M2Z OccuSync Glitch = Lost Drone

[Kazzaaz]

I lost connection for no reason which resulted in what I could only assume was an erratic spiral towards my M2Z's death in some unknown location.
Luckily I was attempting to break some speed records and was recording my phone screen as this happened.
The rest of the details are in the video.

I spent about 30 minutes at that location in case the drone was hovering and was landing. (Lost connection mode set to "Hover")
And hours looking in the area with no success.

WTF happened here!? Feeling pretty unlucky especially after buying DJI care just hours earlier (thinking, I'll definitely crash again but there's no way I'll lose it).
Trying not to be too upset but... I have a hard time not pointing a finger at DJI.

 

[JAW]

You had a couple high wind warnings.....maybe it drifted and did a water splash down. No comment on the rest of the flight.

 

[BluePuffin]

To me it looks like it ended up drifting away due to high wind. You can see the dramatic change in speed when going against and with the the wind. When you were flying south the speed stayed around 22 MPH (indicating you are fighting the wind), and when flying back to the east the speed went up to 60 MPH at some points (likely flying with the wind). I could be wrong but it also seemed like you may have been flying outside of VLOS, and you likely didn't notice it drifting. I also noticed the drone was not in sports mode when it disconnected so it likely tried to return home and couldn't keep up with the wind. Unfortunately, this seems to be the common cause of most flyaway and DJI often does not cover it.

 

[BD0G]

@sar104

 

[JAW]

@sar104View attachment 67337

Crack me up, BDOG!

 

[sar104]

@sar104View attachment 67337

10-4

 

[Mavic Pro Platinum]

I lost connection for no reason which resulted in what I could only assume was an erratic spiral towards my M2Z's death in some unknown location.
Luckily I was attempting to break some speed records and was recording my phone screen as this happened.
The rest of the details are in the video.

I spent about 30 minutes at that location in case the drone was hovering and was landing. (Lost connection mode set to "Hover")
And hours looking in the area with no success.

WTF happened here!? Feeling pretty unlucky especially after buying DJI care just hours earlier (thinking, I'll definitely crash again but there's no way I'll lose it).
Trying not to be too upset but... I have a hard time not pointing a finger at DJI.

70+ MPH!

 

[sar104]

The wind was certainly a problem - around 30 mph out of the west when the aircraft reached 500 m AGL. I'll refrain from commenting on the wisdom of doing that in Detroit airspace. The reason for the disconnect is not clear though - there were no signal issues leading up to it, which suggests that it may have suffered a battery disconnect or an FC crash. If that happened then it should be possible to predict the landing area fairly accurately. I ran a few wind drift simulations with fixed and variable winds based on the computed wind speeds as a function of altitude during this flight, and generated the following descent tracks:



I would go with the green and blue tracks first - the red one is a uniform 15 m/s wind field as a bounding condition. I'd take a look around 38412 Kingway Ct. for starters.



If it did not disconnect due to a catastrophic failure then it will have drifted on the wind during hover and then autolanded, which will take a bit of careful estimating, but is certainly doable.

While I said I wouldn't comment, I'm pretty unimpressed by the idea of your aircraft drifting out of control at 1600 ft AGL in busy airspace, or falling from that height on a residential area.

 

[sar104]

On a closer look at the wind data, I'll revise that estimate to 38420 Kingsway Ct.

 

[AnzacJack]

Being in illegal airspace, is it a possibility someone hit it with one of the jamming devices? Seems strange to just drop of the grid like that.

 

[sar104]

Being in illegal airspace, is it a possibility someone hit it with one of the jamming devices? Seems strange to just drop of the grid like that.

It's not common, but it does happen - several such cases have been reported here since the M2 was released. I'm not sure how anyone would jam the comms link from at least half a kilometer away.

And note that it was in Class G airspace, just much too high. Not technically illegal under Part 101 though - it would be challenged under 101.43 - endangering the NAS.

 

[AnzacJack]

Yea I guess. Unlucky for such an event to happen.

 

[sar104]

Yea I guess. Unlucky for such an event to happen.

Or karma.

 

[Meta4]

I lost connection for no reason which resulted in what I could only assume was an erratic spiral towards my M2Z's death in some unknown location.

Why? .. Loss of connection doesn't result in a fall-from-the-sky incident.
It just leaves the drone hovering in place.
Loss of power is what causes fall-from-the-sky incidents.

I spent about 30 minutes at that location in case the drone was hovering and was landing. (Lost connection mode set to "Hover")

Why ?? Not a good thing to do if you want to get the drone back.[/QUOTE]

 

[tmtwomey]

And note that it was in Class G airspace, just much too high. Not technically illegal under Part 101 though

Are you referring to the height not being technically illegal? Because my understanding is that no drone in the US can fly above 400' AGL unless within 400' of a structure or operating under a waiver.

Here we are flying at 1600' at night over a populated area trying to break some speed records. We are flying in the same airspace as manned aircraft. If a Mavic 2 falls from this height and kills someone, think of all the good press. All I can say is don't complain when new regulations further restrict your ability to fly.

 

[sar104]

Are you referring to the height not being technically illegal? Because my understanding is that no drone in the US can fly above 400' AGL unless within 400' of a structure or operating under a waiver.

Here we are flying at 1600' at night over a populated area trying to break some speed records. We are flying in the same airspace as manned aircraft. If a Mavic 2 falls from this height and kills someone, think of all the good press. All I can say is don't complain when new regulations further restrict your ability to fly.

You are confusing different laws. Your understanding is correct in terms of Part 107 but, if a pilot is flying recreationally under Part 101, there is no fixed altitude limit. As for your opinion of flight, you won't get any disagreement from me. Although a falling Mavic reaches 95% of its terminal velocity in less than 150 ft, so the altitude is not really a factor from that perspective. My issues are with the airspace and that falling from 1600 ft there is going to be no warning on the ground. At least if it is within 400 ft there is a chance that it will be heard.

 

[sar104]

Why? .. Loss of connection doesn't result in a fall-from-the-sky incident.
It just leaves the drone hovering in place.
Loss of power is what causes fall-from-the-sky incidents.

Maybe. This event looks more and more bizarre. The pilot appears to have been switching back and forth between sport and manual (flycState 0). That's full manual mode - he had modified the parameters to make it accessible via the flight mode switch. And the aircraft was in manual when the connection was lost. I'm not sure what happens on connection loss in that mode.

 

[sar104]

Here's the overview of this flight. I've seen some gratuituously stupid piloting but this takes things to a new level. It's definitely what the FAA had in mind when they wrote 101.43.

 

[Kazzaaz]

First things first, I completely deserve the scornful comments and can assure you I feel like quite an idiot here. "One more high flight for the logs, then that's it!"
I didn't anticipate losing control of course, but now I've seen firsthand how hard **** can hit the fan if you do. Not that I am in a position to defend myself, but I did check FlightRadar to make extra sure I was out of the way of any aircraft... Just had to throw that out there. Stupid, stupid extreme-risk low-reward flying - that has effectively lost all of its appeal anyway after this incident! Karma for sure.

Back to the incident, there's a few things to clear up.
The first few comments about wind preventing an RTH don't apply here. The signal-loss action was set to hover. I do this because I often run out-of-range Litchi missions that will RTH early as it is on the fringe of connection. I have briefly lost connection countless times in this mode and have never had an issue restoring it.

It certainly is a bizarre incident, and yes - it was in the mode I will never EVER use again - full manual. With all motor control relying on the pilot, I also can only guess as to what happens when it loses connection.
Many drones that operate normally in this mode will "freeze" on the last command received and zoom out of control. If this is the case then the drone could be a mile away.
If you look at the aircraft's data right before it loses connection, it is actually tilted slightly backwards and could have actually ended up west!

This mode was how I was able to achieve such high speeds but this is the second malfunction that has led to a catastrophic failure and I learned the hard way the aircraft is simply NOT built for it.
Don't put your Mavic in this mode folks.

With a laughable degree of irony and humility I appreciate your help, @sar104 . I just went out to the area for kicks but it is too dark to search and would feel sketchy walking around. If it makes any significant difference, I can say with high certainty (using pitch/roll profiles while in manual) that the wind at that exact time was headed at about 70 degrees. (Slightly north as well)

I'll start the search again tomorrow but honestly I should probably just chalk this one up to a hard lesson learned and stick to the cheapo drones on Amazon.

 

[sar104]

First things first, I completely deserve the scornful comments and can assure you I feel like quite an idiot here. "One more high flight for the logs, then that's it!"
I didn't anticipate losing control of course, but now I've seen firsthand how hard **** can hit the fan if you do. Not that I am in a position to defend myself, but I did check FlightRadar to make extra sure I was out of the way of any aircraft... Just had to throw that out there. Stupid, stupid extreme-risk low-reward flying - that has effectively lost all of its appeal anyway after this incident! Karma for sure.

Back to the incident, there's a few things to clear up.
The first few comments about wind preventing an RTH don't apply here. The signal-loss action was set to hover. I do this because I often run out-of-range Litchi missions that will RTH early as it is on the fringe of connection. I have briefly lost connection countless times in this mode and have never had an issue restoring it.

It certainly is a bizarre incident, and yes - it was in the mode I will never EVER use again - full manual. With all motor control relying on the pilot, I also can only guess as to what happens when it loses connection.
Many drones that operate normally in this mode will "freeze" on the last command received and zoom out of control. If this is the case then the drone could be a mile away.
If you look at the aircraft's data right before it loses connection, it is actually tilted slightly backwards and could have actually ended up west!

This mode was how I was able to achieve such high speeds but this is the second malfunction that has led to a catastrophic failure and I learned the hard way the aircraft is simply NOT built for it.
Don't put your Mavic in this mode folks.

With a laughable degree of humility I appreciate your help, @sar104 . I just went out to the area for kicks but it is too dark to search and would feel sketchy walking around. If it makes any significant difference, I can say with high certainty (using pitch/roll profiles while in manual) that the wind at that exact time was headed at about 70 degrees. (Slightly north as well)

I'll start the search again tomorrow but honestly I should probably just chalk this one up to a hard lesson learned and stick to the cheapo drones on Amazon.

Actually I suspect that your flight maneuvers eventually caused the battery to dislodge. I still think that the most likely endpoint is where I indicated earlier, on Kingsway Ct.

If the comms link is lost in ATTI mode then the aircraft reverts to P-GPS. If I were programming the firmware then I'd have the same behavior occur in Manual mode. That needs to be tested though - I'll run that one at the weekend.

If you find it - please don't do that again.