New Chartplotters and Features

[Jackdaw]

We're going to have to agree to disagree. 30 years ago (in the early days of NMEA 0183) who would have thought that microwaves would be found on sailboats. Now, they are pretty common on new boats.

With networking, transmission speed, bandwidth, and ease of installation will rule the day. Sonar and radar require much faster transmission speed and bandwidth than N2k can support. Wifi connectivity is pretty standard on new devices, and that uses an ethernet protocol. Larger recreational boats will have video monitoring of different parts of the boat. My chart plotter (Zeus2) already supports video, albeit, through a separate cable. Certainly Flir wants to sell more cameras and an ethernet based network will broaden their market. Instead of 4 or more connections on the back of an MFD, there will be one or two, power and ethernet.

Initially the OneNet devices will be pricey, however, like all technology with time the price goes down and the technology improves. Small start ups are already producing NMEA to Ethernet bridges and my MFD is home and connected to my home network via ethernet.

For many sailors all of this will be overkill, however, simplicity and familiarity for the end user and increased capacity will drive the market.

Have to agree. While NMEA2k took FOREVER to be fully standardized and put in widespread use, its a total success now. But it is not without it limitations. Tied to slow and inflexible network topology, its already showing its weakness. As noted all NMEA 2k RADARs use a proprietary side-channel Ethernet to deal with the data-rate needs. And current NMEA messages and topology are ill-sited for streaming, video and large file types because it is not packetized. Access for non-NMEA devices? Companies will create plug-in gateways, just like they do now.

A NMEA standard for wired and wireless IP-style marine data will be a boon.

 

[Davidasailor26]

A NMEA standard for wired and wireless IP-style marine data will be a boon.

Definitely. And once you start having wireless devices on the network security does become a worthwhile consideration. While I might not worry much about someone plugging into my NMEA 2000 bus, I would want any wireless devices that can send data to my chartplotter to be properly authenticated.

 

[Pizzazz]

Time will tell. May be a good example would be the stereo aftermarket for cars. New cars and new boats adopt vendor specific solutions, while the aftermarket has much more capability and faster refresh cycle (mp3s, video, cameras). As long as you stay close to a standards based solution (USB, Ethernet, serial ports), you can get this advanced functionality. Otherwise you are at the mercy of the vendor. Some people like the simplicity of a vendor solution. Others want the functionality. There will be a market for both.

I think the biggest mistake vendors made was not to lock down the radar protocols which allowed PC control of recreational radar. I think it is a matter of time before we see a good, quality screen mfd solution from a third party that blows away the functionality of vendor solutions (Open plotter). Nmea 0183 and Ethernet make this possible. Nmea2000, oneNet and certification only hinder this process. I could see the case for locking down autopilot control, in the same way car vendors lock down engine ECU control but anything else is just being greedy. Think about how you prepare for a passage. You do it on a PC, with all the weather data available from different sources, integration with an SSB or Inreach or Go, maps, pictures, guides, the web, etc. Or the simplicity of a weather routing app on Android. The is no way an Axiom can compete with this. All real racers use PC solutions. I like the reliability of a vendor mfd, it always works but I spend most of my time in front of the PC. I do not need more vendor lock-in. Regarding security, it is typically done with a firewall, you do not need end to end encryption, it just adds complexity (and it can be easily done open source as well).

While I agree that NMEA2000 is well established today, in terms of market share on all boats, I think nmea0183 and Seatalk (also serial based and possible to debug by a human) have a much bigger share than nmea2000.

Interesting discussion. I would not mind paying Raymarine some money for apps from there app store but I wish they open up the APIs, so that third party vendors can design really useful apps for video, wether routing, music players, visualization, integration with satellite services, etc. May be they will make more money this way.

 

[Gunni]

Really enjoying this discussion as I am in the process of further build-out of 10 year old Navico networked system. Half NMEA 2000, half Seatalk. Adding Navico AP to existing MFD/Radar/AIS/Engine monitor. The existing wind, boat speed, depth remain RM seatalk and I'm not interested in fixing what ain't broke. It works, very well. An ABYC Navico shop will do the work, but I will control the scope and plan. Too much of the tech that is offered is not user friendly, I am the only person who can fully use the existing MFD, and it will get worse with AP's that are controlled by submenus on the MFD. So reaching 100% driver navigation usage is more important than a bunch of wiz-bang features. Wish more manufacturers understood that.

