Re: [OSM-talk] Defective GPS trace

[Roy Wallace]

On Sun, Jan 3, 2010 at 9:17 AM, Jean-Marc Liotier  wrote:
> Steve Bennett wrote:
>
>> I'm even tempted to draw a massive straight line between several
>> towns to indicate roads that I know exist but that I haven't
>> surveyed. Would this offend a lot of people here?
>
> That would be using a map as an item in a to-do list.

IMHO that's ok provided you comprehensively use source:*=* and FIXME=*

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Steve Bennett]

On Sun, Jan 3, 2010 at 10:17 AM, Jean-Marc Liotier  wrote:

> That would be using a map as an item in a to-do list. It would look ugly
> to me and I doubt that many people would support that use. Better keep
> the todo list separate - in another layer if you want it represented
> geographically.
>

Yes and no. A straight line works almost as well for routing as a correctly
surveyed road, and much better than a non-existent way. So rather than a "to
do list", it's more like a first draft.

You might say that in order of priority, we would like the following
information about every road:
1) Start and end point
2) Major junctions
3) Name
4) Minor junctions
5) Exact route.
6) Surface
7) Width, lanes, speed limit...

The order is debatable though.

Steve
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[John Smith]

2010/1/3 Steve Bennett :
> completely missing way. I'm even tempted to draw a massive straight line
> between several towns to indicate roads that I know exist but that I haven't
> surveyed. Would this offend a lot of people here?

I've done this, I knew a road ran between 2 towns but wasn't on OSM,
so drew a straight line, and when I was able, I surveyed it properly.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Jean-Marc Liotier]

Steve Bennett wrote:
> On Sat, Jan 2, 2010 at 3:40 AM, Craig Wallace wrote:
> 
> But I'd still agree with Shaun - a single GPS trace is not really
> accurate enough for adding ways to OSM IMO.
> 
> Hmmm...is there consensus on this view? My approach so far has been "any
> information that is approximately correct is better than nothing". The
> quality of information can be improved over time, and a way whose
> endpoints are correct but with a fictitious route is far more valuable
> than a completely missing way.

Anyway you put it, the map will only ever be an approximative
representation of reality, so approximation is the name of the game and
 improvement is part of it. But from approximative to fictious, there is
a line that I would not cross.

> I'm even tempted to draw a massive straight line between several
> towns to indicate roads that I know exist but that I haven't
> surveyed. Would this offend a lot of people here?

That would be using a map as an item in a to-do list. It would look ugly
to me and I doubt that many people would support that use. Better keep
the todo list separate - in another layer if you want it represented
geographically.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Steve Bennett]

On Sat, Jan 2, 2010 at 3:40 AM, Craig Wallace  wrote:

> But I'd still agree with Shaun - a single GPS trace is not really
> accurate enough for adding ways to OSM IMO.
>

Hmmm...is there consensus on this view? My approach so far has been "any
information that is approximately correct is better than nothing". The
quality of information can be improved over time, and a way whose endpoints
are correct but with a fictitious route is far more valuable than a
completely missing way. I'm even tempted to draw a massive straight line
between several towns to indicate roads that I know exist but that I haven't
surveyed. Would this offend a lot of people here?

Steve
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Ulf Möller]

Martin Koppenhoefer schrieb:

> while this might help in the case your devices calculations/capabilities 
> create the offset, this is still no help in the case of atmospheric 
> interferences.

True, but it does help in practice.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Martin Koppenhoefer]

2010/1/1 Ulf Möller 

> > You can check the satellite screen on the Garmin. It should show an
> > estimated position accuracy.
>
>
> The eTrex often claims 10m accuracy when in fact it is 50m off, so that
> doesn't really help. Using two different GPS receivers is a good idea if
> you don't want to survey twice.
>


while this might help in the case your devices calculations/capabilities
create the offset, this is still no help in the case of atmospheric
interferences. Generally I'd say the more traces you can get, the better is.
If you got only 1 trace that shouldn't prevent you from mapping though:
enter the data the best you can (detailed tags are at least as valueable as
positional accuracy), and probably someone else will optimize the track with
new data in the future.

Cheers,
Martin
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Dave F.]

