Hello,

I've been reading lots of forums and posts out there about RTK, affordable equipment like Reach RTK and Piksi, but Piksi seems it requires lots of configuration and according to some posts lots of things could go wrong, and as for the Reach RTK, and I haven't found much, in fact I haven't found any videos of people using them, they both claim to integrate with Pixhawk, but has anyone test them?, similar? or cheaper?, and specially on Rovers?

http://store.swiftnav.com/s.nl/it.A/id.4216/.f

http://diydrones.com/profiles/blogs/affordable-new-rtk-gps-designed...

https://emlid.com/shop/reach-rtk-kit/

I'm sorry if this question was already answered but I'm researching before buying anything.

Thanks.

Views: 5694

Reply to This

Replies to This Discussion

1. Get Reach

2. No camera feedback port on drotek and more heavy than reach

3. PPK is the way for mapping. You don't need precise vehicule location but precise camera feedback

4. Same as QX1, no feedback until you hard hack it => https://community.emlid.com/t/hacking-the-sony-qx1-for-output-trigg...

Super interesting questions by @electrolics.

+1 here

Why do you need feedback from the camera? Is it not enough to know GPS + Compass + acc data when you trigger the camera?

Shutter lag is always present in any camera. If you use autofocus and other metering (shoot full manual to avoid this) it even adds more delay from the trigger pulse to the actual taken photo. Without feedback you won't be able to measure the "exact" moment. A whole other story is to calibrate everything up to sub cm precision as x y z offset from camera sensor to gps antenna and ofc accurate acc data to measure the exact heading...

too much for a "hobby grade" implementation IMHO.



electrolics inc said:

I was willing to buy the Reach RTK, and I am quite new to these RTK and PPKs. To bad these are not fully workable with pixhawk. Really interested to know more about it rather than spending a lot of time an money.

Questions on my mind:

1. Which cost effective solution to get then?

2. What about the Drotek one? Is it better then the both Reach and Piksi?

3. Any other PPK or RTK solution out there?

4. Not exactly relevant but is there a way to get feedback from Sony A5100? There is no hot shoe on it.

TIA

look into a M8P plattform like drotek

thank you for your reply @kiki boy

1. Please share your experience with Reach.

2. Thank you, so drotek is out.

3. Reach can be used in PPK mode, right? What accuracy (x,y & z) do you get with no GCPs and just the recorded feedback of the camera?

4. I had similar solution in mind to hard hack the A5100 and use shutter motor output. There is also one more way I am thinking to get feedback without tearing down the A5100. Can optical slave trigger be used to detect the flash and record the trigger event? I don't think it will add much more of the delay to the system. Please let me know your thoughts about this.

Are there any other RTK/PPK solutions out there which can record camera trigger and are plug and play?

The base station requires exact centimeter level location data which then translates to the rover. How do you get the precise GPS coordinates when there is none available before? Which hardware/model do you use? How long does it take to lock survey grade GPS location?

thanks again


kiki boy said:

1. Get Reach

2. No camera feedback port on drotek and more heavy than reach

3. PPK is the way for mapping. You don't need precise vehicule location but precise camera feedback

4. Same as QX1, no feedback until you hard hack it => https://community.emlid.com/t/hacking-the-sony-qx1-for-output-trigg...

Theoretically, you should be able to eavesdrop on the SD card starting to write. On the first "data clock" + "write" signal or so. It starts some millisecond after the picture is taken, but it should of course be possible to compensate for in the software (back log position the same amount of microseconds).

Handy cable from Ebay for 3USD will help you:

what is the range of wi-fi ?

can we connect it to internet for correction?

what happens when drone fly out of rtk wifi range? Does the rover gps is used as static gps?

Juan Diaz said:

I bought an EMLID Reach Kit. Here is my experience so far.

