABYC's standard? You pay for for them. They are copyrighted so you can't disseminate them. They are for your personal use. How is this different from buying NMEA's standards? One is a copyright, the other a licence. Although there are some legal differences, from the users viewpoint they both take you to the same place, and neither organization will let you see the materials before you pay for them. If you let people read them first, the income disappears, and then the organizations follow.

Posted by: Bill Bishop at June 3, 2016 9:33 AM | Reply

Last time I brought up something critical of N2K I got chastised and told to take it somewhere else. I see this time is no different even though my comments directly relate to the discussion in this thread, although they are a bit delayed as I acknowledged in my original comments.

Yes, anyone can license N2K. I said that in my very first post as well. And I never voiced an objection to paying for standards publications or even licenses per se. But that doesn't change the constraints imposed by the N2K license which are fundamentally different from the constraints imposed by a copyright standards spec like ABYC or any of the others discussed so far. It's not semantics, it's fundamental differences in what's allowed and disallowed. Call it open, closed, proprietary, good, bad, or Fred. It doesn't change the licensing terms.

But per your request, I'll take it up elsewhere.

Posted by: twistedtree at June 3, 2016 12:08 PM | Reply

So instead of taking his N2K issues to a more appropriate Panbo thread, "twistedtree" (who I forgot is Peter Hayden) went to his own site:


I still don't understand what he's gotten so worked up about, but his post is an opportunity for Ben bashing ;-)

Posted by: Ben at June 4, 2016 7:02 PM | Reply

What's really interesting, I think, is the interplay of NMEA companies like Digital Yacht and Signal K. This week I had email with Paul Sumpner (who heads up the iKommunicate project) and he told me:

"​I have published an iKommunicate Developers Guide at https://github.com/digitalyacht/ikommunicate which includes some simple Apps (with source code) that I wrote. Bearing in mind I have no previous knowledge of writing web apps and I am a hardware designer, not a software developer, it proved to be surprisingly simple and if someone knows what they are doing with HTML5 web design, then the skies the limit."

He also said that DY is "aiming for a fortnightly firmware update on iKommunicate, at least to start with as we get feedback on features, conversions, bugs, etc." and that they're hungry for NMEA2000 log files. Logging is built into iKommunicate, though it's currently cumbersome (as in using Telnet Terminal Client to enable and disable it). But Paul said that one of first new features planned for iKommunicate is logging control from the Web interface.

Who's got their iKommunicate running?

Posted by: Ben at June 4, 2016 7:39 PM | Reply

Ben, Thanks. I'll take another look at it.

Posted by: Anonymous in reply to Ben at June 6, 2016 10:50 AM | Reply

Mine is up and running and I love it! I wrote up a short article on unboxing and installing, along with lots of photos and some screen shots at https://www.sailbits.com/blog/2016/07/ikommunicate-unboxing-install/

This is definitely the future of marine electronics interconnectivity, and I'm so happy to see it in a commercial product.

Posted by: Steve Mitchell at July 6, 2016 2:30 PM | Reply

Have all of these shipped? I was backer #48, have not received.

Posted by: Dan Corcoran (b393capt) at October 3, 2016 7:55 PM | Reply

Dan, you should definitely drop them a line. I believe they all shipped out in June.

Posted by: SteveD at October 4, 2016 4:49 AM | Reply

Hi Dan, no need to drop us a line, I will investigate this immediately as I thought we had shipped everything, except a few units where people were out sailing and wanted us to delay shipment.

We will get this sorted ASAP. Sorry for the inconvenience.

Posted by: Paul_DY at October 4, 2016 4:54 AM | Reply

Hi Dan, Nick has emailed you again this morning as we need the shipping address to send your unit to, and then you will get your iKommunicate later this week.

Now that all of our Kickstarter Backers have received their units, iKommunicate is part of our normal product catalogue and available through all of our normal sales channels, priced at $299 USD.

Posted by: Paul_DY at October 4, 2016 5:25 AM | Reply

Paul, please feel free to brag!

The latest news is that Digital Yacht has developed an iKommunicate instrument app that runs on inexpensive, sunlight viewable Kindle tablets:


The app is open source so any developer can mess with it, and a working version is included in the latest iKommunicate SD card image (V1.60 dated 28th September):


I don't know much about Kindles but I think that a 6" 5th Gen with E Ink display and WiFi, available used from Amazon for $30, would work fine:


Posted by: Ben at October 4, 2016 8:32 AM | Reply

Thanks Ben, for the Kindle App mention.

