Mavic 2 Pro switched to GPS redundancy and eventually crashed

[TZ_Photo]

Yesterday I was about to shoot some evening photos of a house with my Mavic 2 Pro. the app prompted me to calibrate the compass which I did (a day earlier I was shooting 200 km from the location, it always requires me to calibrate the compass when I move to a location 100km or more). It was a calm evening, with no wind, normal temperature, and no nearby electricity grid, I launched the drone, it was hovering at a low altitude close to the home while I was checking the composition and moving things on the terrace. When I was ready to shoot I noticed that the screen switches from normal to some slightly different magnified view back and forth, I lost my image settings (no BKT) and a message appeared similar to "Swiched to GPS redundancy" twice. After that I decided to land but was unable to control the drone as it started to move erratically, finally, it hit a wall. One motor is damaged, but the big damage is my trust in the machine.
Can anyone help me explain what happened?
Here is the log file:

DJI Flight Log Viewer - PhantomHelp.com

 

[Meta4]

a message appeared similar to "Swiched to GPS redundancy" twice.

There was no problem with GPS, it worked properly for the whole flight.
At 3:43.3 there was a message to say that the drone had switched from the primary IMU to the backup IMU.
A second message a second later confirmed that the navigation system had switched.
This is not necessarily indicating a problem and the drone seemed to be functioning normally before and afterwards.

After that I decided to land but was unable to control the drone as it started to move erratically, finally, it hit a wall. One motor is damaged, but the big damage is my trust in the machine.
Can anyone help me explain what happened?

You mostly left the drone hovering until 4:32.5 but after that you made many confusing full stick moves with both joysticks to move the drone backwards, forwards and sideways at full stick.
These rapid, extreme moves make it hard to find a time when it was unable to be controlled and it started to move erratically.

You started to bring the drone down at 4:41.9.
As the drone descended there are still joystick inputs to move the drone backwards and to the left at full stick.
The data shows you put the joysticks in the CSC position from 4:49.6 and the collision happening at 4:50.9.

Were you consciously making the full stick joystick inputs for the last 20 seconds of the flight??

 

[TZ_Photo]

There was no problem with GPS, it worked properly for the whole flight.
At 3:43.3 there was a message to say that the drone had switched from the primary IMU to the backup IMU.
A second message a second later confirmed that the navigation system had switched.
This is not necessarily indicating a problem and the drone seemed to be functioning normally before and afterwards.


You mostly left the drone hovering until 4:32.5 but after that you made many confusing full stick moves with both joysticks to move the drone backwards, forwards and sideways at full stick.
These rapid, extreme moves make it hard to find a time when it was unable to be controlled and it started to move erratically.

You started to bring the drone down at 4:41.9.
As the drone descended there are still joystick inputs to move the drone backwards and to the left at full stick.
The data shows you put the joysticks in the CSC position from 4:49.6 and the collision happening at 4:50.9.

Were you consciously making the full stick joystick inputs for the last 20 seconds of the flight??

The drone was not responding to my joystick inputs, all the confusing full stick moves were my response to sudden drone movements not initiated by me, When I decided to CSC the drone it also did not respond and finally crashed onto a wall.

 

[Phantomrain.org]

When you have a Drone that is Requiring Calibration everytime you laucnh from a new area , its a good chance that the Compass has been Compriimised wtih magnetic intereference .

Magnetic Intereference could also explain the Drones Behaviors , im curious as to where you took off from the last couple of flight ?

Phantomrain.org
Gear to fly in the Rainl. Float on the Water.

 

[Meta4]

When you have a Drone that is Requiring Calibration everytime you laucnh from a new area , its a good chance that the Compass has been Compriimised wtih magnetic intereference .

He was flying the Mavic 2.
DJI (for some bizarre reason known only to themselves) made the Mavic 2 ask for a compass recalibration (even though there was no real reason for it), when it had moved more than X kilometres since the last flight or every 30 days.

Magnetic Intereference could also explain the Drones Behaviors , im curious as to where you took off from the last couple of flight ?

It doesn't sound like a magnetic interference problem.

 

[Cafguy]

Have you Calibrated The IMU also?

 

[EricJT]

He was flying the Mavic 2.
DJI (for some bizarre reason known only to themselves) made the Mavic 2 ask for a compass recalibration (even though there was no real reason for it), when it had moved more than X kilometres since the last flight or every 30 days.

It doesn't sound like a magnetic interference problem.

I have a Mavic 2 Pro and about a week ago, I took it north to the cottage which is over 200 km away....I was never asked for a compass recalibration when I flew there. So do you think the recalibration requirement sometimes doesn't appear, and if so, why?

 

[Meta4]

I have a Mavic 2 Pro and about a week ago, I took it north to the cottage which is over 200 km away....I was never asked for a compass recalibration when I flew there. So do you think the recalibration requirement sometimes doesn't appear, and if so, why?

There is no physical reason why transporting the drone any distance on earth would necessitate a compass calibrationbut DJI did force that with early firmware versions.
Perhaps they changed it in more recent firmware updates?
I haven't flown a Mavic 2 to check.

 

[Kuiper Denizen]

If the compass is using magnetic north for navigation it would seem one must recalibrate anytime there is a significant change in location. If the drone uses G.P.S. for determining “true north” then it should never need to be recalibrated, I would think.

