Landing

[Andrew F]

ll

Because when I'm out over the ocean it's easy to lose perspective and I like the extra protection from accidentally setting it down. Highly unlikely, but still. The technology is well considered even if a bit buggy.

Also, there is nothing of value between .04 and .05 that I can find. Why not keep the easier and more logical Functionality? My use is likely atypical as I'm over water the vast majority of the time.

 

[JLane]

I thought you guys had a bit of a bee in your bonnet about this issue until just now. Well I was wrong.

I thought that when it starts spuriously auto-landing from altitude because of a VPS glitch, I can just cancel it wiht Sports mode. This would be the case when flying at height over the water for instance.

What I did not consider is the case when an autoland goes haywire in the last meter down to the ground.

If you are in a sports field no dig deal. But if you are standing on a pier with water nearby and it unexpectedly starts drifting away when 2' off the ground what are you going to do?

Just happened to me and Mavic went for a swim.

My instinct was to fly out of the situation and to climb. I did not even consider using the mode switch which I have been such a proponent of. You would have to be consciously guarding this switch with a view to bang out on every landing for this to be viable.

I would definitely recommend disabling landing protection until DJI restore the abilty to cancel with throttle.
More later. Got some drying out to do.

Seriously? Oh man, that really sux... I can't believe that just happened to you. And it's pretty much exactly one of the situations I envisioned could lead to a Mavic in the water in this post #67

I see you posted over on my thread in the DJI forums. Hopefully DJI hears and listens to our feedback.

 

[Andrew F]

I guess only that is is not a default.

You guessed wrong.

 

[Andrew F]

Seriously? Oh man, that really sux... I can't believe that just happened to you. And it's pretty much exactly one of the situations I envisioned could lead to a Mavic in the water in this post #67

I see you posted over on my thread in the DJI forums. Hopefully DJI hears and listens to our feedback.

That's exactly what is going to happen in .05 repeatedly if DJI doesn't return abort option.

 

[Robbyg]

I thought you guys had a bit of a bee in your bonnet about this issue until just now. Well I was wrong.

I thought that when it starts spuriously auto-landing from altitude because of a VPS glitch, I can just cancel it wiht Sports mode. This would be the case when flying at height over the water for instance.

What I did not consider is the case when an autoland goes haywire in the last meter down to the ground.

If you are in a sports field no dig deal. But if you are standing on a pier with water nearby and it unexpectedly starts drifting away when 2' off the ground what are you going to do?

Just happened to me and Mavic went for a swim.

My instinct was to fly out of the situation and to climb. I did not even consider using the mode switch which I have been such a proponent of. You would have to be consciously guarding this switch with a view to bang out on every landing for this to be viable.

I would definitely recommend disabling landing protection until DJI restore the abilty to cancel with throttle.
More later. Got some drying out to do.

Oh Snap, so sorry to hear about that. Was this fresh water or salt water? My suggestion is to pull the mavic apart and flush all the components with distilled water and then dry them out with a hair dryer. Make sure to clean out the snap on connectors and dry them well. Then shove the disassemble mavic in some rice for a day or two to get rid of any water you might have missed. Get new Thermal paste and reassemble her back together and hopefully it should be good to go.

Rob

 

[Robbyg]

Also, there is nothing of value between .04 and .05 that I can find. Why not keep the easier and more logical Functionality? My use is likely atypical as I'm over water the vast majority of the time.

Yep, that's why I went back to .400 and 4.04
I think the only feature that might have had some use for in .500 was the voice recording, but I just realized the other day, Duh just turn on the iPhones voice recorder app and I can sync back the audio track in post.

Rob

 

[Kilrah]

Because when I'm out over the ocean it's easy to lose perspective and I like the extra protection from accidentally setting it down.

I do not think it will prevent you from landing on water at all. Firstly the recommendation to disable over water, and the fact that on the opposite it's confirmed it actually CAUSED several water landings shows that's not a case where it's intended/going to protect against.

 

[Robbyg]

Thats true but we have also had several guys bust up their gimbal by landing on uneven ground with rocks.

The fix is that we need the landing protection and the stick to work. I agree with the other poster that this might not happen because DJI is probably freaking out over the Canada situation and the China Situation and is probably working the programmers to death on NFZ updates.

Rob

 

[Strafe1]

If you can see the landing spot of your Mav, and your happy with it, is there any reason to have Landing Protection on?

 

[JLane]

If you can see the landing spot of your Mav, and your happy with it, is there any reason to have Landing Protection on?

I actually feel like "Landing Protection" isn't the best description of what the option enables/disables, or at least not a complete description. Aside from giving the Mavic a chance to analyze the landing spot, you can also think of it as downwards Obstacle Avoidance, because what it really does it try to always maintain a height of 0.5m above the ground or obstacles. That includes causing the Mavic to rise up if you fly it too close over the top of something. Some people might like the extra protection that this downward OA provides, as it theoretically prevents you from accidentally flying the Mavic into the ground if you make a mistake. However, as we can see, it does have its drawbacks, and it's not foolproof. It's also quite dangerous if you're flying in tight spaces with objects or structures overhead. The Mavic has no way to sense what's above it, so it will happily rise up into some branches if it senses something below it.

ps. I have no idea how it actually decides what is a "safe" spot or not to land. I've heard weird stories where the Mavic won't land on a perfectly good spot in a carpark, for example.

 

[Logger]

Oh Snap, so sorry to hear about that. Was this fresh water or salt water? My suggestion is to pull the mavic apart and flush all the components with distilled water and then dry them out with a hair dryer. Make sure to clean out the snap on connectors and dry them well. Then shove the disassemble mavic in some rice for a day or two to get rid of any water you might have missed. Get new Thermal paste and reassemble her back together and hopefully it should be good to go.