You are right any of the latest Kindles can run the app and it is pretty straight forward to take the source code and modify what data you want to display.

Now that iKommunicate is released, I am trying to create a few simple but interesting reference designs to show what is possible now with Signal K and perhaps fire the imagination of other developers out there that want to take the source code and expand on it.

It is amazing what HTML5 allows a web app to do and when developed, web apps can be copied to the iKommunicate SD Card, allowing them to be run by anyone onboard with a mobile device.

Watch this space !!

Posted by: Paul_DY at October 4, 2016 9:15 AM | Reply

I'd like to hear why SignalK is "better" compared to NMEA 0183. To transfer the position of a ship, I need one NMEA sentence of 77 characters. To transfer the position of any ship, add a TAG block of 15 characters to add its IMO number. To do the same in SignalK, I need over 650 bytes (!) requiring a lot more parsing of field names and opening/closing tags of fields and groups. Even from a software engineers point of view, I'd rather parse an NMEA sentence than a SignalK info block.

If the only advantage is that you don't need to pay for the SignalK definition, than you're not making the right business decision. The time spent on writing a SignalK parser is easily exceeding the cost of the NMEA 0183 Document.

I'd love to hear other views though :-)

Posted by: Meindert Sprang at October 19, 2016 2:32 AM | Reply

The verbosity of Signal K and inappropriateness for constrained devices has been discussed in plenty of places. Its target market is enabling app programmers in high level languages on 802.11 or ethernet networks. It's model which seems to have been adopted in quite a few commercial home automation products. Why might we not use NMEA-0183? Partly because it's a point to point protocol and not defined for use over IP networks (even if people are using it that way now, and thanks for at least standardising the port, if not the transport for us :-). Of course Signal K doesn't solve that problem either: it's a data format without a defined transport. You mention TAG blocks: they're not well documented in non-NMEA documentation as that was largely written before V4. So you'd have to buy the standard. But if you buy the standard you're bound by non-disclosure and can't use the information you just bought in open source software: so you couldn't use a purchased version of the official NMEA-0183 standard in OpenCPN for example. People will also point you at the extensibility of Signal-K. If you don't think of Signal K as an alternative to the NMEA protocols you'll find less to dislike about it. Of course that means that we still need a free (as in speech) direct alternative to the NMEA protocols for constrained devices and networks...

Posted by: Keith at October 19, 2016 4:06 AM | Reply

Meindert, your right, to a point. NMEA 0183 is very easy to parse. It's just an ASCII string with commas used as deliminators. It's also a 4800 baud point to point comm system that appeared 36 years ago (1980) in the days of the Comodore Pet computer. That being said it's a durable protocol that is still supported by most manufacturers, although at a lessor level each passing year. As an installer I still interface to it but mostly to communicate with legacy boat gear such as autopilots. So if your data needs don't require speed, and the number of interfaces are few NMEA 0183 works very well. In the real world I install more WiFi access points and routers on boats than NMEA 0183 interfaces nowadays.

Signal K was specifically designed to be app developer and cloud based applications friendly. It had several goals. Making data easily accessible to mobile devices was the first. Via a gateway the data is already parsed into individual JSON strings meaning no licences needed for the data use. Beyond that Signal K's easily extendable data model not only contains NMEA data, but also data never contemplated by either NMEA 0183 or 2000 such as performance racing data, vessel related
data like sail inventories, GPS offsets, and owner data. Signal K has become the defacto open source standard for moving data off your boat and into the cloud.

NMEA 2000 has some additional benefits, it's 50 times faster than 0183, can connect to far more vessel devices, it's plug and play, and supports more vessel data than NMEA 0183. So it's really about your needs and both approaches work well. A bicycle is simple and easy to use, but it's not as practical as a car if you have to travel far, or need to go faster.

Posted by: Bill Bishop at October 19, 2016 8:05 AM | Reply

Hi Meindert,

Not sure where you're getting 650 bytes from. Yes, JSON is verbose, but it is self-describing and compresses very well. A well formed transfer contains many parts that are either optional or shared with other data coming from the same source.

Whether a Signal K parser is difficult depends on your point of view. A JavaScript developer will love Signal K since JSON is just valid JavaScript, and the parser is already written (= it is the interpreter/compiler itself). Other languages that allow dynamic class construction such as PHP, Ruby or Python have built-in parsers as well. There are public free implementations of JSON parsers for every computer language worth its salt out there.

