An example of a GPS 'failure'

Re: An example of a GPS \'failure\'

You know what I mean and I suspect you are being deliberately argumentative as I thought I was agreeing with you. People use GPS in pilotage situations when there is a more appropriate visual reference point and then when their reciever encounters noise or the software is bad they blame the infrastructure. If the software in the gps reciever used only 100 bytes of 8 bit storage in its data segment then that represents a state machine of the order of 10^240 states before considering any OS the reciever may have, any hardware issues or any issues concerning the integrity of the satellites. It clearly can't have been exhaustively tested as the testing would probably not finish before the predicted heat death of the universe so at best, like any system involving software, the best that can be said is that it appears to work but spurious things can and will happen.
 
Re: An example of a GPS \'failure\'

Look at my explanation above - sitting at your berth, your GPS is applying a 0/0 DR so it will maintain a reasonably accurate position even when fix quality is low. Underway, it is not likely as easy to spot deviations.

BUT my GPS sounds an alarm and displays a message if it goes into DR mode. It never has, either at sea or on the berth.

(The ECS does the same, if DR mode fallback is enabled on it).

Sorry /forums/images/graemlins/smile.gif.

John
 
Re: An example of a GPS \'failure\'

- or an rf interference, other problem (voltage dip) on an inadequately built boat.

(Gawd, I can see I am inviting disaster here, this summer our GPS will be alarming every 10 mins and we will also be bouncing off the rocks due to total system failure /forums/images/graemlins/smile.gif).

John
 
Re: An example of a GPS \'failure\'

I think Brendan is referring to propagation matters to do with ionosphere (eg delay), not noise.

John
 
Re: An example of a GPS \'failure\'

I thought those things stopped when SA was turned off. Were they not a result of timing shifts in the satellites to decrease the accuracy of the system.
 
Re: An example of a GPS \'failure\'

We drove to Portugal last Feb using GPS and autoroute on laptop, the syatem was so accurate that when pulling in to service stations on the motorway, the message "off route" came up so the accuracy is not a question. Whist passing Bordeaux we were momentarily placed about 100 miles into Biscay, the track spike was recorded and within 20 seconds we were back on track, this happened three or four times over 5 minutes and all returned to normal, the spooky thing was it happened once on the return journey, not at the same place but very close.
I wont burn all my paper charts just yet
 
Re: An example of a GPS \'failure\'

That is still noise.

Er, no it isn't. There may be code "noise" and receiver noise (which are not to do with the ionosphere or atmosphere in any way) associated with the signal, and charged particles and collisions of such particles (eg as in aurora) may create noise at rf as may lightning, but the ionospheric delay errors which affect the GPS are definitely not noise.

The lower atmosphere has a lesser delay error associated with it too (due to moisture, etc) and that is not noise either.

I suggest that you may not wish to argue on this point unless you have a degree in radio physics, electromagnetism or similar /forums/images/graemlins/smile.gif.

John
 
Re: An example of a GPS \'failure\'

To reinforce someones argument, probably yours John, we have 3 plotters on board and 3 GPSs. One plotter (Navman 5500i) is totally independent it has an integral GPS with WAAS/EGNOS (but not used currently until EGNOS is reliably functional) and is on a separate battery bank. The second plotter is a Navman 5600 with a separate GPS (also has WAAS/EGNOS but turned off). This Plotter/GPS sends NMEA data to the ST50 Seatalk system via the ST6001 Autopilot. Plotter #3 has a separate Raymarine WAAS/EGNOS (turned off again) GPS and can also be connected to the Seatalk system if needed via a changeover switch, this plotter is normally reserved as a spare but is turned on occasionally to test it and keep it's almanac memory intact.

We occasionally get a lost data alarm from the Seatalk system which clears within seconds, the position data is lost (in the Seatalk system) as is evident on the cockpit Navdata repeater or the 6001 Autopilot head. HOWEVER at these times the cockpit Navman plotter (independent of any instruments) has NOT lost position and nor has the chart table one which is the one that actually feeds the instruments - to me this says the alarm heard is for momentary DATA loss around the Seatalk circuit, not a GPS failure. All our GPS/plotters have the averaging function set very low if not actually 'off' these days now S/A is off (I must check next visit), each one leaves a continuous track visible on screen and never have I seen any deviations from a steady track. Additionally we have 'routes' programmed in which we have tried and tested for regular passages including some difficult and complex ones where it is essential to stay on track despite strong cross tides (like in Chenal Du Four). On these routes not only are we watching the plotter but also the Off Track Error, this doesn't show sudden 100m jumps either but is stable.

Now we are not using handheld GPSs but fixed sets which are connected to (separate) big battery banks and not the engine bank, so voltage drops, variations and spikes are far less likely.

We did one time have a 'gone to DR mode' alarm on Seatalk which was when a Raymarine GPS failed totally. This wasn't a problem as the cockpit plotter/GPS was running and we switched to another set immediately for the ST50s/pilot which our set up allows, Raymarine later replaced the faulty GPS head FOC. This at least illustrates that we have planned backups in place - 'disaster management' of a kind?

Do we trust plotters for critical passages in bad visibility or the dark? Well like most folks I prefer to be able to 'see' but sooner or later circumstances will arise where this isn't feasible and yes we have used a combination of plotter and radar to very good effect. We have entered L'Aberwrac'h in 25m visibility with a 4 kt tide running across the leading lines through the reefs for example, gone into Isles De Glenans at night several times as well as through Chenal Du Four in fog. I should stress we are using well tested waypoints and routes that have been previously followed in good visibilty, even using the pilot's 'follow the track' feature.

