RTK GNSS experience

Dear all,

I try to do some outdoor experiments and would like to have good altitude/position tracking, even in dynamic flights.
About 2016 some first tests using the RTK GNSS technologies were done. As I searched the internet I did not found recent reviews which show the performance of the RTK.
Does anyone of you did gather some experience on this technology and how it incorporates with px4?
Especially, how is the performance static(are 2.5cm or so really achieved?) or dynamic?
Which setup do you use? Drotek, Here+ or another, how is the hardware(qualitiy, customer service, availability etc.)?
Is it really worth to invest 600€?

Thank you for your help and a nice day.
Chraebi

1 Like

I have done experiments with Real Time Kinematics in 2018. But I haven’t used any of the commercial products instead I created the PCB my self and used Neo M8P-0 and Neo M8P-2 modules. With active antennas had good GPS signal strength. the experiment was successful. PX4 integrations are very solid and actually if we have clear visibility of the Sky, we can archive the 2.5cm accuracy in RTK fixed mode. In RTK float mode it will within 10cm range. The location of the drone is relatively to the base station. So fixing the base station accurately is mandatory if absolute location is important.

Thank you very much for your help!
Did you also observe the altitude? As you said, static 2.5cm (I think horizontal and vertical) can be achieved. How about the altitude tracking during filghts?
Might you explain short what experiment you did?

For my part I would like to compare different algorithms and see how e.g the altitude error varies.

Kind regards
Chraebi

TLDR: if you need absolute accuracy, then yes RTK is absolutely worth the price tag.

I don’t have anything to add to @Aruna 's response, but I can report I have been seeing similar results. If you need absolute accuracy and repeatability, RTK is pretty much your only option.

RTK will provide you with a significant increase in altitude tracking performance and remove the drift-y behavior observed when only using the barometers for absolute altitude tracking. I’d say that I’ve effectively observed the absolute altitude estimation within ± 20 cm. You just need to make sure your parameters are tuned appropriately (found here).

Actually, I was involved in a drone delivery project. I was using a Lidar sensor for the altitude estimation. But during the testing, it was observed that RTK can fix the altitude also up to some level. The 2.5cm level accuracy was observed in longitude and latitude readings. QGC can automatically detect the base station and once Base is fixed, it will automatically send the corrections to drone. But as mentioned by @tonysam1 we have to configure flight controller parameters to utilize the RTK. I was trying to implement the wireless link with GPRS but it was failed due to network latency, Then I have used Long range radio for communication.
According to my experience, If your project requires a higher accuracy only in altitude, It is better using another option rather than RTK when compared with the price.

Thank you guys for your help. I also thought about using a Lidar sensor for alitutde estimation, but I have some concerns because the EKF will be only using those measurements until 2m/s.This is defined in EKF2_RNG_A_VMAX.
@Aruna Did you use the Lidar measurement mainly for take off and landig or also during the flight?

Kind regards and thanks a lot!

I only used it for landing and as the height estimator for the optical flow sensor. Lidar was accurate only up to 12m of height. But there are Lidars that can handle more than that.