I recently tried to fly my MA from the deck of my boat sitting on my mooring, something I've done many times with less expensive drones.  After lifting off instead of hovering in place it started wandering off to the side. I immediately tried to get it back on the boat but managed to clip the side of the boat and my MA went to the bottom of the harbor.  Question: Does the air gyroscopes calibrate on power up and get thrown off if they are sitting on a moving platform?  The boat on the mooring has some side to side drift due to tide and any breeze.  If this is a problem is there some way to counter it?  BTW I recovered the drone with scuba gear after it sat on the bottom after an hour.  

Hi,

Sorry that you lost your MA like this. Only your flightlog can tell you and us more about the cause of this 'accident '.

If you like upload your flightlog and post the link here. [ link:  http://www.phantomhelp.com/LogViewer/Upload/  ]

cheers
JJB

Whether it was low gps or compass problems caused by magnetic interference on the boat, it will show up in your log.

Hi there, I'm really sorry to hear that. Kindly send the drone to us so that we can undergo damage assessment. Our engineers will explain to you why the incident happened. Please contact us using this link:https://www.dji.com/support?site=brandsite&from=navs

To me its not a good Idea to launch off a moving boat .  To me the wind would be a huge factor in controling the drone.

i had exactly thr same problem in bermuda , mavic air at bottom of the sea

makker01 Posted at 2018-7-7 09:11
i had exactly thr same problem in bermuda , mavic air at bottom of the sea

look at this youtube video  same issue

i have some experience flying from moving boats and vehicals..   it sounds like you panicked when it didn't react the way you expected.   remember, 95% of crashes come from human error, probably more like 99%.  when you are returning back, always return to your hands, directly!  do not try to return it to take off location or to the boat or car..but directly to you.  

then thing you have to keep in mind, every drone reacts a bit differently to such situations, it is not expecting to take off from a moving platform, basically initializing its systems.. then its taking off and like WHOA!.  and every drone has to suddenly compensate using IMUs/VPS and GPS and some drones are better at figuring things out smoothly than others... dont panic it will figure it out way faster then you will, just trust its computers.  and when returning back, return it to your hands..always catch it...

=

Not a total loss, you got a scuba dive in!   As JJBspark point out, need FlightLog file to look at for possible causes.  As for flying from a boat that is moving slightly while moored, I would think all it would do is slightly slow-up the initial acquirring GPS position fix.

Or buy an RC sub like Dustin Dunhill who is using it to find his drowned Mavic Pro.

HereForTheBeer Posted at 2018-7-7 09:23
i have some experience flying from moving boats and vehicals..   it sounds like you panicked when it didn't react the way you expected.   remember, 95% of crashes come from human error, probably more like 99%.  when you are returning back, always return to your hands, directly!  do not try to return it to take off location or to the boat or car..but directly to you.  

then thing you have to keep in mind, every drone reacts a bit differently to such situations, it is not expecting to take off from a moving platform, basically initializing its systems.. then its taking off and like WHOA!.  and every drone has to suddenly compensate using IMUs/VPS and GPS and some drones are better at figuring things out smoothly than others... dont panic it will figure it out way faster then you will, just trust its computers.  and when returning back, return it to your hands..always catch it...

http://www.phantomhelp.com/LogViewer/5TRM2MZ714LB1BKYM18Q/

my log showing loss was unable to control my mavic air

makker01 Posted at 2018-7-7 09:43
http://www.phantomhelp.com/LogViewer/5TRM2MZ714LB1BKYM18Q/

my log showing loss was unable to control my mavic air

oh, well i only have a mavic pro which is very stable at taking off from moving platforms.  maybe the mavic air is less stable and less able to compensate.

looks like however your VPS was causing issues for you coming in and out and disagreeing with your IMU, possible the water tricking it.

