How to PPK 50-100 Survey Points using 2 Reach RS Units with no RTK correction?

Reach RS / RS+ Getting started
1

I found an earlier closed post on this but I was still unclear on the proper solution after reading it. I need help with the following:

  1. No internet and therefore no RTK corrections.
  2. Using 2 Reach RS Units. One will be the “base” and used for reference
  3. Second Reach RS will be “rover” and used to collect Survey Points.

I can follow the guide and Post Process the Rinex file from the Rover. But what is important is “Post Processing” the survey points. Let’s say I go out and survey 50-100 points using the above scenario. Using a pair of Reach RS Units, how do you batch Process all the Survey points so that they are now <5cm accurate? (or as accurate as possible)

So whether you collect 100 points using a Reach RS “rover” or 100 points using a Reach Module on a UAV, I’m trying to understand how you will process all of those “survey points” to be accurate?

Thanks for any help!

2 Likes
2

Tim, If you are going to post process for your corrections, the CORS logs will be the base, not one of your Reach modules. You would use the files created by your Reach rover and plug them into RTKLIB with the correction files from the CORS station to come up with your corrected points.
Note that this is my understanding but I have not received my Reach kit yet so I have not done this myself.

Maybe this will help you use RTKLIB even though it is not using the Reach. The process would be the same though I think. Maybe one of the guys here or Emlid will put together a little tutorial using the Reach.

3

I have 2 Reach RS units and one Reach (RTK Kit) that will go on my Phantom 4. To do PPK you need a base unit. It does not have to be a CORS base and for single frequency receiver like the Reach, I am pretty certain that it needs to be within 10-12km of your Roving Unit. So if you don’t have a CORS unit close by then you really need to use your own base, i.e. Reach RS Base (or even a Reach Module) for the best accuracy.

I am beginning to wrap my brain around processing with RTKlip but I am now wondering how to post process “survey points” that are taken using the Emlid Reach. When you go to export your Survey, the only options are GeoJSON, ESRI_Shapefile, or DXF. So in my above example, if you surveyed 50 points, the coordinates for those points are not processed for accuracy. They are precise because the location assigned to each Survey point is relative to the base unit. However, the base has error that has not yet been processed.

So once you process the Base, there has to be a way to apply that processing to the batch of 50 survey points (or in the case of a UAV, it could be 500 points (one for every photo). There has to be a way to do this.

When traditional land surveyors go out and survey a bunch of points, they have to use some type of software to batch process their Survey Points for pinpoint accuracy. The log file and your survey are two different things. I am just trying to figure out how to process Survey points taken with ReachView or an app like Mobile Topography Pro. I sent a message to TB who has been very helpful in teaching me this stuff. If he sends me a PM with good info, then I will update the information here.

4

In your scenario, you would need to correct for the base and then correct the rover points. Why not just directly correct the rover points. You say there is no CORS station within a reasonable distance from which to get corrections for your rover. That means that there is also no CORS station within reasonable distance to correct for your base either. In that case, you will need to set your base up on a known location to begin with, correct?

After this comment, I’ll hang back and let folks that actually know what they are talking about respond because I am just learning too and my data may need correcting. :wink:

5

I would think that one option is to setup your base and let it observe for 15-30 minutes. You could then do PPK to get a an accurate location of the base. You would then manually enter the corrected coordinates in the base unit and now you have RTK on your survey results because you know your base coordinates to be true and accurate from the PPK processing.

But I would still think there is a way to batch process the survey points back in the office. But not 100% sure on that.

6

I did not think of that :blush: I will do like you and see if we can get some guidance on this as I certainly don’t know and I will refrain from speculating.

I am mainly thinking of not having RTK due to two reason:
A) No internet connectivity
B) I am too far from the one NTRIP base that I have access to.

But in north Mississippi where I am, CORS units are in fact about 50 miles apart when I look them up on ArcGIS Web Application . I am still learning a little every day. BTW Dave, thanks for the input and feedback!

7

If you have a log of when you occupied each point then you could post process the file with the correct base location and then extract the positions for each point from the text .pos file. If you were static at each spot for 30 seconds (say) then the fastest and dirtiest method is to load the whole .pos file into excel and just average the positions for each occupation.

