Performing PPK with the Autel Evo 2 RTK and CORS Data

The Autel Evo 2 RTK drone has a built-in RTK GNSS receiver. When you fly the drone, it automatically records the GNSS data to the SD card. Even without a dedicated base station, these data can be post-processed in order to achieve centimeter-accurate positioning for each captured image.

1. Download and install Emlid Studio.
2. Download and extract RTKLib Explorer.
3. Download the captured data from the SD card. In each session folder, along with the images, there should be three files produced by the GPS: a .bin file, a .obs file, and a .mrk file.
4. Figure out the nearest CORS base station. The easiest way to do this is to use the map provided by NOAA. Determine the four-letter code for the station. For instance, the Tifton station has the code “GATF”.
5. Download the corresponding CRX data for this base station at the correct time. The format for these data is as follows:

   curl -O https://noaa-cors-pds.s3.amazonaws.com/rinex/YYYY/DDD/ssss/ssssDDD0.YYd.gz

   Here, “YYYY” is the four-digit year, e.g. “2025”, “DDD” is the day of the year, “ssss” is the station code, e.g. “gatf”, and “YY” is the last two digits of the year, e.g. “25”. The easiest way to find the day of the year is to use Python’s datetime library, and subtract December 31st of the previous year from the flight date.

   Extract this file and rename it with a .crx extension.

6. Use the “crx2rnx.exe” program from RTKLib Explorer to convert the CRX file you just downloaded to an RNX file. You should be able to simply run it in the terminal with the path to the CRX file as the only argument, and it will automatically output the RNX file in the same directory.
7. Open the “rtkconv.exe” program from RTKLib Exporer. As the input, select the RNX file you just created. Make sure the first checkbox for the .obs file is selected. Click the “Convert” button to generate the .obs file.
   
8. Download the navigation file from NASA’s archive. Note that this archive requires a login, so you might have to create an account. Select the proper year, and then the “brdc” folder. Inside this folder, the file you want is named “brdcDDD0.YYn.gz”, where “DDD” is once again the day of the year, and “YY” is the last two digits of the year. Download this file and extract it.
9. Open Emlid Studio. For the Rover file, select the .obs file from the drone. For the Base file, select the .obs file you generated from the base station data in step 7. For the navigation file, select the extracted .YYn file that you downloaded in step 8.
   
10. Press the “Process” button to perform the PPK analysis. A good PPK result should be mostly fixed. Note that you might have to play around with the settings (accessed through the gear button) to get a good result. Some settings that I’ve found work well with this type of data:
    – A low elevation mask, of around 5-10 degrees
    – A slightly lower SNR mask, perhaps around 20.
    – Set “Integer ambiguity resolution” to “Fix and Hold” for GPS and GLONASS
    – Set “Filter type” to “Combined”