I have been involved in selling and maintaining a high-end alarm system for boats, and your post brings up some talking points that you might consider.

Using the internet to warn of fires, sinkings, and the like is full of problems. The boat has to be near a wifi hotspot to send an alarm. And, once the alarm is on the internet, it has no place to go but to an inbox, where it might be read the next day.

It makes a lot more sense to use the SMS service provided for cellphones. The coverage area is broad, and the infrastructure is in place and access can be purchased for $10/month. The alarm goes directly to the owner's cellphone, so is likely to be read immediately. The captain's or dockmaster's phone can be alerted at the same time. And, most GSM cell companies provide an SMS to email gateway so the message can go to that inbox as well.

A lot of these points are covered in Frequently Asked Questions at http://www.geocities.com...aic_FAQ/Mosaic_FAQ.html . I also have some schematics up that show how to wire various sensors (it shows how I wire them, there are lots of ways.....) but the LiessemUSA website is down (it was costing money to maintain and was not an effective sales tool). If anyone emails me with an address I will round up some brochures. Or, you can get to the description in Norway by going to http://www.liessem.no/indexen.php, selecting "English", and then selecting "MOSAIC" on the bottom right of the window.

I know this sounds like an illegal commercial use of the forum. It is not, Marinette owners are not the target market. Just thought it might be helpful to see other ways of doing it....

I did some work with mobile phones for monitoring remote water stations. We found with SMS it could take anything up to 24 hours for a message to be delivered, though most came through pretty quick.

The old fashion way was using a paging service, you sent a data message to the centralised paging number and that would be routed on, in real time by the service, to your pager. The data would appear on the pagers screen.

The paging type service would be possible with a mobile phone, most mobile phones not only have a voice number they also have an associated data number. In the case of the remote water stations we would phone in on the data number and a status report would be sent back to the central control room where it was logged and displayed on a screen.

I am not sure if any standard mobile phones are equipped to display data messages on the screen but it would not be difficult using an industrial mobile phone module. No doubt with future programmable mobile phones on the way this will be easily done.

=> SMS messages are indeed in batches, like email and most other computer stuff. They are sent to the cell phone using the "gaps" in bandwidth, since the voice data is real-time. Your provider's server holds the message if you are out of touch, then tries to deliver it until your phone acknowledges receipt.

Usually I can query the boat in FL from here (DC) and get an answer back in ten or fifteen seconds. Good for an alarm, bad for realtime control.

=> Mobile-based data services are not actually real-time, the requests are cued until the bandwidth is available. Your mileage may vary. The SMS is a data messaging/paging service with guaranteed delivery, but provided by the cell company (at no or little extra cost).

=> 24 hours for delivery means that two GSM providers were not talking to each other. There are legal and financial problems in getting competing networks to work together, batch or no. This would be VERY RARE if sender and sendee were both on the same network.

=> I did encounter a problem getting SMS from Google. They go direct to the provider's server, and my prepaid service just wasn't hooked in. No problem with an SMS sent from a phone, however.

For the hobbyist it comes down to cost and challenge. Writing a custom GUI like the Maretron NK2 software is the challenge while NK2 still needs the USB100 gateway, which in the UK is $560.

The development board Marinettejoe is using is perfectly capable of doing what the USB100 does, in fact that what this forum is about.

Reading the data sheets the problem with the Maretron NK2/USB100 is that it is only focused on Maretron products, so how well it would cope with NMEA2000 products from other manufacturers is not clear.

Oddly the USB100 is really about converting NMEA2000 to the old NMEA0183, so that old charting programs can still be used, rather than creating an "OPEN" USB interface.

A small business, particulary in the early stages, has to adopt the hobbyists approach in evaluating if there are any practical business opportunities in NMEA2000.

Contains opinion....

ON a marina area specification. Doug, all. I am not sure I'd agree.

You are correct in that there are limitations with ethernet, but it's ubiquity makes a good choice for a low cost alarm. An remote alarm system that costs $1500 would go into few boats, whereas an monitoring system that with a costs< 300 would go into many boats. Generally speaking USCG regs follow a failure and effects approach to equipment, which tends to look at failures and not behavior. It's old thinking of a system of systems. Expensive separate systems. Each single system is evaluated, so multifunction systems (Fire, CO, Bilge) don't get built for the US because a device would come under different labs and testing. In order to cover my self for each of these, I buy a separate system. A low cost "good enough at the dock" system is a reasonable capability, that the community is better off having than not. Since bandwidth is limited in wireless ethernet, we can have the system activate and connect when an event occurs, rather than have a constant connection. Not perfect, but enough to save many boats and maybe a number of marinas from fire and sinkings.

Some Cellular phones can direct dial a number and create a data link (or play a message). This is common is some alarm systems. Most boats sink at the dock (BOAT US accident) and http://www.boatinglife.c...e_content.jsp?ID=44439. But as Cellular companies are unwilling to allow many to use the dialer interface directly, we are limited to "specialty equipment and recurring costs".

While USCG FMECA MIL-STD-1629A (failure mode, effects, and criticality analysis) approach is good, it tends to favor single systems at higher costs, rather than a ensuring system at some cost. I believe this is why CG* is being discontinued, as the risk of the connection failing is is higher than not having a communication system. For some small boaters, a cell phone is the most likely communication path, a VHF is the CG preferred, while the best solution should be multipath (connect to any system). Oddly the USCG does allow different types of Life vests (Type I-IV) based on some behavioral recognition that some types won't be worn by the average boater. FMECA has it's place, but so does behavioral threat analysis.

A behavioral threat analysis approach assesses human activity in response to cost and actions. This factors the behaviors on probability of installation, use and cost is assessed with the threat. This needs to apply to electronics regs on Boats.

I also feel that mandated unified safety capability (functionally specified not by system) needs to added to most boats. A GPS enabled/VHF Radio/Cell/Weather/AIS unit, and a bilge vapor/fire/CO/flood (for closed bilges) warning unit. If any new boat had functional requirements for safety, then the lowest cost and easiest to install unit would win. This changes system of systems thinking to system of capability thinking.

Here is another example, If I have a vapor sensor that connects to my NMEA 2000 network, and it uses the network to display a warning, can I be in compliance with the USCG regs. No, the function is there, but it does not fit the reg as a testable device.

So we have a separate vapor alarm, a separate CO alarm, a separate xxxx and yyyy and zzzz, which too few enough of us install. That may not be true I am a huge manufacturer of boats and make my own electronics and engines, as then I can afford to fight the regs.

Thanks