2 FlightLog review requests - one fell during landing (mini 4 pro) and other was lost mid-flight (mini 2 SE)

[flyernj]

Hi all!

First time poster here. I wanted to see if anyone can help with 2 flight log reviews:

1 - This is from a Mini 2 SE. I lost this in Nepal more than a year ago. I was using Litchi app and it was night time. The drone was flying without any warnings but the video feed suddenly froze and never came back. It was at approx. 80-85% battery life then. So I immediately drove towards the last location with RC and phone still on Litchi app but found no trace of the drone there. While driving there also I kept checking if the video feed resumes but it never came back. Flight logs - DJI Flight Log Viewer - PhantomHelp.com

2 - After my Mini 2 SE was lost, I bought a Mini 4 Pro. No issues until a few days ago. I was at around 10% battery level and it started to auto land. Home point was 1st floor balcony. But when forced auto landing, I tried to maneuver it back to the balcony (19:38 I give full throttle hoping to back into homepoint). I think the obstacle avoidance sensors prohibited the drone to come back to the balcony due to the door. So, it started to land right outside the railing (few feet away) on the lawn. At about the homepoint altitude (floor of the balcony ~ 10-12 feet higher than the lawn), motors stopped and it just fell off, damaging propellers. At time of crash, battery was 8% and it was bright and sunny outside . Flight logs - DJI Flight Log Viewer - PhantomHelp.com

Any help with either is appreciated!

 

[Meta4]

I bought a Mini 4 Pro. No issues until a few days ago. I was at around 10% battery level and it started to auto land. Home point was 1st floor balcony. But when forced auto landing, I tried to maneuver it back to the balcony (19:38 I give full throttle hoping to back into homepoint). I think the obstacle avoidance sensors prohibited the drone to come back to the balcony due to the door. So, it started to land right outside the railing (few feet away) on the lawn. At about the homepoint altitude (floor of the balcony ~ 10-12 feet higher than the lawn), motors stopped and it just fell off, damaging propellers. At time of crash, battery was 8% and it was bright and sunny outside .

The mini 4 pro is a simple case of flying until the battery didn't have enough power to sustain flight.
You reached critical low voltage (3.2 volts per cell) at 17:46.2 with the drone still 570 feet from home.
You continued flying for another two minutes, depleting the battery beyond critical low voltage level to cell voltages of <3 volts per cell.

Your description suggests the drone fell 10-12 feet, but the VPS data indicates the drone was 13 feet above the ground (?) when you moved out from the balcony and it autolanded slowly down to 0.6 feet above the ground and rolled over on landing, presumably on an uneven surface.

 

[Meta4]

1 - This is from a Mini 2 SE. I lost this in Nepal more than a year ago. I was using Litchi app and it was night time. The drone was flying without any warnings but the video feed suddenly froze and never came back. It was at approx. 80-85% battery life then. So I immediately drove towards the last location with RC and phone still on Litchi app but found no trace of the drone there.

You had an in-flight incident at 4:30.7, that resulted in the drone tilting forward about 80°, rolling to one side at 117° and falling to earth.
At the time the drone was about 200 feet above the ground below it and should have been clear of obstacles.
The data stream ends with the drone still about 100 feet above the ground

The incident is unusual in that the data indicates the drone initially fell from 200 feet but signal was lost for 19 seconds from 0.3 seconds after the incident started.
If a drone was falling from that height, it should reach the ground in approximately 5-6 seconds.
But when signal was returned after 19 seconds, the drone's indicated height was 59 feet ... and the drone's height was increasing despite being rolled over at 106° and pitched forward at 28° !!!
During the last 3 seconds of data, the drone continues to rise to 100 feet above the ground which would be impossible under its own power rolled at that angle.

The only explanation I could imagine would be that the drone was grabbed by a large was grabbed by a bird of prey and taken away?
If it was at night, the bird was likely to be a large owl
That's just a wild guess but I can't think of anything else to fit the data.

 

[flyernj]

Thanks for replying Meta4.

2 questions:

- I feel that I have had battery levels below 5% in the past and still stable landing. In case #2, last status was 8% so it is hard to believe that motors stopped due to low battery.

- Do you have any comment on the Motor stuck message in the last few seconds?
I was standing right at the door of the balcony. The drone started landing outside the first floor balcony right on the lawn. I may be slightly off but it definitely didn’t roll off 1 foot. Actually there was a chair on the ground and it hit the chair. I am not sure how accurate VPS is since it jumps a lot in comparison to IMU throughout the flight. If I believe last IMU of -7.9 feet then it fell about 4 feet (assuming total distance between balcony homepoint and lawn is 12 feet or more).

