Why did my Mavic fly itself into the ocean?

[BudWalker]

I should clarify. It was active in the settings, but it wasn't in effect in the end, so I couldn't see the drone sink as a result of the VPS. Undoubtedly, if the log continued, the VPS would have taken effect at some point.

You're right. When I read your post I was thinking about @Jackslinger 's supposition that bad VPS data could have been a cause.

One other thing to note is that the VPS stops reporting much closer to the water than if it were launched from ground. At least in this incident where it stopped reporting at 3.2 meters.

 

[tomnjfl]

It's because in the air it uses the GPS to gauge altitude, but if you're close to the ground, the VPS takes effect to keep the Mavic stable above ground. Since the water is fluid, the sensors can't gauge the ground. In many cases the Mavic thinks it's drifting up from the water, when in reality it isn't, and as a result lowers the thrust a bit to compensate and hold steady, when in reality it sinks into the water.

Ah ok, that makes sense. So if you're at a higher altitude over the water it shouldn't really effect it.

 

[Cyberpower678]

Correct.

Ah ok, that makes sense. So if you're at a higher altitude over the water it shouldn't really effect it.

 

[cheertest]

It's possible there's a problem with the link cable from the handset to the phone, it's also possible that moisture may have been the cause with the link cable...ive had similar with my s7 edge which detected moisture in the micro usb when I plugged in new cable.

 

[Jackslinger]

What phone and app were you using? Were you in airplane mode? I'm betting Samsung / android and not in airplane mode.

Guilty as charged. Ive read briefly that airplane mode is a good idea but I havnt heard why Samsung and Android are bad?

 

[Jackslinger]

It's possible there's a problem with the link cable from the handset to the phone, it's also possible that moisture may have been the cause with the link cable...ive had similar with my s7 edge which detected moisture in the micro usb when I plugged in new cable.

In that case did the remote continue functioning properly?

 

[Jackslinger]

Actually, the VPS sensor stopped reporting at 48 secs and never came back on. The .txt converters use 0 to indicate that vpsHeight has no data. DatCon uses a blank to represent this. I've been bitten by this several times. Anyway, the incident isn't related to the VPS sensor reporting bad data.
View attachment 8574

To summarize my contribution we know even less about the cause of this incident

Thanks for posting that, I was trying to figure out how to retrieve that data on datcon. Was hoping to find out at least something regarding the VPS system near the end of the flight. I suppose you did find something, unfortunately that something is nothing.

Perhaps the VPS system kicked in again shortly after the flight log ends?

I still wonder though, even if the app lost a useable connection why the remote became unresponsive and did not give any warning of auto land engaging?

 

[Jackslinger]

Personally, if I have video glitching, I don't fly. Haven't had that problem since many Go App updates ago.

Yes, and this is why I hovered within reach for 30 seconds before flying off. To make sure everything was ok. And it was until moments after I started climbing. This is reflected in the healthy drones analysis.

 

[Andrew F]

Guilty as charged. Ive read briefly that airplane mode is a good idea but I havnt heard why Samsung and Android are bad?

It's seems that Samsung's wifi interferes more so than Apple. I like airplane regardless. Also, it seems the apps are natively designed for Apple then ported to Android so the Android app seems to have more bugs.

 

[Oingo Boingo]

What phone and app were you using? Were you in airplane mode? I'm betting Samsung / android and not in airplane mode.

Nothing wrong with Android. Im on a Huawei mate 8. Never had software crash or malfunction

 

[Andrew F]

Nothing wrong with Android. Im on a Huawei mate 8. Never had software crash or malfunction

Well, Oingo Boingo, since you've never had a problem does that mean no one else has? Read up then reply.

 

[cheertest]

In that case did the remote continue functioning properly?

The app and remote functions work in harmony to allow the mavic to function, there's been reports of cable errors causing all sorts of issues. however I would still suggest flying to close to an airport with radar interference is the probable cause....pilot error

 

[Andrew F]

Nothing wrong with Android. Im on a Huawei mate 8. Never had software crash or malfunction

Just type "Android App Problem" into the above search box and you will see that there have been MANY more problems than with Apple. I have both Samsung and Ipad and have only had problems with the former. But that is anecdotal.