The fact that the instructions warn about not just magnetic interference, but any metallic material close at hand, it would seem the former, magnetic north, is what the drone uses, not the latter.

Nonetheless, that doesn’t explain what happened. As a victim of the stress caused by proximity alarms and low battery warnings, coupled with a sudden return-to-home announcement (and maybe the mad-hornet-buzz of a close drone), I know it is easy to lose control of the situation, even if only for a few frantic seconds, and send the drone bouncing off trees.

There are many pilots here capable of reading posted data (I’m not one) who can interpret what commands were being sent, but I expect they can’t glean the mental state of the pilot at the moment when the totally unexpected happens.

 

[TZ_Photo]

I have two Mavic 2 Pros. The first one very rarely prompted me to do compass calibration, the second one does it as described earlier (I use it because the lens is sharper at the edges). Reading forums you may quickly conclude that you should calibrate IMU and compass before every flight as well as check the Kp index and the state of all drone parts. Every single flight. Similar to a car where before every trip you should check all your tires, engine oil levels and check if there's no cat under the car, otherwise you risk being called reckless.
I've been flying DJI drones since 2015, I still make mistakes sometimes but this particular day was a calm one, with no stress, no strange behavior of the machine, and then 10 seconds happened which were completely unexpected.
The guy I sent the drone for repair said that the sun Kp index might have been high at the time but it wasn't, I was standing close to a heat pump fan installation with e metal fence at the edge of the plot but I had no issues in much more complicated areas. I'm confused.

 

[Tolip]

You can ignore the Kp index for drone flying. Magnetic fluctuation due to geomagnetic storms is a nonissue for any drone compass.

To measure most geomagnetic activity takes a magnetometer (typically a proton magnetometer) that can resolve to 5 or 6 digits, or .001% to .0001%. For example, my local horizontal magnetic field is about 20,000 nT (nanoTesla). Typical geomagnetic storm variation is 10s to 100s of nT. This is the reason you cannot "see" a magnetic storm on a compass which can only resolve to 0.1% to more likely 1% or so.

You can see your local magnetic field data here: NCEI Geomagnetic Calculators

 

[Meta4]

If the compass is using magnetic north for navigation it would seem one must recalibrate anytime there is a significant change in location. If the drone uses G.P.S. for determining “true north” then it should never need to be recalibrated, I would think.

In normal situations you never need to recalibrate the compass of your drone.
The idea that you need to recalibrate after travelling some distance is 100% myth.
Compass calibration isn't about outside magnetic influences, it's only concerned with the magnetic fields that are part of the drone.
To get an understanding of what this is all about read the excellent first post in this thread.

A short explanation of compass function, calibration and errors.

BACKGROUND The aircraft compass is often misunderstood, both in terms of function and purpose, but it is important to understand what the compass measures, why, and what happens to that information. At the most basic level the purpose of the compass is to measure the earth’s magnetic field in...

mavicpilots.com

The fact that the instructions warn about not just magnetic interference, but any metallic material close at hand, it would seem the former, magnetic north, is what the drone uses, not the latter.

That DJI mention all metals, is just ignorance on the part of their writer or translator.
Aluminium, copper, brass, lead, etc have no effect on your drone's compass.
Iron and Steel and live electric cables are what matter.

Nonetheless, that doesn’t explain what happened.

That's because what happened had notjhing at all to do with any compass issue.

 

[Meta4]

Reading forums you may quickly conclude that you should calibrate IMU and compass before every flight as well as check the Kp index and the state of all drone parts. Every single flight.

Lots of myths get spread in forums.
The three you mention are completely false.
Don't unnecessarily recalibrate your compass or IMU.
Flying in a high Kp index has never been the cause of any drone flight incident.

Every single flight ... otherwise you risk being called reckless.

If you do every single flight, that's a superstitious ritual.
It's better to understand how your drone works than rely on superstitious rituals.
Read the first post in the link I posted in my previous post to understand what compass does (and doesn't) do and when it's actually needed.

I've been flying DJI drones since 2015, ... and then 10 seconds happened which were completely unexpected.
... I was standing close to a heat pump fan installation with e metal fence at the edge of the plot but I had no issues in much more complicated areas. I'm confused.

If you'd had someone analyse the flight data, you could have found out what actually caused your incident. reduce the confusion and help prevent a similar incident in future.

 

[Felix le Chat]

You can ignore the Kp index for drone flying. Magnetic fluctuation due to geomagnetic storms is a nonissue for any drone compass.

To measure most geomagnetic activity takes a magnetometer (typically a proton magnetometer) that can resolve to 5 or 6 digits, or .001% to .0001%. For example, my local horizontal magnetic field is about 20,000 nT (nanoTesla). Typical geomagnetic storm variation is 10s to 100s of nT. This is the reason you cannot "see" a magnetic storm on a compass which can only resolve to 0.1% to more likely 1% or so.

You can see your local magnetic field data here: NCEI Geomagnetic Calculators

Keeping a watch on the KP index before/during drone flight has nothing to do with potential magnetic disruption to the onboard compass - the only time solar phenomena would directly affect electronics is if there was a significant CME hitting the upper atmosphere, which would cause a large scale EMP. The KP level affects the number of satellites the drone can lock on to. The higher the KP: the greater the attenuation: the more satellites you lose.