Craig Wallace wrote:
> But I'd still agree with Shaun - a single GPS trace is not really 
> accurate enough for adding ways to OSM IMO.
Whilst I agree - more the merrier, just because you have only one trace 
it's not a valid reason not to upload it.

If you feel it's an accurate representation,  good. If not, add a note 
or Fixme tagged explaining that it needs retracing to get a more 
accurate average.

IMO, if your policy was followed, OSM would be a fraction of it's size & 
poorer for it.

Cheers
Dave F.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Dave F.]

Steve Bennett wrote:
> I've got a trace from today which is significantly out of sync with a 
> path I traced from Nearmap:
>
> http://www.openstreetmap.org/edit?lat=-37.880138&lon=145.193417&zoom=19&gpx=594988
>  
> 
>
> The trace looks like I was wandering through the grassy paddock, but I 
> was actually following exactly that northern most "highway=path" in 
> the bush. So it looks like the trace is incorrectly recorded something 
> like 50m north of where I actually was. Now, since the discrepancy 
> seems to go away on that track a bit further east (later 
> chronologically), presumably the explanation is the GPS data is 
> faulty. Is this common? I'm new to GPSing, so I'm just surprised. It's 
> a Garmin Oregon 550. Is there anything I can do to reduce, or at least 
> detect, such errors?
>
> Steve
There are some good points in the previous messages, but I think there's 
an overall strategy & that's:

Don't be a slave to your GPS.

Be aware of your surroundings - Are you in thick undergrowth? Are you 
traveling at the bottom of a steep cliff?

As has been said, keep an eye on the accuracy reading (except if you're 
driving of course!)
If you are in an area where reception is poor make a note of it, either 
with a waypoint or, as I do, with old fashion paper & pencil.

I take photographs to help me remember what my route looked like. I find 
it extremely useful for recalling road signs & street names.

If your trace goes straight across a field, but you know you walked 
around the edge of it, mark it as you walked it, taking a best guess as 
to where you went. Then tag it with a note or Fixme explaining that it 
needs updating.

Cheers
Dave F.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Dave F.]

Liz wrote:
> On Sat, 2 Jan 2010, Michael Hufer wrote:
>   
>> On the Oregon 550(t) you will find the satellite almanac-screen if you
>>  touch  the "five-bars" satellite reception indicator.
>>
>> 
> Thanks, will try it.
> Later I might read the instructions.
>
>   
I wouldn't bother, they're not much cop :-(


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Liz]

On Sat, 2 Jan 2010, Michael Hufer wrote:
> On the Oregon 550(t) you will find the satellite almanac-screen if you
>  touch  the "five-bars" satellite reception indicator.
> 
Thanks, will try it.
Later I might read the instructions.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Steve Bennett]

Thanks for the comments, everyone. In all my playing with nearmap, I have
little reason to doubt their accuracy. There are a couple of little seams
here and there, but nothing more than a couple of metres. Giving the way
this trace here meanders all over the place, I'm pretty confident that the
nearmap data is a safer bet than the trace - but I'll be paying more
attention to accuracy in the future.

Btw, can everyone actually see the trace in the URL I referred to (
http://www.openstreetmap.org/edit?lat=-37.880138&lon=145.193417&zoom=19&gpx=594988-
then press G)? I'm a bit uncertain about the different privacy
settings in
Potlatch.

Steve
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[andrzej zaborowski]

2010/1/2 Anthony :
> This seems like somewhere that the wisdom of crowds actually applies.  I
> think I'd trust the average of a bunch of independent GPS traces to a single
> orthorectified aerial - especially in an area which isn't extremely flat.
> But I guess I might be convinced otherwise, if I'm actually shown the
> "sub-metre accuracy" claim, which presumably outlines the methods utilized
> to ensure such accuracy.

Average sub-metre error would still allow for a place or two with 50m
error especially over a big area such as Australia, and Steve might
have visited one of these places.  It would be best to just go there
again and see if on the second survey the error is siginificantly
lower (or there's still error but the the vector is opposite
direction).

For example based on GPS traces I can say with some confidence that
Yahoo is on average < 3m off in my city even though there is a couple
of places that are badly mangled.

Cheers

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[John Smith]

2010/1/2 Aun Johnsen :
> Than they have done a hell of a job on aligning the fotos, cudos to them. I
> know that Yahoo imagery varies from less then 1 meter to at least 30 meters
> on the hi-res, havn't seen nearmap, and as I understand its only for
> Australia, so I would not have any data to compare with.