 

[jon hansen]

i am enjoying this read. i know nothing of all this electronic wizardly. i have no microwaves on the boat. first thing i do when arrive on the boat is turn my phone off and put in a drawer.
it's fun to learn what you all enjoy though and i do understand why you all enjoy this magic.
my lead line is functional.

 

[HMT2]

Anyone familiar with the Ikomunicate by Digital Yachts? It takes N2K data and translates it to something they call Signal K which is freely accessible. I threw in $200 for a Kickstarter campaign and several months later received one. The idea is that app developers will jump in to make apps that use the signal K protocol. The apps connect through a wireless network on your boat. I have a couple of apps that have been developed and they are okay. I think the apps have not developed as significantly as Digital Yachts hoped. My iKommunicate plugs into my N2K network then into a standard WiFi router. Its all wired into a separate switch called WiFi so there is no power draw if I’m not using it. It’s been fun experimenting and learning. Interesting tech.

 

[dlochner]

Anyone familiar with the Ikomunicate by Digital Yachts? It takes N2K data and translates it to something they call Signal K which is freely accessible. I threw in $200 for a Kickstarter campaign and several months later received one. The idea is that app developers will jump in to make apps that use the signal K protocol. The apps connect through a wireless network on your boat. I have a couple of apps that have been developed and they are okay. I think the apps have not developed as significantly as Digital Yachts hoped. My iKommunicate plugs into my N2K network then into a standard WiFi router. Its all wired into a separate switch called WiFi so there is no power draw if I’m not using it. It’s been fun experimenting and learning. Interesting tech.

Yes, I once had an iKommunicate that I purchased as part of the kickstarter campaign. I think Signal K has a future as an opens source product, much like Open CPN for navigation. Ultimately I decided not to install and sold it on eBay. At the time I was looking at upgrading the MFD and instruments. The boat came with a Garmin MFD, which I really disliked as the interface was poor. The other instruments were B&G Network series. This was a very early NMEA 0183 implementation and the version of NMEA 0183 was not fully compatible with later versions and any bridges to N2K. There were enough things to fix and update on the boat that the last thing I wanted to do was horse around trying to get a simple network to network. As for marine electronic specialists, I live in a marine services desert.

The upgrade went in stages with some wheeling and dealing with Active Captain/Defender specials and ebay. First came a 9" Zues2 MFD with a 4G radar. The radar features were not fully implemented because it needed better heading data with rate of turn, pitch and roll data, which the original compass couldn't provide. In the end, last year I installed new instruments, new AP computer, and a new precision compass. It was an easy install. Once the cables were run, I plugged them all in and turned it on. It's worked well since then. Adding a Vespermarine AIS was also plug and play.

The only significant difficulty I had was self induced. I foolishly thought I could save some money by making my own backbone cable with field installed connectors. If someone would like a good deal on some N2K cable, let me know I have bunch and some field connectors.

 

[Davidasailor26]

I think it is a matter of time before we see a good, quality screen mfd solution from a third party that blows away the functionality of vendor solutions (Open plotter). Nmea 0183 and Ethernet make this possible. Nmea2000, oneNet and certification only hinder this process.

Don't get me wrong, I'm all for open standards, but NMEA0183 isn't really one anyway. It's just simple and old enough that it's been largely reverse engineered. Ethernet and TCP/IP give you the open integration that you're after, but data from both 0183 and 2000 can be put on a TCP/IP network with readily available components (including modern chartplotters themselves), and that will likely remain the same with OneNet.

 

[Stu Jackson]

and it will get worse with AP's that are controlled by submenus on the MFD. So reaching 100% driver navigation usage is more important than a bunch of wiz-bang features. Wish more manufacturers understood that.

That is so very true. My example is the old Garmin b&w that first came out with simple charts, not even real ones. To input a waypoint you had to dig thru four menu levels. Forget about a simple GoTo button either when you're trying to find a WP you, finally, got to enter by the time you were 400 yards away!!! The GPS76Map series fixed that very well, one button shopping. Yippee.
It's one of those things where the programmer thinks he's covered all the bases, and most likely has, but never bothered to try and use it in the real world.
I also believe that because of our sailing environment, touch screens and spoken word don't work so well. It's not only us old fogies who like, need, and want buttons anymore.

 

[Meriachee]

Data security? Seriously?
So, what is the plan going to be when the certificates expire 1/2 way to Bermuda and everything stops.
Don’t think it won’t happen either. We had two very major phone systems go offline this week because somebody on the data side of the house didn’t pay attention to certificate dates.

