Something dodgy going on here

[Murky 1]

Below is a photo of our GPS track in the annual race out from Tutukaka around the Sugarloaf yesterday.

 

Or should I say, below is what the GPS believes was our track but is in fact a long way off - it is showing us having rounded some non-existent feature to the north of the Sugarloaf.

 

This is a handheld unit which we have had for three years and which I would previously said - from comparing it with what I could see - was probably accurate to within a metre.

 

Yesterday's trail is probably 250 metres out of whack.

 

Fortunately the race was in broad daylight and we rounded in close proximity with two other boats, so there is no issue in terms of the outcome of our race. However I notice in the Balokovic thread, Fineline is mentioning reviewing GPS trails as a way of resolving disputes over who went round what.

 

I can say in this case, we did go round but the GPS trail would leave us without a leg to stand on.

 

I am always mindful of the "do not rely..." warning that appears every time the unit is switched on but ironically, as we approached the Sugarloaf yesterday I was intending to use it to see if I could shave any closer than the other boats around us.

 

A one-off issue with the charting software perhaps, or is the system coming down faster than we thought?

[Guest]

as we approached the Sugarloaf yesterday I was intending to use it to see if I could shave any closer than the other boats around us.

That's probably what the dudes that killed themselves by slamming into assort solid features, inc Gt Barrier, also thought.

 

A few months ago 3 of us had GPS handhelds on all at the same time and all in the same cockpit. All 3 units were saying they were reading the max number of satellites they could, 1 x 8, 1 x 10 and mine by 12. The 3 reported positions were approx 70mts apart.

 

Aren't we all lucky eyes don't rely on software.

[John B 106]

I noticed discrepancies when we were threading/ gunkholing through the Rimarikis in jan Murky, although it corrected itself later in the day . It was enough for me to watch a Riv going through the inside at speed in a waiting for it kind of way.

 

Barrier is quite a bit out too I noticed again this year. One of the 'lagoons' we frequent had us high and dry on the plotter and man o war passage is about 50 metres or so out.

[Deep Purple 512]

I've noticed with my iphone on google earth and maps that the dot moves all over the place, probably about 50mtre radius from actual position. Not expecting the most acurate GPS from a phone, but a good reminder that GPS is a DERIVED position, not an actual one.

[Marshy 30]

Got stung with that one on Whale rock doing the coastal, Said we were heading right over top of it so we altered course only to then work out the flashing light was on it so the GPS had a shocker for some reason...

[Guest]

Whale rock is some 50-100m inside the bouy/light.

[cam 0]

stuff had a story on this issue today actually.

http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10698665

 

much of the time the accuracey seems very good, it is easy to see how people come to rely on them without appreciating the variations in accuracey that occur. Good reminder.

[Guest]

I seem to remember Sir Peter Blake's Antartic voyage where he was sailing over land!.... (I seem to think its was more E charts than GPS at fault though)

[Murky 1]

stuff had a story on this issue today actually.

http://www.nzherald.co.nz/nz/news/artic ... d=10698665

 

much of the time the accuracey seems very good, it is easy to see how people come to rely on them without appreciating the variations in accuracey that occur. Good reminder.

Interesting Cam. I also heard of a grounding in the BOI by an outfit that would normally not be expected to do do.

 

It's not only the fact that you come to rely on them but also that in situations of doubt (e.g. white-out rain squall, darkness) it would have taken a lot to ignore/override what the GPS is saying to the extent that we would need to.

[Grinna 2]

You do have to make sure that your GPS is working from the right datum .... or at least the same datum as the mapping software. There are multiple datums (data) that a GPS unit can work from ..... there's a dazzling list of them lurking in the setup system of each and every GPS unit. It would be a big assumption that all GPS units are set to the same datum or that each GPS is automatically set to the best datum for your use.

[Guest Rocket]

I foolishly assumed we had all learnt from recent events in Oz. GPS is a guide in an electronic virtual world - your eyes are in the real world. It is the real world that bites - at all times confirm what your GPS is telling you with your eyes and if there is any doubt believe your eyes!!!

 

GPS and radar in poor visibility is quite good - but only if you know how to use your radar. Remember the olden days when crusty navigators stuck there head out of the hatch to look at each turning mark or obstruction - that is still the way to do it.

 

And I can personally confirm that rocks marked by bouys on long cables need to be treated cautiously. The rock stays still and the buoy moves around - a lot!!!

[Bad Kitty 252]

You do have to make sure that your GPS is working from the right datum .... or at least the same datum as the mapping software. There are multiple datums (data) that a GPS unit can work from ..... there's a dazzling list of them lurking in the setup system of each and every GPS unit. It would be a big assumption that all GPS units are set to the same datum or that each GPS is automatically set to the best datum for your use.

 

There are other options in the setup menu that affect accuracy & offset also. DGPS for instance which is not active here but if not turned off can affect the position shown.

But most importantly it is only an "aid to navigation"

Talk to any marine loss adjuster about how many claims relate to solely relying on GPS & plotter for navigation.

[cam 0]

I foolishly assumed we had all learnt from recent events in Oz.

What events were they Rocket? Don't tell me the cyclone and floods were caused by the gps satellites having a wobbly? I thought they were just too much of a coincidence.

[Rangi1 51]

I suspect he means the loss of PWC (ex Shockwave) onto Flinders Islet with the loss of two experienced sailors.

[cam 0]

Oh. The Oct 2009 race. actually I had not realised the significance GPS errors (and the dependance on it's use) played in this incident. Thanks for bringing this incident up as I must have been busy at that time and defintely worth knowing more about this one.

 

Most people here have probably already been all over this but the investigation here has a good summary of the incident and GPS errors (from page 26 down):

http://www.cyca.com.au/sysfile/downloads/CYCA_Flinders_Islet_Internal_Inquiry_Report.pdf

 

if you are short of time the quotes below help understand some of them.

 

"While these systems are capable of being accurate to within a few metres they are also subject to a number of variable errors which can degrade the accuracy significantly. These errors can be associated with the:

 satellite constellation and the visibility of satellites,

 system time errors,

 motion errors associated with a poorly placed aerial on a yacht,

 accuracy of the original chart survey,

 transposing government chart data to commercial products, and

 use of different datums."

 

"110. Closer examination revealed very high Dilution of Precision13 (DOP) (see Diagrams 4 and 5) readings shortly before 2.00am (1500 UTC) until after 3.30am (1630 UTC) that could corresponded with an error possibly in excess of 100 metres. At that time there were 6 or 7 satellites visible (Diagram 1) at Flinders Islet. Of these only 3 or 4 were useable (Diagram 2). Greg Halls advises that on a yacht, with a low aerial and subject to heel, that data from a satellite whose elevation is less than 30° should not be trusted because of rapid kinematic characteristics."

 

The report has alot else to say (from a hydropgraphic surveyor) on how hard it is to predict the magnitude of the errors involved.

[cam 0]

It's not only the fact that you come to rely on them but also that in situations of doubt (e.g. white-out rain squall, darkness) it would have taken a lot to ignore/override what the GPS is saying to the extent that we would need to.

 

Yes that is a big problem, every time I have used my GPS, the accuracey is kind of reinforced as it shows the headland or the landmark just a few meters over there and sure enough it is.

 

I recall learning about many of these GPS limitations at a boatmasters course several years ago, but as I have not had cause to use the knowledge, or even notice inaccuracies in my GPS, much of the info about the inherent inaccuracies in the system I had almost completely forgotten.

 

This is a blimin good reminder.

[Guest Brucey]

Whale rock is some 50-100m inside the bouy/light.

 

 

You could check that with George Hendy for an acurate position???

 

Cant the US put an INACURACY in to the whole world GPS system when they want to, was certainly publicised when they launched GPS? Maybe the shite in the Middle East has meant they turned the scramble back on?

[Guest]

Thread drift - when I read that report it sounded more like a human error than a technical error to me.

[Rangi1 51]

I think relying on GPS to get you as close as possible to a rock which is a rounding mark probably classifies as a human error. There were a lot of factors in that tragedy, to me it seemed like an overreliance on the GPS to locate the rock was one of them.

[Deep Purple 512]

The most damning aspect of that report is that they were fully aware the GPS position was inaccurate yet used it for close navigation anyway. Less "human error" more "gross negligence"

40. In interviews with the Inquiry some PwC crew commented that they had previously observed errors on the chart plotter. An example was given that on entering Sydney Harbour the plotter indicated the boat had crossed South Reef and the Sow and Pigs Reef when the yacht had clearly been in safe water. A similar example was provided during the Hamilton Island race week. This could have been a problem with the update interval in creating the past track or possibly a setup, installation or calibration problem on PwC.