Is it the most suited format for a tiny microcontroller with a few KB of RAM that would need to *parse* it? No. Solution: get a bigger CPU that can handle this. For instance, the new Simrad IS42/B&G Triton2 instruments run Linux(!) so that they don't have to rewrite AIS handling in a too-small-system.

Would it be possible for a small sensor mc to *generate* it? Sure. The Signal K team hasn't focused on that, but they are getting requests from various developers that would like to build sensors, so this is an area of further expansion.

For apps developers making small software applications, the cost of doing business in the NMEA world -is- a consideration. The biggest problem however is that once they buy the standard they cannot make their software open source, as that would violate the IP. SK gets around that by having hardware vendors that are NMEA members read NMEA data and generate Signal K. Digital Yacht has seen a market, apparently, and I don't think they regret it.

If you are interested in discussing Signal K, maybe a better place to have this discussion would be https://groups.google.com/forum/#!forum/signalk or http://slack-invite.signalk.org

Posted by: Kees in reply to Meindert Sprang at October 19, 2016 8:31 AM | Reply

I think there are a number of forces at play here, but fundamentally Signal K exists because it solves three problems that NMEA does not. It's easy to program to, has a defined transport over IP networks, and is open in a way that N2K is not.

First, SignalK is a more convenient format/structure for app developers. Making it easy to program means more apps. That's good.

Second, there needs to be a way to get nav data on an IP network in some standard way. Signal K does this, but it's by no means the only way. In fact, OneNet is NMEA's effort to standardize the transport of N2K (and maybe 0183?) over IP. Once that is complete and stable, it will create a competing data transport mechanism to Signal K. And there already is a defacto standard for transporting 0183 and N2K over IP used by most of the products out there today. Who knows which will win, but I think Signal K has the advantage since app developers will favor whichever is easiest, and I think that will be Signal K.

Third, as Keith pointed out, and has been discussed before, the N2K specification is accompanied by a mandatory non-disclosure agreement. So anyone who gets the specs and sees how N2K actually works, is also obliged to keep it a secret. That means they can't discuss it in a technical forum, can't release open source code that implements it, can't correct open source implementations that are incorrect, etc. It's a rather unique set of terms that I have never seen in any other standard specification, resulting in a closed spec (it's contents are confidential), yet open to anyone who wants to joint the group who know the secret and become part of the secret. Signal K solves this problem by recasting all the NMEA data in an open format that can be viewed, discussed, critiqued, and evolved by anyone.

One could easily create an open source parser for NMEA 0183 and NMEA 2000 (in fact some exist today), but because of the non-disclosure that accompanies the official N2K spec, any open source implementation is the result of reverse engineering rather than following a spec, so its validity and correctness will always carry some uncertainty.

I N2K were not encumbered by a non-disclosure agreement, and if OneNet had happened in a more timely manner, it's entirely possible that Signal K would never have emerged, but nature abhors a vacuum, and Signal K stepped right in.

Posted by: twistedtree at October 19, 2016 10:24 AM | Reply

Who decided it was a horse race with a winner and a loser? I don't see it that way at all.

I think it's great that Signal K was created and that NMEA tacitly approved the bridging of N2K data to SK, and vice versa. I'm just testing some SK apps now and I think the future is promising.

But most boaters use marine hardware/software systems, not just apps, and it's primarily those developers who are creating OneNet. We haven't even seen what OneNet can do on boats, and it will take a while, but I think it's quite possible that both it and SK will progress quite a ways and may in many situations be complimentary.

Posted by: Ben in reply to twistedtree at October 19, 2016 11:34 AM | Reply

OneNet is a adaptation of N2K over IPv6/UDP/Ethernet. It should be a simple technical matter to support OneNet translation with SignalK. Its the same N2K message decoding.

Posted by: Sparky at October 19, 2016 12:05 PM | Reply

@twistedtree: perhaps just our differing terminology, or maybe things have changed recently but Signal K specifically doesn't define a transport. This was a subject of some debate on the Signal K google group last year and the Signal K team wanted the format to remain transport neutral. In practice a reliable transport is assumed with websockets and http the de-facto standards. Someone is welcome to correct me on that if incorrect.

Posted by: Keith at October 20, 2016 3:30 AM | Reply

I agree with you Ben, the existing NMEA formats do a very fine job and will be around for many, many years to come. Signal K provides a new, interesting and accessible layer of information and technology that will compliment and enhance the NMEA systems found on every boat.

Posted by: Paul_DY at October 20, 2016 7:26 AM | Reply

Leave a comment