 

[Meta4]

- I feel that I have had battery levels below 5% in the past and still stable landing. In case #2, last status was 8% so it is hard to believe that motors stopped due to low battery.

The battery percentage isn't what matters.
It's the cell voltage that's important

- Do you have any comment on the Motor stuck message in the last few seconds?

That would normally suggest that a prop has hit something.
But I can't see that message in the data I'm looking at.

I am not sure how accurate VPS is since it jumps a lot in comparison to IMU throughout the flight.

VPS is an accurate measurement on the distance to whatever is below the drone.
The IMU height sensor is subject to drifting during the time of a flight and it doesn't indicate the height above whatever is below the drone.
Your VPS data suggests the drone steadily descended to the ground.

 

[eEridani]

Once upon a time DJI batteries did need the occasional calibration. That may explain the full depleted battery showing 8%. Also, it was not too warm that day, and batteries don't do as well when they are chilly.

Just guessing based on the basic data shown, but that flight in Nepal looks like there's a chance you impacted the Dharahara - your drone was at 200' and the tower is 236' and the route puts you directed towards it at 10 mph when you moved the controls to stop and gain altitude. Or, perhaps you were so close you flew up into the balcony as you tried to go over it?

 

[flyernj]

It's the cell voltage that's important

Understood! I guess then 3V per cell is absolute minimum when it needs to be on the ground?

 

[flyernj]

Once upon a time DJI batteries did need the occasional calibration. That may explain the full depleted battery showing 8%. Also, it was not too warm that day, and batteries don't do as well when they are chilly.

Just guessing based on the basic data shown, but that flight in Nepal looks like there's a chance you impacted the Dharahara - your drone was at 200' and the tower is 236' and the route puts you directed towards it at 10 mph when you moved the controls to stop and gain altitude. Or, perhaps you were so close you flew up into the balcony as you tried to go over it?

View attachment 179992

Great picture! Yes that looks like a possibility from the data. But I remember getting a great shot of the top of the tower from close and then going past it a little and doing a 180 degree turn, now facing the remote. I feel like signal was lost then during the turn - as it often did from past experience. So then I tried to raise it up with no avail. It must have initiated RTH and it could have possibly hit the tower then - but I do keep RTH at 600+ feet.

I do notice that it drops from 200 ft to 60 feet during signal loss period. Not sure how that would happen. If anything, it should have risen to my 600+ RTH altitude.

I did check with the security there and they said while I needed a permit to fly close there, they do not have any anti-drone tech. We all looked inside the compound for a bit but no signs.



How does one re-calibrate Mini 4 Pro batteries?

 

[Meta4]

Yes that looks like a possibility from the data.

Your drone never hit that tower - see post three for details.
If it had hit the tower, it would have fallen to the ground or if you were very lucky, collided, recovered and come back.
It wouldn't have tilted to a crazy angle, lost altitude and then climbed while still at a crazy angle.

I feel like signal was lost then during the turn ...
It must have initiated RTH and it could have possibly hit the tower then

Signal was lost for 10 seconds, but but that was after the incident started.
The drone never entered RTH during the flight.

I do notice that it drops from 200 ft to 60 feet during signal loss period. Not sure how that would happen. If anything, it should have risen to my 600+ RTH altitude.

See post #3.

 

[flyernj]

You had an in-flight incident at 4:30.7, that resulted in the drone tilting forward about 80°, rolling to one side at 117° and falling to earth.
At the time the drone was about 200 feet above the ground below it and should have been clear of obstacles.
The data stream ends with the drone still about 100 feet above the ground

The incident is unusual in that the data indicates the drone initially fell from 200 feet but signal was lost for 19 seconds from 0.3 seconds after the incident started.
If a drone was falling from that height, it should reach the ground in approximately 5-6 seconds.
But when signal was returned after 19 seconds, the drone's indicated height was 59 feet ... and the drone's height was increasing despite being rolled over at 106° and pitched forward at 28° !!!
During the last 3 seconds of data, the drone continues to rise to 100 feet above the ground which would be impossible under its own power rolled at that angle.

The only explanation I could imagine would be that the drone was grabbed by a large was grabbed by a bird of prey and taken away?
If it was at night, the bird was likely to be a large owl
That's just a wild guess but I can't think of anything else to fit the data.

Sorry, I missed this post entirely!

