Dual GPS with RTK and georeferencing images

Is there anyway to geoference images with the secondary (not switched) GPS rather than the primary one?

I have dual GPS setup, M8N as the primary, and an RTK system as the secondary. Id obviously like the RTK systems data to be the one used for referencing data, but dont want to put it in control of positioning the machine quite yet :wink:

Cheers, Chris

1 Like

Hi Chris,

I am interested in alike Duel GPS with RTK setup.
I am in touch with Ublox to order M8 raw data GPS units.

Tell me if your first GPS is ground control point static fixed
and what type of data transmission do you suggest : radio, WiFi, 3G
to work with rover on your drone ?

And finally tell me if you have noticed better GPS accuracy in altitude
to keep drone immunized against fly-away.

darius
thank you

Hi, I will add something to do exactly this,

please check the latest beta MP.

^ You sir, Rock :ugeek:

Thanks!

[quote=“darius”]Hi Chris,

I am interested in alike Duel GPS with RTK setup.
I am in touch with Ublox to order M8 raw data GPS units.

Tell me if your first GPS is ground control point static fixed
and what type of data transmission do you suggest : radio, WiFi, 3G
to work with rover on your drone ?

And finally tell me if you have noticed better GPS accuracy in altitude
to keep drone immunized against fly-away.

darius
thank you[/quote]

Im using a Piksi on the GCS, and in the air as the secondary GPS, M8N is the primary GPS.

The airborne Piksi talks to the ground station one using a second SiK 915Mhz radio.

Accuracy is certainly better, as long as the position that is fed to the ground station piksi to broadcast is accurate.

The datalink between the two seems to be the weak part in the chain. I dont trust it enough to put it in control of the aircraft yet.

With EKF etc, something has to really go wrong with a Pixhawk to get an actual flyaway. Ive seen a GPS totally die in flight and the Pixhawk failed safely into altitude hold (and did it multiple times!)

I dont think RTK is the solution to “flyaways” as while it is more accurate, it is also more fragile than standard GPS due to the high sensitivity of the antennas and the datalink between the two.

Hi,

I’d like to do the same as Chris.
How is it possible to do that? Checked the latest beta MP, but haven’t found anything related.

Gabe,

in beta MP i added a tick box to base all messuremetns off the seconds gps, instead of the first.

Heya Michael,

Id appreciate your insight into this if you have 5minutes?

One mission, images processed with GPS1 (M8N) and then processed again using GPS2 (Piksi with Fixed RTK lock, using second SiK radio)

Difference between the two in accuracy… negligible :smiley:

The RTK Base position was gathered using a gps stat 4 fix on the M8N, so probably isnt helping the accuracy, but I wouldve expect the error to be consistent.

One thing I was not able to do was to enter the cameras latency, anything other than 0ms in the MP tool resulted in an error in preprocessing (something about time not found, I shouldve written it down)

https://drive.google.com/open?id=0B8h_12QqMnqpeWNOZXpCd1A0YmM
https://drive.google.com/open?id=0B8h_12QqMnqpbEtuLUV2OGpSdWs
https://drive.google.com/open?id=0B8h_12QqMnqpMl9SdktXalViX2c

Rai from Swift Nav replied with this

Dear Chris,

I looked into the mission-planner tool that you used to extract geo-reference data. I ran a data set of one of our survey missions through that and compared it to our post-processed tools. The errors were all over the place between them. Ranging from 0.001m to 1.8m, these errors had no pattern to them. I leaned on trusting our data as we know how it is being calculated.

A bit of an insight as how we post-process Piksi data to extract geo-reference data. We are linearly interpolating at the GPS TOW (Time of Week) the trigger message is captured. I believe that mission planner is not doing that. I think, mission planner is taking the closer point in the logs to the CAM messages.

Could you describe how you landed on 0.28 seconds delay ? It can’t be an exact number for all images captured or is it ?

Looking forward to your reply.

Best

Rai Gohalwar

Bit more information over in the Swift nav forum if your curious.

Im going to take a look at hot shoes :slight_smile:

Thanks, Chris

Hi to all,
i see that someone had configured an M8n to store raw data.
I have the same configuration but no way to extract rinex.
Is it possible to have a configuration step to enable logging raw data from the M8n as secondary gps?
thank you,
Marco.

just use the standard df log conversion, and it will create a obs file with rinex obs in it by default.

Sorry, but I didn’t understand the answer about the rinex obs file with the M8N gps unit. I have a dual gps with a M8N as first gps and M8N as second gps plug in serial 4/5. The GPS_RAW_DATA=1. What do you mean with "standard dataflash log conversion¨? I convert the log file into ¨KML+GPX¨ but i don’t find the rinex file.

thank you in advance

m8n does not support raw data output.

Hi,

Thats what i thought: m8n doesn’t support raw data.

I am thinking about changing the 2nd m8n unit with an old LEA 6T, which I
think it supports raw data. Is there any configuation profile recommended
for using this gps units configuration in order to get that rinex files?

Thank you