Going back to Benbow's original post (for which I await a cuddly toy for the correct answer!), they did NOT have a GPS system fault either from the satellites or the set itself, there was no high voltage zapping nor did the USA turn off the system to bomb Anglesey, the error was a simple human one because someone had re-set the averaging way too high.

Oh and before someone smart chips in we DO have charts on board and both SWMBO and I are very competent pencil and paper navigators.
 
Re: An example of a GPS \'failure\'

[ QUOTE ]
Going back to Benbow's original post (for which I await a cuddly toy for the correct answer!),

[/ QUOTE ]

Sorry, here it is:

Sailor.jpg




[ QUOTE ]
the error was a simple human one because someone had re-set the averaging way too high.

[/ QUOTE ]

Exactly, and the result was quite subtle. Anyone who said they would rely on GPS until it went haywire would have come unstuck in this scenario. OK, it was an unusual set of circumstances, but strange things do happen.
 
Re: An example of a GPS \'failure\'

[ QUOTE ]
Exactly, and the result was quite subtle. Anyone who said they would rely on GPS until it went haywire would have come unstuck in this scenario. OK, it was an unusual set of circumstances, but strange things do happen

[/ QUOTE ]

Thanks for the bear, I have 3 on board so he will have company!

Your situation was a little unusual too in that I would guess you were not on your own boat so it was possible for Mr Fiddler in the crew or a previous crew to play with the settings unnoticed! Many years back when I was roped in to skipper charter boats for a works sailing club I had such a person who used to 're-adjust' anything and everything. I missed shipping forecasts because he had switched the radio to another station, had a severe fright when he had switched the echosounder from feet to metres, and was blinded one night when he helpfully turned on the deck floods just as we were entering the narrow channel to a marina.
 
Re: An example of a GPS \'failure\'

Anyone who said they would rely on GPS until it went haywire would have come unstuck in this scenario.

First I agree, your post was interesting.

Not sure I have heard anyone say they rely entirely on GPS. And in the same way would be just as risky to just rely on plotting on a paper chart as that is likely to end up in error sometime or another - bearing off wrong headland or misdrawn on chart, compass misread, etc, etc.

In my own case, I have made plotting errors on a paper chart but have never had my GPS/ECS get it wrong so propose that if one was to rely entirely on one method, perhaps GPS/ECS is the most reliable - at least for me it has been but I would be interested in anyone who has regularly used both who could claim otherwise.

It is probably clear that I am quite happy placing alot of reliance on GPS/ECS and also quite happy navigating with it through dangers in low visibility on proven tracks as the risk of failure in that short time is very low and if I had any concerns on a particular occasion I would not do it.

I think in the same way if nav with paper and pencil I would do the same through dangers, if it were feasible (a dark night is a common example), but would always be aware that I could make a nav mistake of some kind. Again in my own case I figure that the likelihood of such a mistake (especially under pressure) is much more likely with paper, pencil, compass, log, and timepiece than a problem arising with the GPS/ECS in the passage time of concern. Yet the paper and pencil only guys (bless them /forums/images/graemlins/smile.gif) tend to regard their method as being more reliable than GPS/ECS.

Ideally we all use several methods to preserve security of nav while in pilotage but I would never dream of maintaining both a paper and an ECS plot and in my opinion the ECS plot is the one most likely to be the most error free. I do maintain a high level of spacial awareness when coasting in pilotage, I counting the major landmarks/features off and brief looks over the steering compass to see they are roughly where expected - and entering harbours, etc I always correlate what I see with the plot.

If it makes sense I am happy, and it always has except on a few occasions where I thought I was somewhere else but GPS/ECS in the end convinced me it was right. So I generally regard it as being more likely to be correct than I am.

John
 
Re: An example of a GPS \'failure\'

Robin is correct, it was not my own boat (sadly!). Normally skipper and mate on this boat are both equally and totally on top of the nav. But there are times, such as this, in fast changing pilotage where skipper just has to trust the mate to carry-out a previously agreed pilotage plan while he covers the deck (or vice versa).

I am sure you are right that human error is now much more likely than an electronic error (as distinct from a failure). Hence the unease when I announced that the GPS was wrong and I was ignoring it! If the GPS had just died or given a position off Bermuda, the skipper would not have batted an eyelid. But the fact that it gave a very reasonable position that conflicted with where I insisted we actually were, certainly raised anxiety levels. If we had not been, among other things, ticking off buoys with time and log, we could easily have convinced ourselves that we were wrong.
 
Re: An example of a GPS \'failure\'

Aghast /forums/images/graemlins/smile.gif - so you now not claiming it as a propagation thing, but just as a loose term for (random?) error.

Given that you have a degree in physics you may wish to wonder why, for the same reasons as with a radio wave eg the GPS signal, we do not refer to the "slowing" (note the "") of light through a medium (delay) as being "noise" or due to "noise" - consider both the reasons from both the quanta (photon) point of view ie delay due to collision with atoms and readmission of another photon by the atom to shed the energy gained and from the electromagnetic point of view ie modelling the medium as a transmission line along which there is delay.

You will see why delay in electromagnetic propagation through the ionosphere (and atmosphere) is not "noise" nor due to noise.

John
 
You must log in or register to reply here.

Other threads that may be of interest

F
Emtrak AIS transponder configured and working, not showing my position in Navionics unless
Replies
5
Views
291
FairweatherDave
F
Y
Vespermarine XB-6000 AIS transponder
2
Replies
24
Views
450
PaulRainbow
PaulRainbow

Share this page

Top