RS+ Base / Reach M+ Rover (add LORA?)

Has anyone experienced an EMLID RS+ base transmitting correction data to an EMLID Reach M+ on a drone for PPK (EDIT: RTK) mapping?

1 Like

No corrections are being sent in a PPK scenario. That’s what the post-processing does. RTK or Real Time sends the corrections. You can get a LoRa antenna for an M+ though. I actually have one that I am going to test by putting it on my vehicle instead of the magnet mount and RS+ rover that I normally use. I thought about trying it on my drone, but without the ability to tag the images in real time it’s kind of a waste. I’m hoping that this is what the a-shot does. Not associated with this thread.

1 Like

Experienced, nop, but is it possible? Yep!

But if you are doing ppk, why would you need Lora and an rtk solution? For navigation?

I guess the LoRa radio add-on would be handy for RTK for a future multi-frequency M++ receiver like the new RS2.

1 Like

But only if the image files can get the coordinates at save time, with the same precision as usual.
I have not yet seen a non-integrated solution, that can do that.

Or of course to get rid of PPK all together, but that has is it drawbacks.

1 Like

Sorry Michael… You’re right. I should have said RTK. We do everything PPK for our sites … slip of the finger. :slight_smile:

Christian, I agree. We fly with GCP’s and process PPK. However, its time consuming on a large construction site to replace GCP’s as earthworks takes place and digs existing ones out. We are currently exploring an M210 with a topodrone customized camera that is connected to a Reach M+. This is still a PPK process, as we take the logs from the base and rover, perform the corrections, update the EXIF image tags with the corrected data and then perform a final process on the cloud. I guess I started asking myself, could we not add a lora antenna to the reach M+, send corrections from the base… But as previously discussed, if we hit a FLOAT, which we have, then data is not good and you have to rerun. and yes, wouldn’t it be nice if the updated metadata were written directly to the images, rather than the steps needed.

This should not be the case with PPK. I previously used 10-15 GCP’s at 50 acres and then incrementally went up as size increased, but I have not used more than 5 or 6 since I have started using PPK. All the GCP’s are doing at this point are keeping the model in check and localizing. I do agree with @timd1971 statement above that if we can get a multi-frequency M+ then paired with the RS2 is a valid RTK solution. This is how the Phantom 4 RTK works, but I still would not trust a pure RTK solution on a drone. Especially if you are in any level of adverse conditions.

1 Like

Thanks Michael. I agree with the 5 or 6. We maintain many due to the number or rock trucks and excavators performing work and stockpiles being moved. I’m a fan of PPK but still testing our RTK solution with Reach M+. We have found float shortly after take off and fix is established on the first few turns and then back to fix, but that just means I continue tweaking with RTK while we maintain PPK for our work. Hence the thought of a Reach M and Lora antenna, connected to an RS2.

1 Like

IF a M2+ is used in RTK, it would most likely perfom better then M+ in terms of float vs fix but the PPK will also be better.
Unless there is a need for the RTK feature i would still pick PPK over RTK with a M+/M2+ unit.
As of now you would still need a ok link between base and rover for the RTK and from what i have seen, a small banking (depends on the situation) in turns will drop a connection, a M2+ would not save you there other then regaining a fix faster ones the connection is up again.
Until SSR is up and running and if the accuracy is good enough, refrence data is crucial in both RTK and PPK.

1 Like

Theory of the number of GCP’s is from setting GPS control networks. The site needs to be “boxed” in with 4 and one in the center. One more may be needed if the site configuration is odd or the points get too far apart. It use to be about 600-800ft, but with PPK 1600-1800ft works fine. You can verify this with midpoint checkpoints.

We are using as few as 7 and as many as 10 GCP’s in our PPK missions. Some are used as GCPs and others as check points. We get 2cm or better accuracy. What we are testing now is an RTK solution where the ReachM is recording on an M210 with a modified X4S. With the base and drone recording seperately, it brought me to think that with both having LORA capabilities…why not send corrections off the known point base… needless to say it was a passing thought and now I’m taking into consideration the changes in LLH and speed of travel.

1 Like

Spot in, thanks for sharing.

Hi @sstacey,

May I ask you to clarify why you need real-time corrections with such a setup?
As was mentioned above, it’s not possible to record events in the RTK solution.

The only benefit of LoRa in your case would be for more precise navigation, given that the flight controller would accept an external nmea stream.

If the base is on a known point, then moving about with a rover on the ground, you can get fairly accurate elevation readings. Corrections coming from the base over LoRa. While all of our drone work is done in PPK, using GCP’s, which we maintain and rebuild as large construction areas tend to take them out as work progresses.

I was thinking that if the base was on a known point, and the Rover was a drone, then it would receive corrections over LoRa. That being said, I’ve realized that a moving drone, vs a rover on a tripod is still applies to oranges when looking for accuracy.

We are exploring the Topodrone camera and ReachM+ solution but have yet to dial in any suitable accuracy, in comparison to our GCP’s and PPK process.

No, we would fly our M210 RTK solution when requiring precision accuracy. Thanks.

1 Like

It’s unlikely that RTK is going to be more accurate than PPK, but more importantly safe and consistent. There are no refinements that can be made with RTK like SnR filtering, Mask adjustments and taking out poorly performing satellite constellations. Personally I would never spend that much money on a “factory” install unless I was so worried about asthetics. The Tuffwing kit performs very well, is easily removable and is modular so that it can be easily self-repaired. Not to mention it is about $2k less as a full kit and $2500 as an upgrade.

Agreed. Perhaps I should just say our M210. While it has RTK capabilities, that is used for positioning of the drone for inspections, etc. All of our drones are using PPK for precision measuring currently, using GCP’s.

Unfortunately, the two RS+'s we are using have been unreliable as of late. Cannot seem to get a consistent lock and have randomly shut down on us as well. One minute we have an AR validation of 999.9 and the next we are 1.1 to 1.7. The conditions have been good with BASE showing 19 and ROVER showing 18 satellites, 9 in GREEN, we are 15 and 35 using GPS and GONLASS, GONLASS AR is ON, LORA was been fine Frequency at at 868, Output at 20 dBm and 2.6kB/s. That being said we have tried 18dBm and 9.11kB/s and no improvement.

Its taken time to get a fix but for the most part, worked well. However this last month or so has been very frustrating. GCP’s are time consuming at the best of times but when you’re solution starts becoming hit and miss… :frowning:

1 Like

Hi @sstacey,

Could you please elaborate in which conditions the units were while it happened? Could you send me Full system reports from both devices in PM?

May I ask you to share raw data logs recorded during the survey from both units and a solution log from the rover one?

You can download them in the Logging tab.

I’d like to examine your RTK solution and observation data to figure out what might provoke such behavior.

1 Like