HereForTheBeer Posted at 2018-7-7 09:48
oh, well i only have a mavic pro which is very stable at taking off from moving platforms.  maybe the mavic air is less stable and less able to compensate.

looks like however your VPS was causing issues for you coming in and out and disagreeing with your IMU, possible the water tricking it.[view_image]

so pilot error ??

makker01 Posted at 2018-7-7 09:57
so pilot error ??

potentially, i dont have a mavic air to know how it reacts and behaves to know if you could have done much to saved it.. some drones like the spark, it can sometimes get completely tricked by the VPS seeing wrong information and it can't be saved.

what i would have tried to have done would been to rocket the drone up in the air 20-30 feet to break away from VPS causing the instability.. if at all possible

HereForTheBeer Posted at 2018-7-7 10:03
potentially, i dont have a mavic air to know how it reacts and behaves to know if you could have done much to saved it.. some drones like the spark, it can sometimes get completely tricked by the VPS seeing wrong information and it can't be saved.

what i would have tried to have done would been to rocket the drone up in the air 20-30 feet to break away from VPS causing the instability.. if at all possible

had  no control using controller

makker01 Posted at 2018-7-7 09:43
http://www.phantomhelp.com/LogViewer/5TRM2MZ714LB1BKYM18Q/

my log showing loss was unable to control my mavic air

Your log shows very little, the VPs alt and IMU alt, is always like this, so wouldn’t worry about this, you were right on the boarder of good gps but you should be ok there. When you took off auto, I think because surface under the aircraft may have been moving and bobbing, and if surface wasn’t well defined this could have caused VPs not to work at its optimum, so aircraft moving about as you said and eventually crashing. Boats are extremely hard both taking off and landing, best practice for this is to take off from hand using sombody to help, and the same when landing, always have someone to catch while you control.

Was it pilot error. I’m not sure here, and I have seen similar get a warranty claim and also been refused. I would follow the necessary steps to go through the warranty procedure and see what dji come up with.

I wish you luck.

hallmark007 Posted at 2018-7-7 10:12
Your log shows very little, the VPs alt and IMU alt, is always like this, so wouldn’t worry about this, you were right on the boarder of good gps but you should be ok there. When you took off auto, I think because surface under the aircraft may have been moving and bobbing, and if surface wasn’t well defined this could have caused VPs not to work at its optimum, so aircraft moving about as you said and eventually crashing. Boats are extremely hard both taking off and landing, best practice for this is to take off from hand using sombody to help, and the same when landing, always have someone to catch while you control.

Was it pilot error. I’m not sure here, and I have seen similar get a warranty claim and also been refused. I would follow the necessary steps to go through the warranty procedure and see what dji come up with.

he mentioned no control  using the controller..  so it reacted more like a spark with one of the systems completely overriding things.

HereForTheBeer Posted at 2018-7-7 10:15
he mentioned no control  using the controller..  so it reacted more like a spark with one of the systems completely overriding things.

You need to move sticks to have any control, his log shows absolutely no stick movements.

HereForTheBeer Posted at 2018-7-7 10:15
he mentioned no control  using the controller..  so it reacted more like a spark with one of the systems completely overriding things.

You need to move sticks to have any control, his log shows absolutely no stick movements.

hallmark007 Posted at 2018-7-7 10:22
You need to move sticks to have any control, his log shows absolutely no stick movements.

i dont know...   would there be any reason why stick inputs wouldn't get logged if the drone was too busy being lost..?    im no rocket surgeon, just like giving people benift of doubt.

HereForTheBeer Posted at 2018-7-7 10:24
i dont know...   would there be any reason why stick inputs wouldn't get logged if the drone was too busy being lost..?    im no rocket surgeon, just like giving people benift of doubt.

Only if the were not pushed, you can see in his log craft moving over and back , I presume it hit something,pretty close and fell into the drink, it all happened in a matter of a few seconds.

hallmark007 Posted at 2018-7-7 10:29
Only if the were not pushed, you can see in his log craft moving over and back , I presume it hit something,pretty close and fell into the drink, it all happened in a matter of a few seconds.

yes was franticly trying to move the sticks  - hit the side and tried to stabilise itself and fell in !!

DJI Paladin Posted at 2018-7-7 05:40
Hi there, I'm really sorry to hear that. Kindly send the drone to us so that we can undergo damage assessment. Our engineers will explain to you why the incident happened. Please contact us using this link:https://www.dji.com/support?site=brandsite&from=navs

Hi Thanks for the offer but I had insurance on it and already have it replaced.  I'm just trying to understand why the problem occurred.

