Hi All,

Not sure if this has been address earlier. I am having an issue with the Terrain Follow feature with the P4 RTK. I went to fly a project the other day using Terrain Follow mode. I set my altitude to 70m. When the drone took off it shot up to 130m and flew the project at that altitude. Has anyone else had this issue with the drone flying much higher in terrain follow that what was planned?

Looks like Aerotas is also aware of this issue as they have noted it on their terrain follow tutorial page.

https://www.aerotas.com/phantom-4-rtk-terrain-awareness

It could be one of two things or a combination of both: 1) Incorrect elevations in the terrain model. It should be elliptical height. 2) You are flying on a non-standard day. The Phantom 4RTK doesn't seem to calibrate the barometer to the home point elevation, isntead it appears to use 29.92.  Please see this post for more information. https://forum.dji.com/forum.php? ... =192850&pid=1999862
My test data is 7 months old. I am hopng a subsequent firmware release has fixed the problem, but I haven't tested the latest release.

Skyveyor112 Posted at 6-9 17:27
It could be one of two things or a combination of both: 1) Incorrect elevations in the terrain model. It should be elliptical height. 2) You are flying on a non-standard day. The Phantom 4RTK doesn't seem to calibrate the barometer to the home point elevation, isntead it appears to use 29.92.  Please see this post for more information. https://forum.dji.com/forum.php? ... =192850&pid=1999862
My test data is 7 months old. I am hopng a subsequent firmware release has fixed the problem, but I haven't tested the latest release.

Sounds like you are correct. This is the response I received from DJI support.

"We are grateful to assist you with your concern. Regarding this issue, if the RTK signal is not FIX, the drone will preferentially use the barometer and GPS to set its altitude. However, the value of the barometer is very easy to be affected by the external environment such as temperature or wind.
In general, there are many factors that can affect the barometer.
For GPS, the most important thing is the number of stars. The more stars there are, the more accurate the value is, and vice versa. Environmental interference will also influence GPS.

Please ensure your RTK signal is FIX when you execute this mission. Or you will see the height in GS RTK isn't accurate as what you set up in App, please execute the flight mission when RTK signal is fixed. Then your aircraft will use RTK to help its position, which is more reliable than GPS and barometer. "

What I don't understand is when I fly this drone I always fly it with my own base station not DJI's, so I am flying it without RTK and then post process using the rinex file on the drone. The results come out great. I notice that when the drone takes off and comes back to land it is literally within a couple of centimeters from where it took off. I am assuming that it is using the survey grade gps on board for navigation to achieve this. If this is the case, why can it not use the same survey grade gps to determine an accurate height for takeoff when using terrain follow instead of using the barometer? I have another drone that has a pixhawk 2 and uses mission planner ardu-pilot with terrain follow with no issues. Why does DJI's have to make this more complicated than it has to be by using the barometer?

CCAM Posted at 6-11 07:44
What I don't understand is when I fly this drone I always fly it with my own base station not DJI's, so I am flying it without RTK and then post process using the rinex file on the drone. The results come out great. I notice that when the drone takes off and comes back to land it is literally within a couple of centimeters from where it took off. I am assuming that it is using the survey grade gps on board for navigation to achieve this. If this is the case, why can it not use the same survey grade gps to determine an accurate height for takeoff when using terrain follow instead of using the barometer? I have another drone that has a pixhawk 2 and uses mission planner ardu-pilot with terrain follow with no issues. Why does DJI's have to make this more complicated than it has to be by using the barometer?

The landing precision is also accomplished optically once the drone is close to its home point.


Your last question made me chuckle.  Many a DJI user (especially of "Pro" and "Enterprise" products) has uttered the same thing (not specifically about the barometer, but rather a host of issues). The DJI solution is not always the most industry standard or elegant one, and even within DJI things are not always consistent across platforms. It's just the way it is.

CCAM Posted at 6-11 07:44
What I don't understand is when I fly this drone I always fly it with my own base station not DJI's, so I am flying it without RTK and then post process using the rinex file on the drone. The results come out great. I notice that when the drone takes off and comes back to land it is literally within a couple of centimeters from where it took off. I am assuming that it is using the survey grade gps on board for navigation to achieve this. If this is the case, why can it not use the same survey grade gps to determine an accurate height for takeoff when using terrain follow instead of using the barometer? I have another drone that has a pixhawk 2 and uses mission planner ardu-pilot with terrain follow with no issues. Why does DJI's have to make this more complicated than it has to be by using the barometer?

the drone does not use GPS to navigate the way people think. The gps position is only accurate to about 1-3 meters. When in RTK then it become precise to 1-3 cm. But otherwise a single gps is always floating around. The drone uses its IMU (inertial M?? Unit) it uses magnetic compass for direction and barometer for height. And knows its tilt etc. Like your smart phone can tell when you tilt. It also uses lidar (very weak signals). And uses cameras, to literally measure its speed. All that and it can return to almost the same spot.

terrain following does work. But it is not perfect. it is all assuming some stuff. And one problem that can happen is the drone robot might think it is going over 400ft above ground and throw an error. I had this happen. Anyhow, make sure the terrain model is ellipsoid height and not ortho elevation. And that it is sampled at a suitable amount. Like 50meter areas or 25meter. Or even 100meter. Depending on what you are doing. you have to give it a terrain model to use. I make them in Global Mapper myself. Terrain following is essential to make some mapping flights easier to accomplish. but overall it is not a trivial thing to do.