Launching from boat - Mini 4 Pro went straight into the sea

[Mrkva]

Hi,
can someone help me figure out what went wrong?
I was on anchored boat (that was swinging a bit), in ~10 knots of wind. After GPS fix, I started propellers and increased throttle. The drone lifted up, however immediately started flying to away, to the side and descending (even though I was giving it full throttle and in the beginning no directional inputs at all).
Then it crashed in a couple seconds and went to the bottom of the sea.

DJI Flight Log Viewer - PhantomHelp.com

 

[akdrone]

was the engine off? Could there have been some electo-magnetic interference? that's all I can think of... Either that or somebody has a voodoo doll cranked up?

 

[mavic3usa]

Which one?

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[mavic3usa]

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[Meta4]

Hi,
can someone help me figure out what went wrong?
I was on anchored boat (that was swinging a bit), in ~10 knots of wind. After GPS fix, I started propellers and increased throttle. The drone lifted up, however immediately started flying to away, to the side and descending (even though I was giving it full throttle and in the beginning no directional inputs at all).
Then it crashed in a couple seconds and went to the bottom of the sea.

The height data in your flight record is difficult to interpret.
The IMU height data (the height you'd see on your screen) is not normal.
When you power the drone on, it starts from zero, but yours starts at 17 metres.
As you launch and climb (as indicated by the VPS sensor), the IMU height indicates the drone is descending.

Although you had full left stuck to climb for most of the time the drone was in the air (from 5.8-14.4 sec), the IMU height data only shows the drone climbing for 1.7 seconds at 6.1 sec and 2.3 seconds at 9.1 sec.
For the rest of the time IMU height indicates the drone was descending.

The VPS data appears to be unreliable too as it shows rapid fluctuations much greater and faster than I'd expect would be caused by the rise and fall of the water level.
Six times in the 14 second "flight", it's showing a drop of 1.8-2.4 metres in 1/10th of a second.

The data doesn't show the motors stopping before signal was lost.
It appears that the drone gradually descended rather than falling and this happened despite full throttle up.

How high did the drone rise?

You mention the drone flying away to the side.
The data indicates the drone ended up approx 12 metres from the launch point and the speed had increased gradually to 2.5 metres/sec .
Does this sound about right?
What can you say about the water surface conditions?
Was it flat calm, 2 metre waves each 0.6 seconds (as the data suggests) or something else?

 

[Mrkva]

I would say it went ~ 1 meter, above me, not more. It not climbing as it is supposed to was the first indication something is wrong. Yes, it flew sideways ~ 10 meters with increasing speed and hit the water with motors still running and no error displayed on the RC.
Sea was calm (maybe 5cm "waves").

 

[Fort Pierce Pilot]

I've been considering a Mini 4 Pro but...that's a problem :-(

 

[Starz]

I've been considering a Mini 4 Pro but...that's a problem :-(

I haven’t seen evidence that the issue encountered by the OP is an across the board problem with all Mini 4 Pro’s.
Have I missed something ??

 

[Mrkva]

I have flown hours from boats, sometimes even in motion without any problem until now. It could be defective unit, it could be some unexpected mistake on my side... I wouldn't rule out the purchase based on one data point.
Let's see what DJI does in my case.

 

[BossBob]

More likely your luck ran out! From the moment the motors are started the logic for the drone is to stay at that three dimensional position. If the boat was swinging at anchor I’ll bet the drone moved away from the direction of swing, suddenly found there was no deck below it so tilted just as you tried to introduce what was insufficient lift and it appeared to fly into the water.

 

[Mrkva]

@BossBob: somehow I don't think that's it. I've experienced drone ignoring full throttle just once - when it was doing emergency landing due to low battery. But it never drifted sideways while descending, unless commanded to. There were no errors displayed, it even responded to rudder input. And whenever I was launching from moving boat, the drone always stopped exactly where I launched it and never moved an inch. Obstacle detection was on "Brake", not "Bypass", so that's not it either.

 

[Starz]

More likely your luck ran out

From the DJI products I have had apart, I don’t remember seeing a container of luck that can diminish with use. I have seen interior components that can and have resulted in failure. The members that decipher flight logs tend to find solid answers.
At times, the logs themselves can be misleading and make determinations difficult.

 

[Bad Santa]

Stories like this make me afraid to fly over water at all.

 

[BossBob]

Stories like this make me afraid to fly over water at all.

Flying over water isn’t a problem. Launching or landing from a boat is the problem, especially if you don’t understand the logic. Even the simple Trello, with no GPS and only rudimentary height sensors, knows when it is dropped if the motors are running and will throttle up itself to maintain its altitude. Add all the sensors on a much more modern drone and put them on an unstable surface all bets are off.

 

[EricJT]

Stories like this make me afraid to fly over water at all.

I used to feel that way, but came to the conclusion that if I flew at much above 30 feet, a crash would be as destructive as losing my M2P in the water....so now I routinely fly over lakes where I am.

 

[mneeze]

I used to feel that way, but came to the conclusion that if I flew at much above 30 feet, a crash would be as destructive as losing my M2P in the water....so now I routinely fly over lakes where I am.

While true, if the drone needs to make an immediate "emergency landing" over water, the result will be much different.

 

[Meta4]

From the moment the motors are started the logic for the drone is to stay at that three dimensional position. If the boat was swinging at anchor I’ll bet the drone moved away from the direction of swing, suddenly found there was no deck below it so tilted just as you tried to introduce what was insufficient lift and it appeared to fly into the water.

But the drone in this incident did not hold position.
It actively flew 44 feet to the north of the point where it launched, without any corresponding joystick input.
One possible explanation for that could be a mild yaw error, but because the OP was using the rudder to rotate the drone, I was unable to find anything in the data to confirm this.
To add to the problems the OP had, the drone failed to climb more than a metre despite the throttle being fully in the up position for the duration of the flight.

Flying over water isn’t a problem. Launching or landing from a boat is the problem, especially if you don’t understand the logic. Even the simple Trello, with no GPS and only rudimentary height sensors, knows when it is dropped if the motors are running and will throttle up itself to maintain its altitude. Add all the sensors on a much more modern drone and put them on an unstable surface all bets are off.

But this drone had a big problem with its height sensors which prevented it from climbing or holding altitude.
The data shows the issue but the cause of the malfunction is unknown.

 

[Meta4]

Stories like this make me afraid to fly over water at all.

Flying over water is safer than flying over land.
No obstacles to hit and no or much less interference to deal with.

 

[Dave Pitman]

I've only launched from a boat a few times, and all went okay. Watching some of the videos where guys that launch from boats all the time only to have one basically commit a forced drowning after multiple good take-offs make me wonder if the initial start up imu cal while being on a moving surface can be a crap shoot.