I have conducted a flight with the Matrice 350 RTK connected to a proprietary base via NTRIP. When processing the LiDAR data, no errors appear, but the last part of the flight is missing. I’m attaching an image of the point cloud and the orthophoto.
I thought it might be an issue with the RTK, so I tried using PPK, but I encountered the same problem. I’m also attaching a post-processed dataset with Emlid Studio (all points are marked as FIX).
Do you know what might be causing this or how I can resolve it?
Thanks.

What source files did you use in Emlid on the rover side?

LV_Forestry Posted at 12-10 09:03
What source files did you use in Emlid on the rover side?

When I set up my base to broadcast via NTRIP, I always make it record raw data to have a RINEX file in case I need to perform PPK.
Rover data, the drone's MMK, and the file `DJI_20241205154951_0002_Zenmuse-L1-mission_PPKRAW.bin`, changing its extension to RTCM3 and then converting it.
Looking into it further, I realized that the total flight time is 24 minutes, but the LiDAR only has 17 minutes of data capture. For some reason, it stopped recording LiDAR data but continued the mission and kept taking photos.

djiuser_LgTryJ7CDX9v Posted at 12-10 23:08
When I set up my base to broadcast via NTRIP, I always make it record raw data to have a RINEX file in case I need to perform PPK.
Rover data, the drone's MMK, and the file `DJI_20241205154951_0002_Zenmuse-L1-mission_PPKRAW.bin`, changing its extension to RTCM3 and then converting it.
Looking into it further, I realized that the total flight time is 24 minutes, but the LiDAR only has 17 minutes of data capture. For some reason, it stopped recording LiDAR data but continued the mission and kept taking photos.

Use .RTK instead of PPKraw in Emlid. Maybe the mistake is there. Anyway it wont solve the issue.

Also you can try with a different base station. Even if located much farther. Sometimes Terra has surprising attitude.

LV_Forestry Posted at 12-10 23:23
Use .RTK instead of PPKraw in Emlid. Maybe the mistake is there. Anyway it wont solve the issue.

Also you can try with a different base station. Even if located much farther. Sometimes Terra has surprising attitude.

Hello! I've used RTK instead of PPKRAW, and now everything is in FIX, although it doesn't actually solve anything :_(  
Anyway, I think the issue is with the sensor, which stopped recording data prematurely. Any idea what it could be?

djiuser_LgTryJ7CDX9v Posted at 12-10 23:28
Hello! I've used RTK instead of PPKRAW, and now everything is in FIX, although it doesn't actually solve anything :_(  
Anyway, I think the issue is with the sensor, which stopped recording data prematurely. Any idea what it could be?
[view_image]

Mine does that sometimes as well.
Format the SD card with Pilot 2 often. It may help.

Don't you have error message,"SD wrting slow" while flying ? Or "overloaded CPU"?

LV_Forestry Posted at 12-10 23:52
Mine does that sometimes as well.
Format the SD card with Pilot 2 often. It may help.

Hi hi,
I dont remember, any idea if i can check old flight logs?
I never try the SD format I will try, the strange thing is that the images were saved without any issues.
Thanks!

djiuser_LgTryJ7CDX9v Posted at 12-10 23:55
Hi hi,
I dont remember, any idea if i can check old flight logs?
I never try the SD format I will try, the strange thing is that the images were saved without any issues.

Yes you can load log file on phantom help. Then if there was error message they will appear.
Its time consuming to do that. There is a forum user who make such powerful dataviewer, JJB. You may contact him if you really want to see all flight log.

Also in Terra, you can generate quality reports. Sometimes there are error written, which are not displayed when the point cloud is built successfuly.

Welcome to the DJI LiDAR world!

LV_Forestry Posted at 12-11 01:30
Yes you can load log file on phantom help. Then if there was error message they will appear.
Its time consuming to do that. There is a forum user who make such powerful dataviewer, JJB. You may contact him if you really want to see all flight log.