Their imagery is a lot better than yahoo, in places yahoo is noted to
be out by 20+m, where as nearmap is pretty consistent. They use
planes, not sats dunno if it makes things easier or not.

As for comparing, you can always use the GPS data uploaded to OSM.

> As far as I know IALA have coverage also in Australia, and I am sure that
> you can get HF also, though it might inicate stations far away (IALA have a
> range of about 150km, HF roughly 1000km), these two systems are focused on
> shipping, so I would guess augmentation stations and transmitters are
> located along the coast. So saying that Australia have no augmentation
> systems are plain wrong. Now the question is rather if your handheld device
> supports IALA.

I didn't think there way, and the only augmentation I've seen is from
mobile phone networks...

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[John Smith]

2010/1/2 Anthony :
> Where are you getting that "sub-metre accuracy" claim from?  This thread
> (http://www.mail-archive.com/talk...@openstreetmap.org/msg03414.html), which
> you contributed to, throws out "3-5 meters", "1-4 meters", and "5 meters or
> so".

They claim the imagery should be sub-metre, from what I've noticed the
differences between different images on different dates are very close
to each other. I can't however verify their claim, I don't have
anything accurate enough, my claim was a best estimate based on GPS
traces...

> This seems like somewhere that the wisdom of crowds actually applies.  I
> think I'd trust the average of a bunch of independent GPS traces to a single
> orthorectified aerial - especially in an area which isn't extremely flat.

They're planning to re-fly once a month over capital cities and you
can access their older images.

> But I guess I might be convinced otherwise, if I'm actually shown the
> "sub-metre accuracy" claim, which presumably outlines the methods utilized
> to ensure such accuracy.

I've cc'd Ben from nearmap on this email, I haven't seen what they do
mentioned before.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Anthony]

On Fri, Jan 1, 2010 at 6:49 PM, John Smith wrote:

> 2010/1/2 Aun Johnsen :
> > Even if you have access to good arial photography, remember that it might
> be
> > out of alignment, it can be a good advise to gather some good fixes to
> check
> > the alignment of your photos, this can be several GPS tracks along your
> > trail.
>
> I'm pretty sure the imagery he's refering to is nearmap.com, which I'm
> not sure how they manage it exactly but they seem to be about
> sub-metre accuracy...
>

Where are you getting that "sub-metre accuracy" claim from?  This thread (
http://www.mail-archive.com/talk...@openstreetmap.org/msg03414.html), which
you contributed to, throws out "3-5 meters", "1-4 meters", and "5 meters or
so".

This seems like somewhere that the wisdom of crowds actually applies.  I
think I'd trust the average of a bunch of independent GPS traces to a single
orthorectified aerial - especially in an area which isn't extremely flat.
But I guess I might be convinced otherwise, if I'm actually shown the
"sub-metre accuracy" claim, which presumably outlines the methods utilized
to ensure such accuracy.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Aun Johnsen]

On Fri, Jan 1, 2010 at 11:49 PM, John Smith wrote:

> 2010/1/2 Aun Johnsen :
> > Even if you have access to good arial photography, remember that it might
> be
> > out of alignment, it can be a good advise to gather some good fixes to
> check
> > the alignment of your photos, this can be several GPS tracks along your
> > trail.
>
> I'm pretty sure the imagery he's refering to is nearmap.com, which I'm
> not sure how they manage it exactly but they seem to be about
> sub-metre accuracy...
>

Than they have done a hell of a job on aligning the fotos, cudos to them. I
know that Yahoo imagery varies from less then 1 meter to at least 30 meters
on the hi-res, havn't seen nearmap, and as I understand its only for
Australia, so I would not have any data to compare with.