  • Integration with Pixhawk is good. It does require some configuration and understanding how the system and differential positioning works. It took me a lot of reading through the forum and around a week.
  • Antenna positioning and the use of correct ground planes is key! Test it many times until you find the set-up with best coverage.
  • After that, it seems that connection through telemetry radio is not very reliable (the corrections have dropouts, which does not permit to obtain a stable RTK FIX). I have experienced this, and read about a lot of people with the same issue. Nobody found a definitive answer to this.
  • Correction link through Wifi works great, I am using this approach currently (because the project in which I am working allows it). Since the drone does not fly very far away from the GCS, I can establish a wifi network.
  • In my current set-up, I am not able to obtain a stable fix every time. Most of the time it is lost after take-off and never recovered. I am working on it, might be an issue with the wifi network (too long distance between base, router and rover.
  • The Forum is fairly active, so doubts and problems can be solved by other users or Emlid workers.
  • The developers seem to be still improving the solution, because I have gone through at least 3 updates of the SW. So I guess there is still margin for improvement.

In conclusion, the technology looks great. When it works, its impressive (centimetre level accuracy). But there are many difficulties along the way. I still have not been able to solve all of them at the same time to have a reliable and stable solution every time. I am using it professionally, and still cannot rely on it. I am working on it though, I have hope!

I would love to hear the experience from other users of Reach or Piksi!


How to calibrate sub cm precision in camera and gps ?

Is there any procedure ?  
Tom said:

Shutter lag is always present in any camera. If you use autofocus and other metering (shoot full manual to avoid this) it even adds more delay from the trigger pulse to the actual taken photo. Without feedback you won't be able to measure the "exact" moment. A whole other story is to calibrate everything up to sub cm precision as x y z offset from camera sensor to gps antenna and ofc accurate acc data to measure the exact heading...

too much for a "hobby grade" implementation IMHO.

The range depends on many factors:

I have an Alfa Network Router, which has 500mW of Transmission power and an omni-directional antenna (supposedly 12dB although I guess it has much less, the stock antenna was 5dB). With this set-up I have been able to maintain stable FIX even with the base 90m away from the Router, and the Rover about 200m away from the router.

With a more powerful router and directive antennas you could have more range.

If RTK link is lost, the Rover starts working in SINGLE mode, and it is just like regular GPS with SBAS. In any case, I use Reach as a secondary GPS for safety.


Sriram said:

what is the range of wi-fi ?

can we connect it to internet for correction?

what happens when drone fly out of rtk wifi range? Does the rover gps is used as static gps?

Juan Diaz said:

I bought an EMLID Reach Kit. Here is my experience so far.

  • Integration with Pixhawk is good. It does require some configuration and understanding how the system and differential positioning works. It took me a lot of reading through the forum and around a week.
  • Antenna positioning and the use of correct ground planes is key! Test it many times until you find the set-up with best coverage.
  • After that, it seems that connection through telemetry radio is not very reliable (the corrections have dropouts, which does not permit to obtain a stable RTK FIX). I have experienced this, and read about a lot of people with the same issue. Nobody found a definitive answer to this.
  • Correction link through Wifi works great, I am using this approach currently (because the project in which I am working allows it). Since the drone does not fly very far away from the GCS, I can establish a wifi network.
  • In my current set-up, I am not able to obtain a stable fix every time. Most of the time it is lost after take-off and never recovered. I am working on it, might be an issue with the wifi network (too long distance between base, router and rover.
  • The Forum is fairly active, so doubts and problems can be solved by other users or Emlid workers.
  • The developers seem to be still improving the solution, because I have gone through at least 3 updates of the SW. So I guess there is still margin for improvement.

In conclusion, the technology looks great. When it works, its impressive (centimetre level accuracy). But there are many difficulties along the way. I still have not been able to solve all of them at the same time to have a reliable and stable solution every time. I am using it professionally, and still cannot rely on it. I am working on it though, I have hope!

I would love to hear the experience from other users of Reach or Piksi!

thank you Juan Diaz...... So we need to have a separate WiFi module and antenna to send corrections? I thought it uses Reach's internal wifi.... 

I tried a lot of expensive commercial soutions about  5 years ago with a helicopter and a hovering rocket.

None of the L1 only soutions were reliable and robust.

I eventually spent the $ to get a L1.L2 trimble setup and it was both stable and reliable. The performance was amazing.

I doubt ANY of the L1 only solutions willbe stable and robust. I look forward to the low cost dual frequency solutions... as I believe they will be a game changer.

Reply to Discussion

RSS

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service