Great analysis. I just looked at the angles and it is just too extreme. If you notice after the connection is restored, vertical speed - despite my full throttle up - decreases from 1.9 mps to 1.1 very smoothly. I know this is too low because after takeoff, it achieved vertical speeds of 7-9 mps. It was always in Normal mode. So that leads me to believe that the props were under a lot of stress and slowed down despite full throttle up. Perhaps because something had already damaged the props.

The two things I can't understand is why it didn't start RTH when it lost signal. Maybe because it was being "carried" by the bird so it never stopped movement?
And what happened afterthe last data point. At this point, the drone is at 100.7 ft with pllenty of battery. Maybe the bird comes again and knocks it down fullly?

 

[Meta4]

that leads me to believe that the props were under a lot of stress and slowed down despite full throttle up. Perhaps because something had already damaged the props.

The drone was tipped over 106° from the horizontal .. see the artificial horizon display here:

While held in that position no amount of joystick input would make the drone climb (or stay in the air).

The two things I can't understand is why it didn't start RTH when it lost signal.

You only lost the downlink.
You need to lose downlink+uplink for the drone to initiate RTH.

And what happened afterthe last data point. At this point, the drone is at 100.7 ft with pllenty of battery. Maybe the bird comes again and knocks it down fullly?

If a bird was the cause of the issue, the drone was held by the bird and carried off.
The cause of height increase was the bird climbing, not the drone.
After signal was lost we can only guess.

 

[flyernj]

The drone was tipped over 106° from the horizontal .. see the artificial horizon display here:

While held in that position no amount of joystick input would make the drone climb (or stay in the air).

You only lost the downlink.
You need to lose downlink+uplink for the drone to initiate RTH.

If a bird was the cause of the issue, the drone was held by the bird and carried off.
The cause of height increase was the bird climbing, not the drone.
After signal was lost we can only guess.

Thanks @Meta4!

I always see birds getting curious and coming close but something grabbing it and trying to fly is just crazy. Mini 2 is substantially louder then mini 4 too so that makes it all the more weird. Additionally, that location is right in the middle of the city and big birds are not common at all.

I know this is DJI recording and saving the logs but could it be that maybe some sensors malfunctioned?
Also, could you explain what downlink and uplink means?
Lastly, what could cause the data to stop abruptly? I assume instant power shut off. The CSV data does show isFlying flag to be 0 which supports the idea that it was being carried by something up. If it actually fell, data should end at IMU of 0 feet when it would crash. Do you agree?

 

[Meta4]

I know this is DJI recording and saving the logs but could it be that maybe some sensors malfunctioned?

When I did my initial analysis, I checked the ground level where the drone was lost and found it to be very close to that at the launch point and assumed that a height of 200 ft would mean the drone was clear of any obstacles.
I wasn't aware that there was a 275 ft tower where you were flying.
@eEridani did well to find this in post #6.

A plot of the drone's flightpath puts it on a collision course with the Dharahara.

My suggestion that a large predatory bird took the drone was my best guess at an explanation to fit what the data showed.
I was baffled by the drone's height data increasing after full signal was restored.
In cases where a drone flies into the ground and power is not lost, IMU altitude can be affected by a hard impact and show the height continuing to fall and reaching large negative numbers that are obviously false.
I've never seen a case where the IMU height showed an increase after a collision.

But given the drone's track was directly toward the tower which was significantly higher than the drone, the only logical explanation is that your drone collided with the tower and fell to the ground at the base of the tower.
The distracting height data is false.
The drone came to rest tipped over 106° from the horizontal and unable to receive good GPS signal in that position.

Also, could you explain what downlink and uplink means?

There are two parts to your signal connection, an uplink (control signal from your controller to the drone) and a downlink (video and data signal sent back from the drone).

Lastly, what could cause the data to stop abruptly? I assume instant power shut off

The data stops if either the drone shuts down, battery is ejected etc or if signal is completely blocked by a solid obstacle.
In your case the drone didn't lose power following the collision, but probably shut down after being tipped over and stationary for two seconds?
I'm not sure how that works with the IMU continuing to indicate increasing height.

. The CSV data does show isFlying flag to be 0 which supports the idea that it was being carried by something up.

There are 194 columns of data in the CSV spreadsheet.
I cannot find one with the label isFlying.
Which column are you referring to?

If it actually fell, data should end at IMU of 0 feet when it would crash. Do you agree?

Not necessarily.
The barometric sensor in the IMU shows height relative to that of the launch point.
If the drone fell, it would show the height of the position at which the fall stopped, which probably isn't the same as where the drone was launched.
Also the barometric sensor commonly shows a drift of 10 feet or more over the duration of a flight.