Mapping and GIS Solutions Community

 View Only
Back to discussions
Expand all | Collapse all

Accuracy and Consistency of DA2 Catalyst 1

  • 1.  Accuracy and Consistency of DA2 Catalyst 1

    Posted 01-05-2024 19:07
      |   view attached

    Hi,

    I subscribed to Catalyst 1 and began collecting points using a third-party app (MerginMap). The DA2 was connected to the RTX service and converged to cm accuracy before I initiated the work. I collected the coordinates of 10 points and repeated the process for four rounds to test the consistency and accuracy of RTX. However, I observed that the coordinates of each point varied in every round I surveyed them. The variation between each round for each point could exceed 1 meter, with the smallest difference being 5 cm. I made sure to achieve <2 cm accuracy before recording the coordinates of each point. I am puzzled as to why this issue occurs, as I anticipated obtaining cm-accurate positions with repeatable results when I subscribed to Catalyst 1.

    Additionally, I noticed that when I stood still at a point, the GPS location would jump around within a range of 20 cm, even though the DA2 was corrected to approximately 2 cm accuracy. Is this considered normal?

    Positions of the same point, which were collected in 4 separate rounds.



    ------------------------------
    Chen Zi Yan
    ------------------------------


  • 2.  RE: Accuracy and Consistency of DA2 Catalyst 1

    Posted 01-06-2024 14:27
    Hi Chen,
    I can't speak for RTX or the mobile software you are using and I really like someone like your self who tests.  Science folks call this calibration.  What your evaluating though is precision, not accuracy.   Here is a more robust test you can do, won't take you long to accomplish.   
    Make sure you in very OPEN SKY.   Setup your software to record XY in meter based projection/CS.  UTM, or state plane (meters).   Sounds like your doing one position and storing.   I would average, if you can 5 positions and storing said point.  Repeat 30 times with the antenna on a pole, steadied with a bipod.   Wait 4 hours for the constellation above you to change.  Repeat.

    Download your data into this spreadsheet, and you'll get the statistics desirable to see precision.  Horizontal accuracy calculation spreadsheet - Federal Geographic Data Committee (fgdc.gov).  Analyze each session (a session seperated by 4 hours), see if you get consistency amongst the sessions.

    You could be getting hit with high solar activity, we are in a high cycle, look for a recent XY HT magazine article on this, or the most likely, if your near obstructions, you could be getting multipath.  Stay in the open!.. Give that a shot and Trimble can better see what is going on with your precision.

    Good luck.




    Joel Cusick
    GIS Specialist





  • 3.  RE: Accuracy and Consistency of DA2 Catalyst 1

    Posted 01-06-2024 15:27

    Another variable is exactly how MerginMaps is getting its position.

    From your screenshot, it appears you're running it on Android.  If this is the case, do you have "Location Sharing" in Trimble Mobile Manager turned on?  Do you have TMM set as the Mock Location Application under Developer Settings?  It appears so, given the estimated accuracy shown on the screen recording.

    The next question is about the method used by MerginMaps for getting positions from the Android OS.  Many applications today are using what's called the Android 'FusedLocationProvider' API which can result in application behavior like you've observed.  Under the hood, locations come from a variety of sources (internal GPS, a-GPS, WiFi-assisted positioning, mock provider, etc.) and are then "fused" by the operating system and then delivered to the operating system.

    Trimble provides several options to bypass the uncertainty caused by the FusedLocationProvider in applications which demand unwavering location accuracy.  If you have a contact at MerginMaps we'd be happy to discuss those options with them.



    ------------------------------
    Dan Colbert
    ------------------------------