Rob

Salt water unfortunately. So even if I get it totally dried and cleaned it is probably doomed.
I put it in a very dry environment (5% humidity for 8 hours while breezing air through it. Gimbal had jumped out of one of the brackets when it crashed into seawall on river bank. Re seated it properly beneath the tangs. 24 hours later with a view to pull the DAT I turned it on. Fired up OK apart from no life at all from gimbal/camera. Was able to pull DAT file which is good. DJI Go connected and showed several VPS sensor messages which went away after reboot. But still hard camera sensor error.
On initial run fan and motors and LEDs, ulstrasound clicker were all working fine, but now no sign of fan coming on. Also Not consistently connecting to RC or even DJI Assist. So powering on may have done some damage.
Will be a Care Assist or a Warranty job if I am lucky.

 

[Logger]

What Happened

The whole indecent took about 3 seconds from transitioning into what appeared to be a perfectly normal autoland about a meter in front of me at 0.5m AGL until it crashed into the river 3m away. When I sensed Autoland had commenced at 0.5m (mavic appeared in a rock solid stable hover and descending vertically) I zeroed the throttle expecting it would continue its usual vertical descent and land at my feet. What could possibly go wrong! Then with no control input from me at it started flying away at 0.9 m/s toward the river that was less than 2m away. River bank here is like standing on a wharf with a near vertical drop of ~1m down to water.
I can see I pitched forward to increase speed away from me (seems counter intuitive) then back to reverse it prior to finally applying full up throttle as a last ditch attempt to climb. After the initial startle response to it flying away, I recall my reaction was to see the clear area and drop off to the water, beyond the edge of the wharf as safety. Big mistake. I sensed it would land right on the wharf edge and topple off if I did nothing. So I pitched forward to successfully clear the lip gaining the extra meter of height above the water below. I then realised I did not actually want to be over the water and reversed the control to bring it back towards me. It was about this instant I realised I was not flying this thing and stuck in a one way autoland. I sensed it was going into the drink regardless of what I did and it might as well be against the seawall where I could retrieve it rather than out in the middle of the river. I applied full up throttle when it was still 0.3VPS above the water. But there was only ever one way this could end with this firmware. That all took 3 seconds!

My mistake was to push forward stick to clear the sea wall. Had I pulled back it may have stopped in time and landed right on the edge. It may not have done so in time either.

Had the Mavic not spontaneously headed off from a stand still at 0.9ms mid autoland I would not have had this event. Had the upthrottle command worked I would have probably have flown out of trouble too.



Where the vertical reference mark is you can see purple velocity line is ~0.9m/s and I have no elevator or aileron applied, having just released the down throttle that triggered autoland. Over to the right you can see where the vpsHeight jumps up to 0.7m as I am now clear of the obstacle and over the water.

Why did it drift away from a stable hover is my question?

 

[BudWalker]

What Happened

The whole indecent took about 3 seconds from transitioning into what appeared to be a perfectly normal autoland about a meter in front of me at 0.5m AGL until it crashed into the river 3m away. When I sensed Autoland had commenced at 0.5m (mavic appeared in a rock solid stable hover and descending vertically) I zeroed the throttle expecting it would continue its usual vertical descent and land at my feet. What could possibly go wrong! Then with no control input from me at it started flying away at 0.9 m/s toward the river that was less than 2m away. River bank here is like standing on a wharf with a near vertical drop of ~1m down to water.
I can see I pitched forward to increase speed away from me (seems counter intuitive) then back to reverse it prior to finally applying full up throttle as a last ditch attempt to climb. After the initial startle response to it flying away, I recall my reaction was to see the clear area and drop off to the water, beyond the edge of the wharf as safety. Big mistake. I sensed it would land right on the wharf edge and topple off if I did nothing. So I pitched forward to successfully clear the lip gaining the extra meter of height above the water below. I then realised I did not actually want to be over the water and reversed the control to bring it back towards me. It was about this instant I realised I was not flying this thing and stuck in a one way autoland. I sensed it was going into the drink regardless of what I did and it might as well be against the seawall where I could retrieve it rather than out in the middle of the river. I applied full up throttle when it was still 0.3VPS above the water. But there was only ever one way this could end with this firmware. That all took 3 seconds!

My mistake was to push forward stick to clear the sea wall. Had I pulled back it may have stopped in time and landed right on the edge. It may not have done so in time either.

Had the Mavic not spontaneously headed off from a stand still at 0.9ms mid autoland I would not have had this event. Had the upthrottle command worked I would have probably have flown out of trouble too.

View attachment 9118

Where the vertical reference mark is you can see purple velocity line is ~0.9m/s and I have no elevator or aileron applied, having just released the down throttle that triggered autoland. Over to the right you can see where the vpsHeight jumps up to 0.7m as I am now clear of the obstacle and over the water.

Why did it drift away from a stable hover is my question?

It looks to me like FORCE_LANDING was cancelled when full throttle was applied at about 67.4. I wish the guy that implemented CsvView would put more ticks on the time axis!!. Can you Dropbox the .DAT so that I can have a closer loo?. Normally I would just ask that you do an Export Workspace and attach the .zip but this incident may help in digging more out of the .DAT.

 

[msinger]

I wish the guy that implemented CsvView would put more ticks on the time axis!!

He is quite the slacker...

 

[Logger]

I think it might have actually been crashing into the seawall (accelX spike) more than the full throttle that cancelled it.
Will pm the link.

 

[Logger]

It just occurred to me I had a very similar lateral fly off a couple of weeks ago. I posted the video and log of it over in the Precision landing thread at Post #39 If you watch the video there you will see how quickly it flies off for no apparent reason with no control input.