Mavic 3 Lost Control Near Boat

[lensimages]

I need to get to a dock and test “shutting off” the wifi during close approach, but it seems a bit unlikely that the wifi would start to cause interference after 50 normal flights in close proximity to it. I’m currently in a remote area on the BC coast, and will not fly from the boat while at anchor until I get this resolved. Need to be at a dock where alternate landing is possible in case it happens again.

PS: my WIFI is part of a Starlink system, which includes router and dish. So perhaps it could be the dish rather than the router.

 

[RodPad]

lensimages

mavicpilots.com

I'm missing something?

Tim3459575

mavicpilots.com

I have 6 WiFi AP's right now, I just checked. When I started looking in to WiFi issues was for a P3S, never found that was issue at my house, but seen it in the past when I recommended the an app on an android. It was people in the large neighborhoods.

It is WiFi analyzer on google play.

There are so many Mavic's, I will admit I have know idea at that level.

What I don't remember I think the files on the on the Black Box SD card where encrypted by DJI for Mavics?

Rod..

 

[Tim3459575]

Been using it for 6 weeks and not had prior issues with the drone while it was on.

 

[RodPad]

Been using it for 6 weeks and not had prior issues with the drone while it was on.

Thanks, that helps my mind set I think.

I'm really curious what you and these guys come up with.

Rod ..

Added: you still got your bird!

 

[Arnold LeVine]

Can I assume the fog misting up the sensors would not have any effect since you had OA turned off?
Does that turn off EVERY sensor, including ground proximity sensors?
You said you have done 50 flights off the boat, but is this the first flight with the Mavic 3?
(side note) Also, I thought flight log viewer in the past shows the location of the drone during the flight, as you go from line to line. The only location I am seeing is a red circle at each warning notation.
Is my memory tricking me, or does the Mavic 3 show differently on flight log viewer.

 

[Meta4]

I thought flight log viewer in the past shows the location of the drone during the flight, as you go from line to line. The only location I am seeing is a red circle at each warning notation.
Is my memory tricking me, or does the Mavic 3 show differently on flight log viewer.

It works as it should and shows the drone's position properly.
If yours isn't just refresh the page and try again.

 

[Tim3459575]

Can I assume the fog misting up the sensors would not have any effect since you had OA turned off?
Does that turn off EVERY sensor, including ground proximity sensors?
You said you have done 50 flights off the boat, but is this the first flight with the Mavic 3?
(side note) Also, I thought flight log viewer in the past shows the location of the drone during the flight, as you go from line to line. The only location I am seeing is a red circle at each warning notation.
Is my memory tricking me, or does the Mavic 3 show differently on flight log viewer.

the viewer on the link I posted shows the complete flight path. I don’t think it’s related to OA or fog. As soon as it got near the boat it lost its ability to hover in one place. as soon as I moved it away from the boat it regained hover ability.

 

[Frank071]

I take it this is a different issue than the one you refer to here...?

 

[thomasthomas]

Exact same happened to me last year flying over a harbour. With a boat almost directly underneath me, the drone suddenly went into atti mode, which is difficult to control when it happens the first time and you don't know what it is.

 

[Meta4]

Exact same happened to me last year flying over a harbour. With a boat almost directly underneath me, the drone suddenly went into atti mode, which is difficult to control when it happens the first time and you don't know what it is.

That's nothing like what happened in the OP's incident, where he had full GPS for the whole flight.

 

[thomasthomas]

That's nothing like what happened in the OP's incident, where he had full GPS for the whole flight.

Yes, I see now. I only read the first post before assuming it was atti mode. It does sound like atti mode, and having to fight the drift with constant stick movement. I thought, maybe it has a bug where some failure causes it to lose its GPS holding ability but fails to report it.

 

[tlyons]

Hi,

I had a very hard time getting my M3 back to land on my sailboat today. I had OA turned off and I've done this many times without problems. The drone behaved normally until it got about 20' from the boat and then it became very erratic wandering all over without stick inputs. Every time I moved it further away, it became rock steady. It seemed like something on the boat was causing interference. My boat has a 60' mast with typical stainless steel rigging - could that be a factor ? In 3 years of flying Mavic drones, I've never seen anything like this. Attached is the flight log that shows the crazy movements as it got close to the boat. Another theory is that perhaps my Starlink WIFI or other electronics may be creating some sort of interference. Could that have affected the drone ? I've done at least 50 flights off the boat before and never turned off the WIFI or seen that kind of behaviour. In any case, apparently my drone has all of a sudden decided it doesn't want to get closer than 20' to my boat without becoming nearly uncontrollable.