Thank you very much for your response. I’ve been talking to DJI support, and in the end, they’ve asked me for the entire project to figure out what might be going on because they still think it’s an RTK issue, even though I’ve done PPK with three different bases. I’ve sent them all the information to see what they say, and I’ll let you know.
I’ve been flying with the L1 for almost two years now. In the end, I had my methodology well implemented, and with the control points, I don’t get more than 2-3 cm of difference, so it’s giving me good results. After all, photogrammetry, being an indirect measurement method, depends a lot on the processing you do to maintain consistency throughout the work.
This is the first time I’ve encountered an issue I can’t resolve. Let’s see what I can figure out.
Thanks so much for your help!

djiuser_LgTryJ7CDX9v Posted at 12-11 06:26
Thank you very much for your response. I’ve been talking to DJI support, and in the end, they’ve asked me for the entire project to figure out what might be going on because they still think it’s an RTK issue, even though I’ve done PPK with three different bases. I’ve sent them all the information to see what they say, and I’ll let you know.
I’ve been flying with the L1 for almost two years now. In the end, I had my methodology well implemented, and with the control points, I don’t get more than 2-3 cm of difference, so it’s giving me good results. After all, photogrammetry, being an indirect measurement method, depends a lot on the processing you do to maintain consistency throughout the work.
This is the first time I’ve encountered an issue I can’t resolve. Let’s see what I can figure out.

They are always very polite at DJI, always ready to help us. Unfortunately I am convinced that nothing much will come out of this data analysis.

A problem with the RTK! Of course yes, in any case if it is not that there are not many other options left. What would be good is to know why this happens and how to prevent it?

And also, have Terra show error messages at the end of the rebuild, not just in the quality report, that would be perfect.
Well no, what would be perfect is if the sensor worked as it is supposed to.

Anyway, you're welcome, let us know if you have news... if there is one.

djiuser_LgTryJ7CDX9v Posted at 12-11 06:26
Thank you very much for your response. I’ve been talking to DJI support, and in the end, they’ve asked me for the entire project to figure out what might be going on because they still think it’s an RTK issue, even though I’ve done PPK with three different bases. I’ve sent them all the information to see what they say, and I’ll let you know.
I’ve been flying with the L1 for almost two years now. In the end, I had my methodology well implemented, and with the control points, I don’t get more than 2-3 cm of difference, so it’s giving me good results. After all, photogrammetry, being an indirect measurement method, depends a lot on the processing you do to maintain consistency throughout the work.
This is the first time I’ve encountered an issue I can’t resolve. Let’s see what I can figure out.

I'm thinking of something here, your base station rinex files, what constellations do they use?

GPS
GLO
GAL
BDS

If these are GPS/GLO only, this may explain why some pieces are missing.


DJI Terra Altitude Error on process | DJI FORUM in this thread the issue is the same.
and i can reproduce it by taking base station from more than 25km, using GPS and GLO only. As soon as I put GAL and BDS back, there is no more missing part.


I play with a Rinex file that has all 4 constelation.
With RTKlib i extract only GPS and GLO:



Terra is giving this results :


When i put back GAL and BDS, it fix the issue :

I can reproduce the problem in 100% of cases, for different datasets. So it seems that this is a cause with a strong potential.

GPS + GLO :



GPS + GLO + GAL + BDS :

It does seem like a systematic error with DJI Terra, but I’ve been checking the RINEX files from my stations, and my personal one has all the constellations, as do those from the Spanish public network service. They all already have at least those four constellations.

Reviewing the data, I still see signs that the sensor stopped recording information. Today, I’m going to extract the logs to see if there was any error with the SD card. In any case, I’ve formatted it using DJI Pilot, and next week I’ll do a test flight.

Thank you so much!

@LV_Forestry interesting find! Of course, halfway decent software would warn you of such an issue...

patiam Posted at 12-12 08:58
@LV_Forestry interesting find! Of course, halfway decent software would warn you of such an issue...

It's us who are too stupid, Terra is much too evolved for us. I think we need a special diploma to use it.

In fact the error message exists, but to see it you have to generate the quality report each time, and scroll a little to find the message.
I don't see the point of doing this at all. It would be so simple to display the message on a popup window like any other software for normal people does.

Good morning!  
I think I’ve found the issue; it seems to be a problem with Terra. When the LDR file exceeds 4GB, a second LDR-1 file is created, and when I process it, Terra only uses the first LDR. I’m not sure if I’ve been doing it wrong for the past two years and just got lucky with flights where the second LDR wasn’t a crucial part of the project, or if it overlapped enough with the next flight to avoid noticing its absence.  
Do you know if this is an issue with Terra or if it’s the normal workflow?  
Thank you very much!