 

[BudWalker]

Thanks for posting that, I was trying to figure out how to retrieve that data on datcon. Was hoping to find out at least something regarding the VPS system near the end of the flight. I suppose you did find something, unfortunately that something is nothing.

Perhaps the VPS system kicked in again shortly after the flight log ends?

I still wonder though, even if the app lost a useable connection why the remote became unresponsive and did not give any warning of auto land engaging?

We are all just guessing since we don't have the .DAT. My guess is that the problem was on the Mavic. You reported that you brought the Mavic back to the kayak; 10 M above and 10 M away. This is shown in the .txt. Then the Mavic drifted and flew on it's own into the water. A failure in the RC and/or App wouldn't have caused that behavior in the Mavic.

After reading what I wrote I realized that another scenario is that the RC failed and the connection was lost. This caused the Mavic to RTH and land (or water).

 

[Jackslinger]

We are all just guessing since we don't have the .DAT. My guess is that the problem was on the Mavic. You reported that you brought the Mavic back to the kayak; 10 M above and 10 M away. This is shown in the .txt. Then the Mavic drifted and flew on it's own into the water. A failure in the RC and/or App wouldn't have caused that behavior in the Mavic.

After reading what I wrote I realized that another scenario is that the RC failed and the connection was lost. This caused the Mavic to RTH and land (or water).

If return to home was triggered the Mavic would have ascended to the specified height and flew to the home position. I thought about that one too... But it just went down, seems more indicative of the auto land function.

I suppose it will remain a mystery :/

Thanks again for the help

 

[BudWalker]

If return to home was triggered the Mavic would have ascended to the specified height and flew to the home position. I thought about that one too... But it just went down, seems more indicative of the auto land function.

I suppose it will remain a mystery :/

Thanks again for the help

If the Mavic is close to the HomePoint (less than 20 meters I think) it won't ascend to the RTH height. It will just descend where it is.

 

[BudWalker]

@Jackslinger the more I think about this the more I think the most likely explanation is that there was an RC disconnect and RTH was initiated. If so, then clearly this is a hardware issue. Six months ago DJI would usually cover this situation under warranty. DJI is a large company and you don't always get the most knowledgeable or cooperative person on first contact. You tried calling; now try chat and/or email. And, demand that the .txt be reviewed by someone who knows what they are doing. I'll leave the exact wording of that to you

 

[HoozierDroneDaddy]

If return to home was triggered the Mavic would have ascended to the specified height and flew to the home position. I thought about that one too... But it just went down, seems more indicative of the auto land function.

I suppose it will remain a mystery :/

Thanks again for the help

But if you already flew it back to your proximity, and your kayak had been drifting, it may have already thought it was Home and had no reason to raise to RTH altitude. Thinking it was already Home, it landed as programmed. Just another possibility IMHO. And if this was the case and your system lost communication, then it should be under warranty. I agree with above, call DJI again until they agree.

 

[cruisinforgold]

A thought on operations over water that involve launching and retrieval.

Would it be a better plan to fly the Mavic over a shore location and set a new RTH point using the app. Then fly your mission. Attempt to fly the Mavic back to your launch vehicle (boat, kayak, canoe). Then RTH automatic/failsafe would be back to a permanent shore-based position that is safe and won't be moving.

I've never tried an updated Home Point based on Mavic's position during flight. If I don't have maps active in the DJI Go App how do I know the Home Point location is actually updated from launch point?

 

[Andrew F]

A thought on operations over water that involve launching and retrieval.

Would it be a better plan to fly the Mavic over a shore location and set a new RTH point using the app. Then fly your mission. Attempt to fly the Mavic back to your launch vehicle (boat, kayak, canoe). Then RTH automatic/failsafe would be back to a permanent shore-based position that is safe and won't be moving.

I've never tried an updated Home Point based on Mavic's position during flight. If I don't have maps active in the DJI Go App how do I know the Home Point location is actually updated from launch point?

Thats a good approach. Even without maps you still get a notice that the home point is set.