[Qgis-user] wishing for accurate lattitude/longitude from a cell phone

Mon May 25 09:25:38 PDT 2020

Dear list,

I would like to add some things here, since I am trying to use RTK GPS
in the field with mobile devices.

> 1) Accuracy of GPS Devices

RTK GPS or any more precice GPS technology will be external, you cannot
get around the missing measurements and lack of algorithms in consumer
grade chips.

Then it depends on the capabilities of the device to receive satellites
of all systems (one or multiple frequencies/signals) and be able to
apply correction data (a question of algorithms). For better equipment
the prices rise quickly.

There are some low cost devices evolving, a new chip was recently
announced: Skytraq PX1122R for about $100. Could be tested with a break
out board and antenna, e.g. see board at navspark shop. The module even
supports PPP, which might be especially interesting where mobile network
is not available.
https://navspark.mybigcommerce.com/px1122r-evb-px1122r-multi-band-quad-gnss-rtk-evaluation-board/

Other low cost solution were mentioned (emlid/REACH, Catalyst, etc).

Am 23.05.20 um 21:51 schrieb Michael.Dodd:
> One app that claims to do a lot of what high precions gps does is> Mobile Topographer Free – Apps on Google Play

> 2) Software

On Android I was able to get the external signal into the system, you
need the app "Bluetoth GPS" (or similar) or a USB2serial + app (better
avoid tiny plugs and large cables during field work). Additionally you
need to override the internal GPS position with the Blue GPS App in the
developer settings (mock provider).

My impression was, that most Android applications target the accuracy of
the device only, so e.g.

* you cannot zoom to cm-levels, only ~100m
* not many proper GIS Apps are available, most are expensive
* Tracking usually does not rely on cm-grade positions, so the apps are
not made for it

One particular thing is, that with high accuracy of the signal, an
internal computation in single precision float will not suffice (all
computations must be in double precision), and you may end up with a cut
off of the last position digits, e.g. if you have 8 digits, 4 before and
4 after the decimal separator (e.g. in DDMM.MMMM format), then you end
up with coordinates cut down to 2-3 decimeters in the real world
(typical GPS mouse output, not so uncommon).

So, looking into that aspect may be required, and only few Apps may
implement that (e.g. expensive "surveyor" apps for professionals? did
not test).
BTW, sponsoring the double accuracy implementation for QField is
welcome, as mentioned ot the webpage. It is already a very usable and
FLOSS GIS solution for the field, especially useful if you have a QGIS
workflow already ;)

Finding proper solutions and recommendations still required...

> ------------------------------------------------------------------------
> *From:* Qgis-user <qgis-user-bounces at lists.osgeo.org> on behalf of
> j.huber at post-ist-da.de <j.huber at post-ist-da.de>
> *Sent:* 23 May 2020 20:35
> *To:* qgis-user at lists.osgeo.org <qgis-user at lists.osgeo.org>
> *Subject:* Re: [Qgis-user] wishing for accurate lattitude/longitude from
> a cell phone
>  
>
> Hi Steve,
>
> I think there are two different aspects to your question:
> 1) Accuracy of GPS Devices
> I agree with others that you can't expect a better accuracy than the 10
> to 30 feet you observed with a smartphone or a simple handheld GPS. This
> might improve by averaging, but this takes time. Survey grade GPS
> devices are very expensive. It is possible to get good modules and
> antennas as components, but building a complete GNSS system requires
> time and skill. This would reduce cost, but only to several hundred
> instead of several thousand dollars, so this is no option in your case.
> Besides, you will still need a correction data service (usually costly)
> or use two receivers (rover and base).
> So you probably have to stick with your phone, which is more flexible
> regarding the software than a handheld GPS.
>
> 2) Software
> The accuracy of the recorded position should not depend on where you tap
> the screen - a good app should allow to record the current GPS position.
> I am using Locus Map (Asamm Software) for a while now, it works quite
> well, although it does not allow position averaging. There is a free
> version, you could try that first.
>
> Regards,
> Jochen
>
> Am 22.05.20 um 20:54 schrieb Stephen Sacks:
>>
>> In order to make widely available some wise advice, I'm sending to
>> this list a message I received from Neil B.  In addition to Neil's
>> message below, I want to mention that Nicolas Cadieux also provided
>> similar information, saying I'd have to pay around $1,000 for
>> equipment that gives consistently accurate location coordinates.  And
>> thanks, also to Falk Huettmann and Bernd Vogelgesang for their replies. 
>>
>>
>> Message from Neil B:
>>
>> Hello Stephen.
>> Glad that you're having success. I would like to start off by saying
>> that it is best to always reply to the mailing list and not directly
>> to the person who submitted the email. Mailing lists work really well
>> in that there is a pool of people out there who may be able to offer
>> advice or may have an alternate method to solve the problem that may
>> turn out to be a better way. On the flip side by maintaining the email
>> chain through the mailing list, the follow up emails that provide
>> information are stored in the archives which benefits anyone searching
>> the internet to have the complete trail of information.
>>
>> As far as your results they are acceptable for the device you're
>> using. GPS in phones are never built to precision survey standards and
>> there is no reason for them to be. If you're within 30ft of where the
>> phone thinks you should be then you can easily navigate the rest of
>> the way by visual sight. High end equipment to achieve sub-inch
>> accuracy is probably in the range of thousands of dollars. One thing
>> to keep in mind is there is a difference between the accuracy of a
>> device and to what level of precision they display. While the app on
>> the phone may display 8 decimal places of a lat/long coordinate and
>> tell you if you have moved a foot, it doesn't help that the coordinate
>> it is displaying is out +/- 30 feet. The accuracy of a device can also
>> be affected by the environment where the device is being operated. In
>> regards to cell phones, they use multiple sources to determine
>> location such as GPS, cell phone towers, and wifi points to perform
>> the triangulation. Lack of line of sight to satellites, signals from
>> cell towers bouncing off of surrounding buildings, or someone's
>> wireless router using inaccurate position information can all affect
>> the accuracy of what is being displayed on your phone.
>>
>> So the question is how are you determining that the coordinates are
>> wrong? If you have information that you trust to be authoritative then
>> adjust your points to those values and carry on. I have no advice or
>> opinions on inexpensive devices that may help with a more accurate
>> reading.
>>
>> Please do not respond directly to me. This email account is not
>> actively monitored and I don't always have the time to follow up with
>> the emails. All the best with your endeavours.
>>
>> ~Neil B.
>>
>> On Fri, May 15, 2020 at 7:52 PM Stephen Sacks <sacks44 at earthlink.net
>> <mailto:sacks44 at earthlink.net>> wrote:
>>
>>     Hi Neil,
>>
>>        With your help, I have successfully brought the corners of our
>>     gardens back from Pennsylvania to the Promenade here in Brooklyn
>>     Heights, New York.  Thank you.
>>        At the risk of wearing out my welcome, I'm now asking for more
>>     advice.  My point features are approximately where they should be
>>     but not exactly, some points are just a few feet off and some are
>>     10 or even 30 feet off.  I imported the data trying both EPSG 4326
>>     and 4269.
>>        I'm now convinced that the problem is due to (1) my Google
>>     Pixel 3 cellphone, (2) the app I'm using ("Latitude Longitude"
>>     published by gps-coordinates), and  especially (3) my
>>     less-than-steady hands.  I capture coordinates by standing at
>>     spot, waiting for the blue dot to settle, and then touching the
>>     blue dot.  Often I don't touch the screen at exactly the right
>>     place.  I tried another app ("GPS Coordinates" published by
>>     Financept) which is better in that it allows me to zoom in, but
>>     I'm still not always getting it right.
>>       I'm thinking now that I need specialized equipment.  That is
>>     what I want to ask you.  Can you recommend some inexpensive device
>>     that will allow me to simply press a button to record accurately
>>     the coordinates of the point where I'm standing?  Keep in mind
>>     that this is a community project with no funding.  I live on
>>     Social Security and a university pension, but I'm willing to pay
>>     something in the range of $50 or a bit more.  Do you know of
>>     anything at such a modest price, or would I have to pay much
>>     more?  Or perhaps you know of better software for my Android Pixel
>>     3 phone.
>>          Thanks in advance for any advice you might offer.
>>                       Steve
>>        
>>
>>
>> _______________________________________________
>> Qgis-user mailing list
>> Qgis-user at lists.osgeo.org <mailto:Qgis-user at lists.osgeo.org>
>> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
>> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>
>
>
> _______________________________________________
> Qgis-user mailing list
> Qgis-user at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-user
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-user
>