As far as I know IALA have coverage also in Australia, and I am sure that
you can get HF also, though it might inicate stations far away (IALA have a
range of about 150km, HF roughly 1000km), these two systems are focused on
shipping, so I would guess augmentation stations and transmitters are
located along the coast. So saying that Australia have no augmentation
systems are plain wrong. Now the question is rather if your handheld device
supports IALA.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[edodd]

> 2010/1/2 Steve Bennett :
>> When I looked up WAAS on wikipedia a while ago, it appeared that we do
>> have
>> an equivalent system in Australia (although the term WAAS is american),
>> but
>> I'm not sure how to tell whether it's functioning in a given area. I
>> switched the WAAS capability on the GPS on, but again, I don't know if
>> it's
>> actually doing anything.
>
> There is no GPS augmentation system in Australia, the closest one is in
> Japan.
>
we will have to disagree on this one.
skim to the fifth page of this pdf to find the first listed AU stations

http://www.beaconworld.org.uk/files/worldDGPSfreqorder.pdf

These of course are the free (unencrypted ones).
Encrypted broadcasts are used by surveying firms, and won't be listed.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[John Smith]

2010/1/2 Aun Johnsen :
> Even if you have access to good arial photography, remember that it might be
> out of alignment, it can be a good advise to gather some good fixes to check
> the alignment of your photos, this can be several GPS tracks along your
> trail.

I'm pretty sure the imagery he's refering to is nearmap.com, which I'm
not sure how they manage it exactly but they seem to be about
sub-metre accuracy...

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[John Smith]

2010/1/2 Steve Bennett :
> When I looked up WAAS on wikipedia a while ago, it appeared that we do have
> an equivalent system in Australia (although the term WAAS is american), but
> I'm not sure how to tell whether it's functioning in a given area. I
> switched the WAAS capability on the GPS on, but again, I don't know if it's
> actually doing anything.

There is no GPS augmentation system in Australia, the closest one is in Japan.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Steve Bennett]

On Sat, Jan 2, 2010 at 8:17 AM, Aun Johnsen  wrote:

> The accuracy shown on your GPS unit is not necessary the actual accuracy,
> but just a calculated accuracy depending on the signals your unit is
> receiving. You can experience athmospheric disturbance, plasma-effects,
> signals reflected off tall buildings, canyon or urban canyon effects, bed
> satellite constillations etc. If you have SBAS (WAAS/EGNOS) activated you
> might see an improvement in the signal, but mind that if you are outside the
> official coverage of such systems you might experience that these
> corrections are in fact increasing the error. If you have access to
>

When I looked up WAAS on wikipedia a while ago, it appeared that we do have
an equivalent system in Australia (although the term WAAS is american), but
I'm not sure how to tell whether it's functioning in a given area. I
switched the WAAS capability on the GPS on, but again, I don't know if it's
actually doing anything.

Steve
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Aun Johnsen]

The accuracy shown on your GPS unit is not necessary the actual accuracy,
but just a calculated accuracy depending on the signals your unit is
receiving. You can experience athmospheric disturbance, plasma-effects,
signals reflected off tall buildings, canyon or urban canyon effects, bed
satellite constillations etc. If you have SBAS (WAAS/EGNOS) activated you
might see an improvement in the signal, but mind that if you are outside the
official coverage of such systems you might experience that these
corrections are in fact increasing the error. If you have access to other
forms of augmentation, such as IALA, make sure that you receive signals from
the closest station.

Even if you have access to good arial photography, remember that it might be
out of alignment, it can be a good advise to gather some good fixes to check
the alignment of your photos, this can be several GPS tracks along your
trail.

The problem of GPS devices drifting off is minimal, though more common in
small and cheap devices, in many cases simple augmentation can counter for
this, but the built in memory of your unit might also help in adjusting (for
better or worse).

On Fri, Jan 1, 2010 at 8:39 PM, Igor Brejc  wrote:

> You can also use Oregon's Waypoint Averaging function to make more accurate
> positioning of waypoints. But you need to do this at different times (say on
> you next hiking trip when you cross the same waypoint) for this to be really
> effective. With couple of accurate waypoints it is easier to detect track
> inaccuracies.
>
> Igor
>
>
> On Fri, Jan 1, 2010 at 9:18 PM, Michael Hufer wrote:
>
>> On the Oregon 550(t) you will find the satellite almanac-screen if you
>> touch
>> the "five-bars" satellite reception indicator.
>>
>>Micha H.
>>
>> > On Sat, 2 Jan 2010, Craig Wallace wrote:
>> > > You can check the satellite screen on the Garmin. It should show an
>> > > estimated position accuracy.
>> > > Also, you can look at which satellites its receiving. If its locked on
>> > > to a reasonable number of satellites in a decent spread across the
>> sky,
>> > > you can be fairly confident in its accuracy.
>> >
>> > The Oregon 550 lacks a pictorial representation of the almanac, and only
>> >  has five bars telling you whether it thinks it has good PDOP or not.
>> > Or it might, but as I've had mine 8 days , the same as Steve has, and it
>> is
>> >  in a menu I haven't found yet ;)
>> >
>> >
>>
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Igor Brejc]