Any suggestions as to the cause of this and how to prevent this situation in future would be much appreciated.

thank you,
Tim

DJI Flight Log Viewer | Phantom Help

Upload and view flight logs from your DJI drones

www.phantomhelp.com

What were the sails doing? Closed hauled or on a reach the wind coming off the sail can be much higher velocity than apparent wind.

Otherwise , turn everything on and try again while tied to the dock close enough to land on a backup spot.

Other thoughts: RADAR? That's a lot more energy than wifi although in a different band.

 

[Meta4]

What were the sails doing? Closed hauled or on a reach the wind coming off the sail can be much higher velocity than apparent wind.

Other thoughts: RADAR? That's a lot more energy than wifi although in a different band.

None of those would explain what the data shows.

 

[coullit]

Maybe the 50mph wind was very turbulent close to the boat, giving the drone some problems trying to remain stable?
No idea, but just a thought

like it was in a 50 mph wind!

 

[tlyons]

like it was in a 50 mph wind!

Like comes off the edge of a sail close hauled in a breeze.
Or like an 8kW (16,000x the power of wifi) beam of radiation was running over the drone every few seconds.
But those aren't it.

 

[Dogpilot]

You local wifi on your boat can be 2.4 and 5.9, Starlink operates and ranges from 17.8-18.6GHz, 18.8-19.3GHz, and 37.5-42.5GHz are used for satellite to gateway transmissions. Now the drone uses Both the 2.4 and 5.9, but each has several channels and they both hop to find the least interference. If you had radar, especially the 10cm band, that could cause problems in the 5.9 band. Harmonic multiples and divisions can also cause interference. I think you had more of the low contrast that comes from PNW clouds and fog.

I flew EW aircraft off carriers as a pilot. Yes the jamming and eavesdropping were my stock in trade. The low contrast flying back on the boat on those marginal days was terrifying. We didn't have a HUD on my aircraft, just indexers. So it was a bit challenging dropping you scan back down to the attitude indications and not on the optical landing device on the deck. It was super easy to get disoriented. I can see see how the visual system would not be all it could be in that situation.

What is weak, are GPS signals. They are easily interfered with, so you drone may have had poor visual and declining GPS information to work with. It does switch to what it thinks is the better attitude source moment by moment. We do have a problem where I live with the military screwing with GPS signals to practice on wartime scenarios. They do usually NOTAM it, but the system in my personal aircraft, a G1000 NXi will drop off line for no really good reason and it seems to correspond to the military having games east of Las Vegas in the huge practice area there, which is just north of me. I do believe that Canada has an offshore practice area off Southern BC where they play the same kind of games with their similar EW aircraft. I don't keep a subscription to those charts anymore, so I can't check, don't go that far north much anymore.

 

[RodPad]

Dogpilot

mavicpilots.com

Rod ..

 

[AnzacJack]

like it was in a 50 mph wind!

Yep…i was already corrected. Thanks for your input

 

[Meta4]

What is weak, are GPS signals. They are easily interfered with, so you drone may have had poor visual and declining GPS information to work with.

His flight data shows GPS reliability of 5/5 and >20 sats for the whole flight.

 

[Dogpilot]

This is what we screw with, you think you have good data from the GPS, but the signal from the sats is very weak, you can easily override it with a slightly stronger signal but with false timing signals. This is exactly how the Iranians diverted one of the USAF's very advanced stealth UAV's. On regular aircraft you have RAIM data to check against GPS timing reliability. We are supposed to check it before we do approaches in poor conditions. The poor little drones don't have any way to check that.

Short version, just having 20 sats with level 5 reception does not mean that the timing signals it is working with are totally valid. Like I mentioned before, the military is playing with this stuff, actively. My bud here in Flagstaff was the head of the Naval Observatory, which is located here. It is the injection source for timing corrections to the US GPS satellite constellation. Learned quite a bit from him, more than I got from my regular exposure to the tech. He sadly passed last year, quite suddenly.

Iran captures UAV