Waypoint Mission not always repeatable.

[KBF]

I've recently done a couple of waypoint missions (Mavic 3 pro) where I was building a structure and wanted to do a daily fly around to document the builds. Both times I did 2-3, maybe 4 flights then the whole pattern would shift around 20ft north. It would still fly the correct mission ... just in the wrong place! Any ideas? (I've ruled out magma shifts, voodoo and butterfly farts ..... well, not too sure about the magma)

 

[DJ Wes]

I've recently done a couple of waypoint missions (Mavic 3 pro) where I was building a structure and wanted to do a daily fly around to document the builds. Both times I did 2-3, maybe 4 flights then the whole pattern would shift around 20ft north. It would still fly the correct mission ... just in the wrong place! Any ideas? (I've ruled out magma shifts, voodoo and butterfly farts ..... well, not too sure about the magma)

No two waypoint mission flights are going to fly the exact same path due to GPS inaccuracies. However, 20 feet would be at the extreme end of inaccuracy.

I have created a number of videos that were created from a blend of multiple waypoint flights flown over the course of construction. In order to get them to match, each segments needs to be scaled up between 10 to 15% and then shifted in either the X or Y directions to get things to align properly.

While a GPS location can be specified with great precision, the accuracy of flying to that GPS location is something less than perfect.Simply stating the accuracy for GPS can be difficult because there are many different types of GPS and many things that affect the accuracy. This is probably why one finds so many different values quoted on-line.

Smartphones tend to use assisted GPS (A-GPS) which combines traditional GPS signals with cellular network data to increase the location accuracy.Therefore, smartphone position accuracy is generally better than that of devices using standard consumer GPS.

Like many consumer devices, DJI drones use what is called "Consumer GPS". Depending on your source, the accuracy of consumer GPS implementations is anywhere between ± 3 to 8 meters (10 - 26 feet). However, DJI states the horizontal positioning accuracy of their older drones to be around ±1.5 m (Mavic 2, Air 2s) and their newer drones to be ± 0.5 m (Mini 3,Mini 4, Mavic 3). I have flown literally 100s of waypoint missions and DJI's stated accuracy corresponds well to what I have experienced when flying and repeating my waypoint missions. I have never experienced a waypoint mission to be anywhere near 8 meters (26 feet) off of the drone's intended course.

Web sites with good explanations of GPS

• How Accurate is GPS?
• An in-depth exploration into how accurate is GPS

 

[KBF]

No two waypoint mission flights are going to fly the exact same path due to GPS inaccuracies. However, 20 feet would be at the extreme end of inaccuracy.

Odd thing is, once it shifted position on day 3 or when ever it was, it tried to fly that new route each day and each day I cancelled it and did it by hand. The second structure I recorded at a later date, I deleted it and made a new mission.
They were quite tight loops around a building so I expect that amplified the shift as opposed to being several hundred feet out and flying a loop.

 

[HoozierDroneDaddy]

Just a thought,, but since you mentioned tight loops around the buildings, are your missions very close to the building that possible there is something in the building that is causing an interference effecting your mission?

 

[Dave Maine]

GPS location will never be perfect. The spec for a well functioning civilian system is +- 16 ft, 95% of the time. The satellites are not in geosynchronous orbit, so the drone swaps in new ones and swaps out the old ones as needed. In addition, the satellites and the drone net to have clear, unobstructed line of sight connection between them. Trees, branches, buildings etc all block the GPS signals as well.

What was the GPS Indicator on the controller telling you? How many satellites? Any other problems showing?

 

[KBF]

Just a thought,, but since you mentioned tight loops around the buildings, are your missions very close to the building that possible there is something in the building that is causing an interference effecting your mission?

It was only 12’ high ag building so I was above it the whole time but good idea

 

[KBF]

GPS location will never be perfect. The spec for a well functioning civilian system is +- 16 ft, 95% of the time. The satellites are not in geosynchronous orbit, so the drone swaps in new ones and swaps out the old ones as needed. In addition, the satellites and the drone net to have clear, unobstructed line of sight connection between them. Trees, branches, buildings etc all block the GPS signals as well.

What was the GPS Indicator on the controller telling you? How many satellites? Any other problems showing?

I don’t recall how many satellites but I always have good numbers and I was in a field away from obstructions.

 

[Droning on and on...]

The configuration of GPS satellites in the sky can affect accuracy. Most of the time there is good coverage, but occasionally you can just have bad luck and just at that time have a large HDOP.

I use GPS Status and Toolbox on Android to examine the sat configuration and calculated PDOP/HDOP/VDOP when it's important, like for waypoint missions.

Here's a screengrab... the little green dots, squares, and x's are the satellite locations in the sky, the middle being directly overhead, the circle being the horizon: