M2P Inexplicably Crashed Into The Pacific

[moonlitnite]

I was flying over the Northwest Pacific with my new Air, 40% battery and, BAM, all of the sudden 0 % battery and force landing kicked in. Luckily wind carried it into pithing 10" of shore and I was able to recover it. I am having a dispute with DJI as they charged me a replacement on my Refresh account and it was product failure. I will let you know the outcome. My flight log has it recorded.

Yeah, especially prior to the P1 and P2 batteries that didn't employ smart partial discharging, folks would "discover" that with damaged battery chemistry the software gives false reading (and sense of security) on remaining battery capacity. It's a terrible feeling to be trying to throttle up on a low battery condition when the software overrides you more, and More, and MORE to descend and land in a terrible place.

 

[GSE D Land]

Hi,

I don't know if this is of importance/help but the last entry in your flight record has the following warning:

Upward Obstacle Detected; Weak signal. Adjust antenna and avoid signal block.

Not sure if it was still in the air at the time or it flipped upside down and the "upward obstacle" was water actually
As home point was high and you are just 6o feet bellow at the time maybe can be a hit?
Good luck with getting something back from DJI or/and try selling the remote but... people tend to loose the aircraft not the remote. Better keep it as spare/secondary remote if you get another Mavic 2

 

[Meta4]

Not sure if it was still in the air at the[ time or it flipped upside down and the "upward obstacle" was water actually
As home point was high and you are just 6o feet bellow at the time maybe can be a hit?

Looking at the flight data, there's no sudden change in the pitch, yaw & roll that would indicate an impact.
It had a 27 degree roll that was due to the sharp rudder input shortly before and had not flipped.

 

[Earthman]

Meta4s interpretation of the flight log reveals several early mission procedural mistakes that are likely contributing factors. However, your perceived lack of response to control input may be real as well if the flight battery reached the low voltage set point and triggered the flight control software to immediately land the drone regardless of where it was. In this case, you still had some control of the drone (assuming you had a good radio link) to help avoid objects, but the drone is landing to avoid damage to the battery. This behavior is built into the software as explained in the manual.

At what voltage did you set for the drone to land after trying to return home on a low battery? Could you have hit that per-programmed set point under load?

Other potential causes or contributing factors include a controller or display device malfunction or low battery. They too are complex electromechanical devices containing microprocessors running software, all of which is in the control loop and subject to failure. Maybe such failures have low probabilities but I assume they can happen.

Personally, given the complexity of sUAS systems and the failure rate (especially for multirotor aircraft), I do not fly over water unless I can afford the loss of the drone and data collected. I never fly over/near unprotected people with the stuff I fly - not reliable enough. Also, I don't control everything in the flight environment, nor I'm I free of making mistakes.

 

[Meta4]

Meta4s interpretation of the flight log reveals several early mission procedural mistakes that are likely contributing factors. However, your perceived lack of response to control input may be real as well if the flight battery reached the low voltage set point and triggered the flight control software to immediately land the drone regardless of where it was. In this case, you still had some control of the drone (assuming you had a good radio link) to help avoid objects, but the drone is landing to avoid damage to the battery. This behavior is built into the software as explained in the manual.

The flight data shows no hint of any loss of control.
The drone responses correspond perfectly with the joystick input for the whole flight.
There's also no hint of any low battery autolanding.
In post #13, the cause is explained ..
from 22:17.6 where you were at 17 ft higher than the launch point, you were flying forward at full speed and pulled the left stick down hard until 22:25.8.
Holding the left stick down hard for 8 seconds, your Mavic descended 77 ft which brought your altitude down to 60 ft lower than your launch point.
The drone was flown into the sea.

Personally, given the complexity of sUAS systems and the failure rate (especially for multirotor aircraft), I do not fly over water unless I can afford the loss of the drone and data collected.

Most of my flying is over the sea and it's a great place to fly.
There are no obstacles to hit or block signal and no interference.
I consider it a safer place to fly than on land.

 

[sar104]

Meta4s interpretation of the flight log reveals several early mission procedural mistakes that are likely contributing factors. However, your perceived lack of response to control input may be real as well if the flight battery reached the low voltage set point and triggered the flight control software to immediately land the drone regardless of where it was. In this case, you still had some control of the drone (assuming you had a good radio link) to help avoid objects, but the drone is landing to avoid damage to the battery. This behavior is built into the software as explained in the manual.

At what voltage did you set for the drone to land after trying to return home on a low battery? Could you have hit that per-programmed set point under load?

Other potential causes or contributing factors include a controller or display device malfunction or low battery. They too are complex electromechanical devices containing microprocessors running software, all of which is in the control loop and subject to failure. Maybe such failures have low probabilities but I assume they can happen.

Personally, given the complexity of sUAS systems and the failure rate (especially for multirotor aircraft), I do not fly over water unless I can afford the loss of the drone and data collected. I never fly over/near unprotected people with the stuff I fly - not reliable enough. Also, I don't control everything in the flight environment, nor I'm I free of making mistakes.

Please just read the thread. The OP confused the elevator and throttle and flew it into the sea. It's that simple - the aircraft behaved properly and there is no mystery left here.

 

[rt60]

What's the 'Not Allowed to change Aircraft Mode. If needed, enable it in Main Controller Settings.' in the log? If the controller changed from Mode 2 to Mode 1, the throttle and elevator sticks are reversed.

 

[sar104]

What's the 'Not Allowed to change Aircraft Mode. If needed, enable it in Main Controller Settings.' in the log? If the controller changed from Mode 2 to Mode 1, the throttle and elevator sticks are reversed.

That's not what happened either. This message refers to the flight mode (S, P, T), which cannot be changed unless multiple flight modes is enabled in the GO app.

 

[carlb60]

Typical BS of 15% off RRP for the drone only. Cheaper to buy one from a shop and sell the bits you dont need.

Ya it’s an odd mistake, I’m fairly experienced at flying these drones and have never Ever screwed up that bad. The flight record clearly shows what happened. It’s hard to wrap my head around. I must have had the left stick down but until the flight record was analyzed, I totally could not figure why it had dropped. In my head, I had that thing climbing hard trying to get out of the descent. It’s a bummer for sure and costly mistake. It won’t be made again.

I have been flying DJI Drones for six years now and the Mavic 2 Pro is the only one that inexplicably crashes like this. Mine has done it four times now. Once in a lake - new camera, Once on a lawn - no damage, once on concrete from about 10 feet - minor damage, last time it crashed from hover above me about 10' hitting me in the head and breaking my iPhone - painful but no damage to the drone. I don't care what the logs show, the Mavic 2 pro is unstable and subject to random crash!

 

[sar104]

I have been flying DJI Drones for six years now and the Mavic 2 Pro is the only one that inexplicably crashes like this. Mine has done it four times now. Once in a lake - new camera, Once on a lawn - no damage, once on concrete from about 10 feet - minor damage, last time it crashed from hover above me about 10' hitting me in the head and breaking my iPhone - painful but no damage to the drone. I don't care what the logs show, the Mavic 2 pro is unstable and subject to random crash!

I'm confused - you don't care what caused this crash, and even though the logs clearly show it was pilot error you are holding it up as an example of the M2 being unstable and subject to random crashes?

And you have crashed yours four times? I'm sure that everyone would love to see the logs from those events.

 

[Meta4]

I have been flying DJI Drones for six years now and the Mavic 2 Pro is the only one that inexplicably crashes like this.
I don't care what the logs show, the Mavic 2 pro is unstable and subject to random crash!

Go back and read post #13.
There was nothing inexplicable at all about this incident.
You should care about the logs.
The recorded flight data clearly shows exactly what happened and why.

 

[JDawg]

I have been flying DJI Drones for six years now and the Mavic 2 Pro is the only one that inexplicably crashes like this. Mine has done it four times now. Once in a lake - new camera, Once on a lawn - no damage, once on concrete from about 10 feet - minor damage, last time it crashed from hover above me about 10' hitting me in the head and breaking my iPhone - painful but no damage to the drone. I don't care what the logs show, the Mavic 2 pro is unstable and subject to random crash!

Hey Neighbor, logs or it did not happen

 

[carlb60]

The logs apparently log what the electronics think happened. I know all the experts say they are infallible, but I know on two of the uncommanded crashes I experienced, I was not touching the controls, and I always had at least 50% battery. Until DJI recognizes that an uncommanded descent can happen because of some electronic glitch, we will continue to see these reports

 

[Meta4]

The logs apparently log what the electronics think happened. I know all the experts say they are infallible, but I know on two of the uncommanded crashes I experienced, I was not touching the controls, and I always had at least 50% battery. Until DJI recognizes that an uncommanded descent can happen because of some electronic glitch, we will continue to see these reports

You obviously haven't read recorded flight data and you are just saying what you think happens.
The data is very reliable.
You can compare the joystick input against what the sensors detected and in this incident, the flight performance perfectly matches the control inputs.