You can also use Oregon's Waypoint Averaging function to make more accurate
positioning of waypoints. But you need to do this at different times (say on
you next hiking trip when you cross the same waypoint) for this to be really
effective. With couple of accurate waypoints it is easier to detect track
inaccuracies.

Igor

On Fri, Jan 1, 2010 at 9:18 PM, Michael Hufer  wrote:

> On the Oregon 550(t) you will find the satellite almanac-screen if you
> touch
> the "five-bars" satellite reception indicator.
>
>Micha H.
>
> > On Sat, 2 Jan 2010, Craig Wallace wrote:
> > > You can check the satellite screen on the Garmin. It should show an
> > > estimated position accuracy.
> > > Also, you can look at which satellites its receiving. If its locked on
> > > to a reasonable number of satellites in a decent spread across the sky,
> > > you can be fairly confident in its accuracy.
> >
> > The Oregon 550 lacks a pictorial representation of the almanac, and only
> >  has five bars telling you whether it thinks it has good PDOP or not.
> > Or it might, but as I've had mine 8 days , the same as Steve has, and it
> is
> >  in a menu I haven't found yet ;)
> >
> > ___
> > talk mailing list
> > talk@openstreetmap.org
> > http://lists.openstreetmap.org/listinfo/talk
> >
>
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Michael Hufer]

On the Oregon 550(t) you will find the satellite almanac-screen if you touch 
the "five-bars" satellite reception indicator.

Micha H.

> On Sat, 2 Jan 2010, Craig Wallace wrote:
> > You can check the satellite screen on the Garmin. It should show an
> > estimated position accuracy.
> > Also, you can look at which satellites its receiving. If its locked on
> > to a reasonable number of satellites in a decent spread across the sky,
> > you can be fairly confident in its accuracy.
> 
> The Oregon 550 lacks a pictorial representation of the almanac, and only
>  has five bars telling you whether it thinks it has good PDOP or not.
> Or it might, but as I've had mine 8 days , the same as Steve has, and it is
>  in a menu I haven't found yet ;)
> 
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk
> 

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Anthony]

On Fri, Jan 1, 2010 at 11:40 AM, Craig Wallace  wrote:

> On 01/01/2010 14:14, Steve Bennett wrote:
> > Well...ok. But in this case I have the aerial photography, so I can just
> > trace it, once I know more or less where the path goes.
>


>  Though yes, this is not really necessary if you have accurate aerial
> photography that you can trace from.
>

Of course, how can you know whether or not you have accurate aerial photos
if you're not sure of the accuracy of your GPS readings?

Aerials aren't always georectified correctly.  The more independent sources
you have, the better.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Liz]

On Sat, 2 Jan 2010, Craig Wallace wrote:
> You can check the satellite screen on the Garmin. It should show an 
> estimated position accuracy.
> Also, you can look at which satellites its receiving. If its locked on 
> to a reasonable number of satellites in a decent spread across the sky, 
> you can be fairly confident in its accuracy.
> 
The Oregon 550 lacks a pictorial representation of the almanac, and only has 
five bars telling you whether it thinks it has good PDOP or not.
Or it might, but as I've had mine 8 days , the same as Steve has, and it is in 
a menu I haven't found yet ;)

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Morten Kjeldgaard]

On 01/01/2010, at 17.40, Craig Wallace wrote:

> But I'd still agree with Shaun - a single GPS trace is not really
> accurate enough for adding ways to OSM IMO.
> I'd say get at least 2, preferably 1 in each direction. If they are
> close to each other you can be confident its probably accurate. If  
> they
> are much different, its worth getting a few more traces and taking an
> average.

I agree with this. I've noticed on my Garmin device, that if the  
satellite reception is lost or very poor, the device assumes it's  
continuing along a straight line using the most recently determined  
vector. Coming from a different direction, the device might be able to  
grab better hold of the the satellite signal and give you a better  
trace.