 

[Gunni]

Forget data encryption, think wireless devices that don’t talk to one another. Dirt tech isn’t ready for marine duty. IPX7 standard for all.

 

[Pizzazz]

Don't get me wrong, I'm all for open standards, but NMEA0183 isn't really one anyway. It's just simple and old enough that it's been largely reverse engineered. Ethernet and TCP/IP give you the open integration that you're after, but data from both 0183 and 2000 can be put on a TCP/IP network with readily available components (including modern chartplotters themselves), and that will likely remain the same with OneNet.

I think you are getting bogged down in the semantics of open vs closed to see the real picture. Protocols get reverse engineered all the time. The software part is easy, especially with modern sniffing tools. The electrical part (layer 1 in OSI) is much harder and often impractical. Nmea 0183 was reversed engineered quickly because it rides on top of the serial protocol and hardware is readily available. Same for the Raymarine/navico/Garmin radar protocols riding on top of Ethernet. You plug in Shark and just observe the commands. Nmea 2000/J1939 does not have an equivalent in the computer world, so you need specialized adapters and few people bothered to reverse engineer it. Even if we know today what most of the PGNs stand for, it is useless since you have to buy expensive J1939 adapters. OneNet has a promise since it rides on Ethernet but they will encrypt it, which will make it much harder to reverse engineer. Hence, it is a non-starter for me. May be some big ships but those have proprietary buses anyway.

Thus, when I say open, I do not necessarily mean standards based, I mean, is it practical to reverse engineer or not. Nmea 0183 is so, same for many of the serial connections used to control ham radio's, etc. Seatalk is propietary but the electrical layer is the same as serial connections, so it is relatively easy to reverse engineer, hence you have many software based Seatalk to serial/nmea adapters. Nmea2000 is electrically different, so you need to buy special hardware -> small market -> not practical. By the way, the same concept applies to sensors, but we got lucky with those, as most come from a single vendor, Airmar, so there is plenty of supply and they are well documented. Plus, you have solid nmea 0183 wind sensors.

Similarly, many engine analog sensors are standards based, so relatively easy to feed via ADCs to a computer. The problem here is that there is no need, i.e. it is better just to connect to a gauge than use mfd's real estate, so we have not bothered. But technically, with a relatively simple tablet or Arduino computer you can have any data (engine/tank/weather/barometer/whatever) show up on your display. We never needed Nmea 2000 for that. Now, if you are running a $1 million yacht or trawler, with digital engine control, you would not mind spending $10K on a Nmea2000 monitoring system. But if you are running a $100K sailboat, with a small diesel engine, and have to upgrade all your sensors or convert them, + spend $50 per waterproof connector (instead of splicing), etc. it just does not make sense to me. Some people do it and then ask the question, why did I do that? So that I can have a window buried deeply in my mft that shows me rpm? Oops!

 

[Meriachee]

As @LeslieTroyer will agree, the NoLand box is not a lot of money and displays engine data from our old tractor motors on the i70s quite nicely via 0183. No sweat. Plug it in.

 

[Gunni]

For about $600 I added a Yanmar engine interface that plugged right up to the Yanmar wiring harness and feeds a whole suite of engine information via NMEA 2000 port to a page on my MFD. I have more accurate oil pressure, engine temp, charge voltage, battery voltage and engine speed than an analog gauge and I don't have to find room for the dedicated gauges. This is actually a good upgrade for owners who are stuck with only engine buzzers and idiot lights.

 

[LeslieTroyer]

As @LeslieTroyer will agree, the NoLand box is not a lot of money and displays engine data from our old tractor motors on the i70s quite nicely via 0183. No sweat. Plug it in.

Actually it transmits it on N2K

 

[dlochner]

I think you are getting bogged down in the semantics of open vs closed to see the real picture. Protocols get reverse engineered all the time. The software part is easy, especially with modern sniffing tools. The electrical part (layer 1 in OSI) is much harder and often impractical. Nmea 0183 was reversed engineered quickly because it rides on top of the serial protocol and hardware is readily available. Same for the Raymarine/navico/Garmin radar protocols riding on top of Ethernet. You plug in Shark and just observe the commands. Nmea 2000/J1939 does not have an equivalent in the computer world, so you need specialized adapters and few people bothered to reverse engineer it. Even if we know today what most of the PGNs stand for, it is useless since you have to buy expensive J1939 adapters. OneNet has a promise since it rides on Ethernet but they will encrypt it, which will make it much harder to reverse engineer. Hence, it is a non-starter for me. May be some big ships but those have proprietary buses anyway.