If you are talking UAV then unless you have a time synchronized trigger event to tie the two files together than you are a little lost. If you used MapPilot or Pix4D to pilot the drone then you have a (somewhat) synchronised set of trigger events on your ipad/android tablet if you used a more opensource controller then you may have a log file with events in it, what is the timing like in that file?

Exif times are not good as they are to the integer second and may be associated with the file save date not the shutter.

8

Why don’t you use the LoRa radio built in to the Reach RS for your RTK correction link. No need for the internet. You may have to make two trips to site to collect enough data to set a known point for your base.

Then you can use the survey app in reach view to log your corrected points.

I have thought of using the time mark pin on the basic reach to mark survey points in the log file for post processing.

The only problem would be it will only log one epoch but this could be worked around by taking multiple observations.

9

Thanks Peter. What I am trying to find out is whether I could setup the Base unit (without any corrections to the base via Ntrip). The Base would be for reference to the Rover and then do the survey with the Rover. Then go back and post process the Survey data with data from a Cors Station. But it seems apparent that this can not be done.

So one option is to do like you mentioned and go a day or so ahead, take observations with the base and then process that for accuracy. Then do your survey on the Rover using the corrected coordinates manually entered into the Base unit.

A second option is to setup your base when you are ready to survey, take observations for a period long enough to Process and determine your corrected and accurate location right then. If you take a laptop with Internet connection then it should not be a problem.

I will begin testing all of this tomorrow and hopefully get it all figured out. I welcome any further advice.

10

Hi, Tim
This is possible , but you need access to downloadable (is that a word?)prehistoric observation data which is an option many has. Possibly your cors.
You log rover, base and download historic .obs data and postprocess this in almost the same maner as descibed in the docs. Its possible somebody has posted a quick walkthrough about it here on the forum.

11

I am also trying to do the same thing. I have used Sokkia GNSS system in PPK mode and their software provides corrected survey points after post-processing.
Also, does anyone knows how to proceed from pos file to extract points in dxf file for CAD purpose?

12

Does your cad handle LLH format?
Right now, only option out from reach is LLH or XYZ (not utm) when logging (not survey mode). There are ways around this, but it might be a bit jumping around to get there.
Edit: using export to gpx with RTKpost is an option. This will bring. POS dato to your cad. But not sure that’s what’s your are looking for?

13

Sorry, I’m coming in a little late here.

  1. The built-in survey project feature in ReachView is geared to RTK for the moment and not PPK (or PPS).

  2. You say you want to post-process for the most accurate result (which is correct). For this method (PPS), forget the ReachView survey project feature. The only thing you need to concern yourself with is the raw log.

  3. There is (currently) no automated way to process your 100 survey points automatically, but you could write a script or batch file to do it for you. And then you could reuse it on your next projects.

  4. I can think of three approaches to handling the raw log for post-processing:

    i. At each survey point, start a new raw log and stop it before you leave. In this method you will have 100 log files to post-process which will result in 100 individual position files.
    ii. At the beginning of your survey, start the rover log, and at each survey point, write down a name and the arrival time (and optionally the departure time as well). You can play back your log with RTKPLOT like a video to find the point and you can also limit the current view to the arrival and departure time and get average position from the statistics on screen.
    iii. Do the same as in ii, but obtain the Reach RS cable and wire a switch to the Event pin wire of that cable so you can mark your points with the press of a button in the field. Say that you sat at each point for 5 minutes, then you could have your script post-process from the Event time then plus five minutes. The upside is it is more suited for automation, but the downside is there isn’t a way to give the points a name (but maybe it doesn’t matter).

Certainly there should be a better way, but this is what we have to work with right now with ReachView and RTKLIB. To others: feel free to say if there is something I missed or to describe third-party software that you use which makes this job easier. (Remember: we are talking about automating the static post-processing of the results - PPS, not RTK)

1 Like
14

If you create a hardware button on the time mark pin that brings it to ground then you can create Photo events which are pulled out in post processing.

image.png1344×732 49.9 KB

I haven’t tried it with the Reach RS,but the spec sheet sems to indicate this could be implemented over pin 4 of the connector. COuld Emlid confirm this?
@egor.fedorov
Mind you the spec sheet also indicates PPS on pin 5.

https://docs.emlid.com/reachrs/files/without-connector.pdf

1 Like
15

We have been down that road before :blush:

1 Like
16

I want to first and foremost thank the people on this forum who have really helped me during my learning process. I can’t say enough to TB and there are others as well (BIDE and Simon).