Cheers,
Morten


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Ulf Möller]

Craig Wallace schrieb:

> You can check the satellite screen on the Garmin. It should show an 
> estimated position accuracy.

The eTrex often claims 10m accuracy when in fact it is 50m off, so that 
doesn't really help. Using two different GPS receivers is a good idea if 
you don't want to survey twice.


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Craig Wallace]

On 01/01/2010 14:14, Steve Bennett wrote:
> On Sat, Jan 2, 2010 at 12:54 AM, Shaun McDonald
> mailto:sh...@shaunmcdonald.me.uk>> wrote:
>
> It is very common for GPSs to give errors for whatever reason.
> Interference is very common from things like buildings. Newer units
> are less likely to have an issue. You simply need to go along that
> track again a few times to get an averaged out reading.
>
>
> Well...ok. But in this case I have the aerial photography, so I can just
> trace it, once I know more or less where the path goes. I was just
> curious if there was a way to detect errors at the time - repeating
> every trace several times "just in case" would be pretty inefficient.

You can check the satellite screen on the Garmin. It should show an 
estimated position accuracy.
Also, you can look at which satellites its receiving. If its locked on 
to a reasonable number of satellites in a decent spread across the sky, 
you can be fairly confident in its accuracy.
You don't have any sort of WAAS like system in you part of the world do 
you? That would help a bit.

But I'd still agree with Shaun - a single GPS trace is not really 
accurate enough for adding ways to OSM IMO.
I'd say get at least 2, preferably 1 in each direction. If they are 
close to each other you can be confident its probably accurate. If they 
are much different, its worth getting a few more traces and taking an 
average.

Though yes, this is not really necessary if you have accurate aerial 
photography that you can trace from.

Craig

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Steve Bennett]

On Sat, Jan 2, 2010 at 12:54 AM, Shaun McDonald
wrote:

> It is very common for GPSs to give errors for whatever reason. Interference
> is very common from things like buildings. Newer units are less likely to
> have an issue. You simply need to go along that track again a few times to
> get an averaged out reading.
>

Well...ok. But in this case I have the aerial photography, so I can just
trace it, once I know more or less where the path goes. I was just curious
if there was a way to detect errors at the time - repeating every trace
several times "just in case" would be pretty inefficient.

Anyway, I'll try SLXViper's suggestion.

Steve
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[Shaun McDonald]

On 1 Jan 2010, at 13:07, Steve Bennett wrote:

> I've got a trace from today which is significantly out of sync with a path I 
> traced from Nearmap:
> 
> http://www.openstreetmap.org/edit?lat=-37.880138&lon=145.193417&zoom=19&gpx=594988
> 
> The trace looks like I was wandering through the grassy paddock, but I was 
> actually following exactly that northern most "highway=path" in the bush. So 
> it looks like the trace is incorrectly recorded something like 50m north of 
> where I actually was. Now, since the discrepancy seems to go away on that 
> track a bit further east (later chronologically), presumably the explanation 
> is the GPS data is faulty. Is this common? I'm new to GPSing, so I'm just 
> surprised. It's a Garmin Oregon 550. Is there anything I can do to reduce, or 
> at least detect, such errors?

It is very common for GPSs to give errors for whatever reason. Interference is 
very common from things like buildings. Newer units are less likely to have an 
issue. You simply need to go along that track again a few times to get an 
averaged out reading.

Shaun

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Defective GPS trace

[SLXViper]

Steve Bennett wrote:
> The trace looks like I was wandering through the grassy paddock, but I
> was actually following exactly that northern most "highway=path" in
> the bush. So it looks like the trace is incorrectly recorded something
> like 50m north of where I actually was. Now, since the discrepancy
> seems to go away on that track a bit further east (later
> chronologically), presumably the explanation is the GPS data is
> faulty. Is this common?

I've experienced the same from time to time. Seems to happen sometimes
with garmin devices... when you restart it, it's at the correct position
again.

> I'm new to GPSing, so I'm just surprised. It's a Garmin Oregon 550. Is
> there anything I can do to reduce, or at least detect, such errors?

Start your gps at a known position and look at the map. For eample when
you're standing near a road and your garmin tells you something
different, restart it.


best regards

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk