Poor Reach Connection In-flight

[spiwakd]

Hello all,
I am trying to integrate Reach systems into my drone flights to geotag photos more accurately for later use in Agisoft Photoscan
I am also very new to these systems so my apologies if I ask obvious questions.

Firstly, I am flying an Event38 E386 Fixed Wing mounted with an R10C camera connected to a Pixhawk programmed with Mission Planner.

                                                   My current settings are:
                          
                                   Base Station:

Positioning Mode: Static
GPS AR mode: Fix and hold
GLONASS AR mode: off
Elevation/SNR mask: 15/15
GNSS Select: GPS,GLONASS,SBAS,QZSS
Update Rate 1Hz
(I am unable to get float/fix solutions… i think this may be part of the issue?)

                                 Rover:

Positioning Mode: Kinematic
GPS AR mode: Fix and hold
GLONASS AR mode: off
Elevation/SNR mask: 15/15
GNSS Select: GPS,GLONASS,SBAS,QZSS
Update Rate 1Hz

My problem:
It seems as though the Reach lost connection with the rover mid-flight, and did not collect data.
Furthermore, the data the drone did collect was poor.
If any of my settings are glaringly incorrect, what should I be doing instead.

What other information should I supply?
How can I improve my data collection, and get more float/fix points rather than the singles?

I have attached the pos file if that helps at all…
raw_201807131511.obs_formatted.pos (184.7 KB)

[andrew.yushkevich]

Hi @spiwakd,

@spiwakd

What other information should I supply?
How can I improve my data collection, and get more float/fix points rather than the singles?

Based on the workflow you’ve described, I would recommend you to use PPK (Post Processing Kinematic). The main advantage of this method is that you don’t need any connection between the base and rover during the flight, everything is done in post-processing.

For post-processing, you need to collect logs from base and rover with your timestamps and post-process it with Emlid RTKLIB version following this doc.

@spiwakd

If any of my settings are glaringly incorrect, what should I be doing instead.

All your settings look fine.

[spiwakd]

@andrew.yushkevich
I followed the Kinematic processing | Emlid Studio
However I still get the same result from my original post.
That was not the full flight, it did a full grid survey of ~25 hectares, and the RTKplot seems to drop out right when it started the grid.

An addition:
For the PPK solution, I do not have to have the two systems talking to each other? I thought this increased the accuracy and allowed for fix/float points?

[andrew.yushkevich]

@spiwakd

That was not the full flight, it did a full grid survey of ~25 hectares, and the RTKplot seems to drop out right when it started the grid.

Could you share your raw logs and hardware setup photo, please?

@spiwakd

An addition:
For the PPK solution, I do not have to have the two systems talking to each other? I thought this increased the accuracy and allowed for fix/float points?

The idea of PPK is that you don’t need any connection between base and rover during the flight, everything is done in post-processing using raw logs from devices. That’s why you could work without cellular or other correctional networks.