Below is what I believe to be the average journey of a newcomer… It is lengthly so if you only want the bottom line go to the end. If you want to know what it is like for newcomers to Drone Mapping and RTK in general, then start here.

There are 2 very distinct groups of users who are purchasing the Reach RS and Reach Modules (maybe a few in the middle). First there is the group of guys who really know this stuff inside and out (ie Bide, Simon, and TB) and I’m sure many more. Then there is everyone else. The “Everyone Else” group consist of many people who have drones and they hear about Drone mapping and think “I can do that!”. They watch videos and see various forums talking about RTK and a few clicks later they find out they can purchase a high quality receiver at a very good cost (ie the Reach RS or the Reach Module). During the next phase they receive their newly purchased Reach unit and start watching YT videos and trying to find out how to even set this thing up. They find great videos on how precise the Reach units can be (which they can because they are awesome!!). Then they even find videos and webpages showing charts with green dots that they have no idea what it means, other than people saying this is great!!.

If they are in the USA they may have gotten lucky and ended up on this YT video channel to see how to set up their pair of units (for those that bought a pair).
https://www.youtube.com/channel/UCKKI6ezzKMOEtlG6vQpIMcg/videos?disable_polymer=1

And at this point they still have no idea what they are doing, they are mearlly doing what the guy shows on the video and they figure they will understand it sooner or later.

Of course there are the folks who only bought one Reach module and thought is going to be sweet. “$275 and I am in the Drone Mapping business!!” They go to test their new toy. They have seen where they want to see the word FIX. They only see Single so back to the drawing board to find out what is wrong. After some more reading they realize that in order to see FIXED you have to have another unit. What???

For those who have 2 units and followed the above video, they see the wonderful word “FIXED” and think they are ready to go. Little do they know that only means they just graduated from the 6th grade. They still have 6 more years of high school, and then college to go. So they have a fix and now it time to figure what to do next.

For the folks who bought one Reach unit and frustrated they don’t see FIXED, they read some stuff on NTRIP servers and NTRIP casters and that this will eliminate the need for a second base. And they even read in multiple places it is FREE. They don’t know that is not entirely true and still think, great I am only in this for $275.

The folks who did buy 2 Reach Units have also become intrigued with NTRIP and are trying to find out what it is and how to use it. So they finally understand what NTRIP is and how it provides correction to their Reach Base or even the Reach Rover. But they also find out that it cost $thousand$ of dollars per year for a subscription. Yikes! But maybe they get lucky and find where they get free access to ONE base. And that is one base that does not change. But hey, it is free access and the testing goes on. Now they throw up a single Reach unit and they see the word FIX. “Hallelujah I am ready to Map!!”… reality is coming.

They read about GCP’s and having RTK on their drone, and they think “No problem! My screen said FIX so I am good to go”…

The majority of the group, “Everyone Else” , is probably flying a P3 or P4. Certainly there are some other UAVs in the mix but the likely majority are DJI Phantoms. So now they are trying to figure out how do I get this RTK thing to even work? As a matter of fact, most have not yet figured out there is a very distinct difference between RTK, PPK, and PPP. Some think it is all just one big happy family, but they carry on.

So in short order, they find out that RTK on their Phantom is a problem. Now they are starting to scratch their head realize this was NOT just something where you buy a $275 quality GPS receiver (which the Reach module is) and you are in the mapping business. They are now eagerly trying to find more videos and tutorials because they are starting to realize the depth of this project and the steepness of the learning curve.

Some have now decided to bow out of the project and realize they would rather just go fly and take drone video.

But for those left in the Drone Mapping and RTK race, after alot of reading and watching more videos, they now actually understand that there is a distinct difference between RTK, PPK, and PPP. They don’t necessarily understand how it is all done, they just know there is a difference. They also realize that they have a UAV that can’t just simply do RTK precision. It is going to take some modifications. They may even find Simon in this forum and order his solution which may end up being pretty darn good solution.

Then they decide to give the RTK stuff a break, because it is making their head hurt. So they turn their attention to the Drone mapping software. They do what alot of people do, Google! “What is the best drone mapping software?”. And now they are off to the races to learn about mapping software. It does not take long to retrieve a list of the top contenders. At each website the first button most press is PRICE, BUY NOW. While some of the prices are pretty high they see the word FREE! Even though they did not read too close on the difference between free and paid, they just know there is a free option. They also find out that is you don’t have RTK on your drone you can use GCP’s for good precision. And they think great, I will can just survey my GCP’s .