Until DJI recognizes that an uncommanded descent can happen because of some electronic glitch, we will continue to see these reports

Until a flyer finds out what really happened and learns from that, they are at risk of making the same mistake.
Put up your own flight data and you'll probably find a different story from what you think happened.

 

[Kilrah]

I have been flying DJI Drones for six years now and the Mavic 2 Pro is the only one that inexplicably crashes like this.

There's been nothing inexplicable at all neither here nor in most other crash threads.

The logs apparently log what the electronics think happened.

No, they log the state and the sensor data. Even if the electronics interpreted that data wrong a human reading the sensor data will spot it and figure out the reason.

I know all the experts say they are infallible

Because a discrepancy would be easy to spot but that's never happened yet...

I know on two of the uncommanded crashes I experienced, I was not touching the controls, and I always had at least 50% battery.

Post them logs then.

 

[carlb60]

BTW thanks for your comments, but I think many of us should consider that there are many sources of

You obviously haven't read recorded flight data and you are just saying what you think happens.
The data is very reliable.
You can compare the joystick input against what the sensors detected and in this incident, the flight performance perfectly matches the control inputs.

Until a flyer finds out what really happened and learns from that, they are at risk of making the same mistake.
Put up your own flight data and you'll probably find a different story from what you think happened.

on two of my M2P crashes (lawn and lake) the M2P flipped upside down (pointing toward me hover about 8' high and flipped clockwise) and crashed that way. Scared my cat half to death on the lawn. I don't know any control inputs that will make it flip. I agree the logs show something, but I think they could easily show incorrect info. It isn't what I think happened or what the logs show, it's what did happen that matters. I really think this forum would be more productive if what the pilot says was taken more seriously.

 

[sar104]

BTW thanks for your comments, but I think many of us should consider that there are many sources of

on two of my M2P crashes (lawn and lake) the M2P flipped upside down (pointing toward me hover about 8' high and flipped clockwise) and crashed that way. Scared my cat half to death on the lawn. I don't know any control inputs that will make it flip. I agree the logs show something, but I think they could easily show incorrect info. It isn't what I think happened or what the logs show, it's what did happen that matters. I really think this forum would be more productive if what the pilot says was taken more seriously.

It would certainly be more productive if you posted your logs, rather than just asserting that they are probably wrong. As for taking pilots seriously - perhaps you could point to thread where the pilot was not taken seriously, or where the logs turned out to be wrong. Perhaps you could explain what you mean by the logs being wrong, and why demonstrably poor human recollection of fast events is to be preferred. Do you have even the faintest clue what is in the these logs?

 

[Kilrah]

I don't know any control inputs that will make it flip.

Nope, so you shouldn't be scared to show the logs since if it flipped it could not be due to your input. But it could be due to something that the logs could allow to determine.

Anyway I wonder why after 4 crashes and being fully confident you're not the cause you're still trying to fly the thing and considering all Mavic 2s are faulty instead of realizing the obvious aka your unit probably being a dud and getting that fixed/replaced...

It isn't what I think happened or what the logs show, it's what did happen that matters. I really think this forum would be more productive if what the pilot says was taken more seriously.

Data is data and is hard fact. People are people and people are known to frequently screw things up without even noticing it or completely losing recollection of what happened, when not "remembering" the opposite. Any accident investigation will show that.
Logs will precisely show what happened.

 

[Meta4]

I agree the logs show something, but I think they could easily show incorrect info. It isn't what I think happened or what the logs show, it's what did happen that matters. I really think this forum would be more productive if what the pilot says was taken more seriously.

Having investigated a few hundred flight logs and incidents, I'll always stick with data that shows sensor data for each 1/10th of a second rather than trusting vague and inaccurate memories and interpretations based on potentially poor observation, or even in a small number of cases outright lies.
I've seen one or two flight logs where the data stream was giving unbelievable numbers because of some complication.
In those cases it was always obvious that there was a problem with the data.

 

[JDawg]

CarlB60, There is absolutely no reason not to post your Log, the guys here are incredible at analyzing it and for some weird reason I really enjoy what they do even though I am clueless about it. Oh, and how many cats do you have...?