Thus, when I say open, I do not necessarily mean standards based, I mean, is it practical to reverse engineer or not. Nmea 0183 is so, same for many of the serial connections used to control ham radio's, etc. Seatalk is propietary but the electrical layer is the same as serial connections, so it is relatively easy to reverse engineer, hence you have many software based Seatalk to serial/nmea adapters. Nmea2000 is electrically different, so you need to buy special hardware -> small market -> not practical. By the way, the same concept applies to sensors, but we got lucky with those, as most come from a single vendor, Airmar, so there is plenty of supply and they are well documented. Plus, you have solid nmea 0183 wind sensors.

Similarly, many engine analog sensors are standards based, so relatively easy to feed via ADCs to a computer. The problem here is that there is no need, i.e. it is better just to connect to a gauge than use mfd's real estate, so we have not bothered. But technically, with a relatively simple tablet or Arduino computer you can have any data (engine/tank/weather/barometer/whatever) show up on your display. We never needed Nmea 2000 for that. Now, if you are running a $1 million yacht or trawler, with digital engine control, you would not mind spending $10K on a Nmea2000 monitoring system. But if you are running a $100K sailboat, with a small diesel engine, and have to upgrade all your sensors or convert them, + spend $50 per waterproof connector (instead of splicing), etc. it just does not make sense to me. Some people do it and then ask the question, why did I do that? So that I can have a window buried deeply in my mft that shows me rpm? Oops!

For a small segment of the boating public, being able to easily reverse engineer devices and protocols is a valued feature. However, I believe for the vast majority we want plug and play and don't want to spend time fiddling to get our electronics working. Your discussion is reminiscent of the old Windows/DOS/Mac OS debates. There are some who enjoy and want command line access, but most of us just want to click on an icon.

To each his own.

 

[Jackdaw]

The next-gen NMEA will HAVE to address security, no question. You WANT it to.

As long as you can physically control access to a hardwired LAN, you have a simple but effective network security layer. Add wireless and limiting physical access becomes meaningless. Look at cars, which for years have used ISO 11898-2 CANBUS (NMEA 2000 is CANBUS). Security was not an issue (except perhaps access to the ODB-II port) until they opened Pandora's box by adding wireless. Now bad actors can access the all-trusting network, and talk of hacking cars abounds.

A security layer will support in each device/node an Access Control List, a list of devices/nodes that are vetted and allowed to issue it commands. So your engine (or that or a MAERSK container ship) will only accept commands from network entries that have been approved.

 

[Meriachee]

Actually it transmits it on N2K

M

Actually it transmits it on N2K

Mines a 0.1 version.

 

[LeslieTroyer]

I think encrypted Ethernet will be orders of magnitude harder to reverse engineer than N2K was. First, sniffers for the underlying protocol of N2K were/are readily available from the auto industry, so you could easily isolate hardware and monitor data flows in and out. Even with that, I don’t see many devices from folks that haven’t paid the big bucks to NMEA (can boat is the big exception). With all the piracy laws around reverse engineering who wants to go to jail to do this???

NMEA Ethernet is going to be a very closed standard for years to come.

Les

 

[Davidasailor26]

Data security? Seriously?
So, what is the plan going to be when the certificates expire 1/2 way to Bermuda and everything stops.
Don’t think it won’t happen either. We had two very major phone systems go offline this week because somebody on the data side of the house didn’t pay attention to certificate dates.

Yeah, with the extra complexity could come new problems. Hopefully it's designed well enough to avoid those kinds of problems. Security does aim to provide three things - confidentiality, integrity, and availability. Sometimes the availability part gets forgotten, but a well designed system will have the right architecture to provide the attributes that matter to your use. In an environment like a boat's network I'd argue confidentiality should be a relatively low priority - all you risk is revealing information about where you've been, where you're going, and the conditions where you are. Integrity and availability are the higher priorities - Don't let anyone tamper with your intended course, and keep the system working at all times. So when you start to design the protocol you look at what tradeoffs you need to make, and what checks should implemented to keep what matters to you. Time will tell whether NMEA did that properly, but in principle it is possible.