For those who have stayed in the RTK race, they begin to find out that there are a lot of uses for drone mapping, including 3D modeling, Agriculture, Construction, and more! They did not know much about some of the applications so the excitement grows as they think about the potential $$$

Now I could keep going with all real phases that many newcomers go thought. But I am going to fast forward a couple of months later…They now realize a few things:

  1. It will cost you a decent bit to get into drone mapping and do it right.
  2. Using the Reach Units and ReachView app is only for RTK solution unless you use additional software and collection means (ie not using ReachView).
  3. If you fly a Phantom P3 or P4, then your problems of RTK are even greater because you have to be able to take the precise survey point collected by the Reach Module and then update your Exif data on the pictures so that each photo contains the precise and accurate point over earth in which it was taken. There is no magic button to make this happen.
  4. It is near impossible for any newcomer to learn RTKLib for processing. There is no teaching material that teaches at a ENTRY level most newcomers need in order to understand it and learn it. So don’t waist your time looking!
  5. (For 99% of users) you can NOT post process a Survey taken in ReachView. So in the real world where you would go and survey points, you can’t post process them. That was hard for me to understand but that is a simple fact.

So here I am months later and I do have a clear understanding of the realities above. I have decided that I am going to use an outside software solution. As Bide has stated, that is really the only option at the moment. And by only option I mean one that fits a real life day to day work flow. I have high expectations that it will allow me to carry one with my investment to do quality drone mapping. I will update this post as I progress and learn more.

So in closing, I will say that the Reach Products are positively a high quality product. They can accomplish the same outcome as other equipment costing thousands more. However, owning high quality equipment at a fraction of the cost is but a tiny piece of the drone mapping pie. In general, there is a huge void of entry level information between buying the Emlid Reach units and the final high quality Drone Mapping output. This will will prevent many people from entering the drone mapping market. But in a way this can be good because those who have resolve to stick it out and learn, will have far less competition than if the information was readily available.

8 Likes
17

Also don’t forget, many states require licensing to be a Surveyor and put your stamp on things. For me in Ohio it’s a four year degree and a few years in the field to even qualify for the exam.

My only option now is working FOR a surveyor and not BE a surveyor. Frankly I’m not going back to school for another degree.

2 Likes
18

@MDS and your journey continues, even once you have the drone mapping software you can make really bad, good looking maps.
Camera calibrations, accuracy assessment and implementations, GCP tie ins, verification points, are all part of the mix.The problem with drone mapping software is that it is really easy to use, but not easy to master.
Photomodeller UAS tried to address this but made the workflow a little too complicated.

1 Like
19

@MDS I really enjoyed that rant. :slight_smile:

2 Likes
20

I greatly sympathize. My years as a journeyman plumber didn’t do much to prepare me for this field. :slight_smile: I started as an absolute beginner as well and still feel like it many aspects. I am just starting to take on the post processing field seriously and I can relate to your feelings. I have made a few tutorials that are targeting absolute beginners for some other procedures, but in my mind there is room for many more. It’s so easy to get lost in the many posts each covering part of the procedure, all full of words that are strange and unfamiliar and each person has a slightly different approach.
However, I don’t see a great alternative to this and I have embraced the learning curve and have enjoyed it (when things are working)
What would be a good way for us users to help with documentation? Maybe we need a list of basic procedures you’d like covered in a tutorial.? Just five or ten minute highly illustrated and explained examples? Just an idea.
I have been working on something a little similar to what you were talking about, but not quite the same. I surveyed a piece of land and had my base station set up on site. Had a great fix the whole time. I logged a control point and then at home post processed that control point against CORS data. I then shifted all my points over together with my control point to get absolute accuracy. This was a rather painful process that had far to much guesswork for my liking. I’d like to see each step explained for guys like you and me that don’t know which coordinate system to choose or how to export a point out of a whole file of .pos data. etc. etc. I think next time it would only take me half a day instead of a day and a half, though :slight_smile:
Keep your chin up! You’ve come a long ways! I remember googling words like “rtk”, “gis”, nmea, all that stuff ,too .:slight_smile:
Please read the last sentence of this post. :slight_smile:

2 Likes