I have spent the past couple of days trying to get my head around the Reach RS system and how to use it to create Ground Control Points to increase accuracy on my 3D drone surveying, but am still struggling to grasp it. I plan to build 3D models for architecture companies instead of them using ground surveyors, they need the models at the CM level of accuracy.
I am in the UK and have the following:
1x Reach RS Unit
1x Reach M+ Unit (The older kind without any casing)
DJI Inspire 1 Pro
From my reading I think I understand that PPK is the best way to go using GCPâs using a DJI drone. I plan to use tiles as GCPâs and understand that I need to keep the RS as a base unit and then take the rover unit around to each GCP to measure them.
Where I am confused is the following:
How do the two devices talk to each other. I can connect to each oneâs bluetooth and see they are receiving signals but how does the Reach RS unit know where the M+ rover is capturing the GCPâs.
Will using the M+ as a rover to collect GCPâs Work or do I need a second RS unit.
What is the best workflow to collect the GCPâs.
Really appreciate any help with this as I have these great bits of tech but struggling how to use them!
RTK is âReal-time kinematicâ as in getting instant results
PPK is âPost-processed kinematicâ as in getting results later after a processing job is done.
For you, the easiest thing will be to do PPK. The devices donât talk to each other. They just record log files. Then later you download them and process them with the post-processing software (usually RTKLIB - see the tutorial in the docs)
There are also some 3rd party software apps that sort of do the post-processing for you.
Hi Bide thanks for your really helpful reply, it is much appreciated.
That makes a lot more sense now! Do you also happen to know if I can use the M+ as my rover unit or do I need to invest in a second RS unit.
From my understanding relative position is fine. I am creating the 3d models so the architect can pull them into a CAD program so as long as the measurements between the building itself are correct I donât think it matters on absolute.
Any Emlid Reach device can be used as either base or rover (original Reach module, M+, RS, RS+). Actually they can even act as base and rover at the same time (which is not normally done, but possible).
Also, and I hope it is not confusing, but when collecting PPK data, technically there is no base or rover, only data (as log files). It is during the post-processing job that you must decide which log file will be used as the âbaseâ and which will be used as the âroverâ.
Of course you will put one unit over a known point and you will call it âthe baseâ and the other unit will be called âthe roverâ. It is just that the data doesnât know that yet. And this info is regarding PPK only. It is different for RTK.
You have seriously been such a great help thank you! I think I was overthinking things a bit.
Does this mean that essentially I can just use the RS device and put it over a point and make a note that it is my base. Then move the same RS device around on top of each of my GCPâs and then when I come to post processing as long as I know which point was my base I do not need a second device? If so do I need to change the base settings to accomodate for it not being static?
Assuming that is correct then my GCPâs will be relative to the base and should provide my 3D drone model relative accuracy.
Two GNSS receivers must log data simultaneously in order to do PPK (or RTK) processing.
The theory is that one receiver is over a stationary âknownâ point. That would be your âbaseâ or a third-partyâs base e.g. CORS/VRS/etc.). By one method or another, coordinates are pre- or post-determined for that base receiverâs location. The other receiver is the rover and it is maybe stationary over an âunknownâ point or maybe moving along an âunknownâ track.
In the most simplistic description, the processing software is told about the âknownâ coordinates where the base was located. Then it looks at the apparent position as calculated from the satellites. The apparent position is almost never âon the spotâ and is moving around in the general area. The difference between the apparent position and the known position is the error.
The error at that specific moment in time is then subtracted from the roverâs apparent position. Since the error is eliminated, now the rover position has become accurate. And that must be done separately for every epoch (every second at 1Hz or 5 times a second at 5Hz)
Behind the scenes, this correction mechanism within the processing software is:
reading the data from each satellite signal and coarsely aligning itself
monitoring the phase of the sinusoidal carrier wave of each satellite signal and fine-tuning the alignment
then when the data of all these signals meets a certain criteria, the position locks in and becomes âfixedâ
So, in a nutshell you need 2 Reach units to record GCPs. That way your data collection is self-contained within your equipment. Otherwise you need access to an NTRIP server that feeds you corrections live so that you can do RTK - or - you need access to historical data from a CORS so that you can do PPK.
p.s. to others reading - if I misspoke somewhere, feel free to correct or clarify, as I wrote this hurriedly.
If you are close enough to a CORS station, you could also use the log from that station as the âbaseâ log and only need to use 1 Reach device. (correct bide?)