Justaguy Posted at 2018-7-7 08:55
To me its not a good Idea to launch off a moving boat .  To me the wind would be a huge factor in controling the drone.

No, not much wind and as I said I've done this many times before with other drones.  Not difficult.

makker01 Posted at 2018-7-7 09:11
i had exactly thr same problem in bermuda , mavic air at bottom of the sea

Sorry to hear that.  Probably didn't recover it??  I have since purchased flotation for the landing legs.  Probable still kill the drone if it lands other than perfectly upright but it'll be a lot easier to recover.

HereForTheBeer Posted at 2018-7-7 09:23
i have some experience flying from moving boats and vehicals..   it sounds like you panicked when it didn't react the way you expected.   remember, 95% of crashes come from human error, probably more like 99%.  when you are returning back, always return to your hands, directly!  do not try to return it to take off location or to the boat or car..but directly to you.  

then thing you have to keep in mind, every drone reacts a bit differently to such situations, it is not expecting to take off from a moving platform, basically initializing its systems.. then its taking off and like WHOA!.  and every drone has to suddenly compensate using IMUs/VPS and GPS and some drones are better at figuring things out smoothly than others... dont panic it will figure it out way faster then you will, just trust its computers.  and when returning back, return it to your hands..always catch it...

Well I hardly panicked.  It was my fault that the drone clipped a strut on my boat but the reason that happened was that when the drone drifted sideways I had to step out onto the deck of the boat to see where it was going.  When I got the drone back close to the boat I had to look away as I stepped back to give it room to land and when I did that I allowed the drone to drift again and that's when it hit the strut.  None of this is relevant to the question I posted however.

HedgeTrimmer Posted at 2018-7-7 09:25
Not a total loss, you got a scuba dive in!   As JJBspark point out, need FlightLog file to look at for possible causes.  As for flying from a boat that is moving slightly while moored, I would think all it would do is slightly slow-up the initial acquirring GPS position fix.

True always a good day when Scuba diving 8^)  I'm working on a test to see if I can replicate the problem (from land this time) because I really want to be able to launch from the boat.  If I make progress I'll post here.  I'll also upload my log to see if there is a clue there.

A CW Posted at 2018-7-7 09:26
Or buy an RC sub like Dustin Dunhill who is using it to find his drowned Mavic Pro.

I have a remote camera, not RC, that I put over the side and played it around for half an hour before locating the drone.  Standard SCUBA underwater search pattern would have found it just as fast though.

fans5760103b Posted at 2018-7-7 12:40
I have a remote camera, not RC, that I put over the side and played it around for half an hour before locating the drone.  Standard SCUBA underwater search pattern would have found it just as fast though.

SCUBA will for sure and with greater success than Dustin bless him

fans5760103b Posted at 2018-7-7 12:35
Well I hardly panicked.  It was my fault that the drone clipped a strut on my boat but the reason that happened was that when the drone drifted sideways I had to step out onto the deck of the boat to see where it was going.  When I got the drone back close to the boat I had to look away as I stepped back to give it room to land and when I did that I allowed the drone to drift again and that's when it hit the strut.  None of this is relevant to the question I posted however.

always hand catch in coming drones on moving vehicals..... always asking for it to take a dump on a moving vehical if your trying to land it on a platform instead of in your hand.. too many physics calculations at work ..   with mavic air and spark all you have to do to stop it is grab it and tilt it 90 degrees sideways and the props will stall out....

makker01 Posted at 2018-7-7 09:57
so pilot error ??

Hi Makker,

Sorry for your loss.

I admit ; i never did takeoff from a moving platform.

I think that taking off a moving platform is different for software than taking off form a steady undergroud. Normally when you start yout rotor they do not have to correct a moving direction, so kind of flat pitch turning in helicopter language.
If you look at my chart you see that the pitch at startup is 7.6 angle, roll 2.6. This means that the drone wants to correct its position in GPS mode. Drone thinks it moves....wich is does ofcourse on a moving platform from the GPS reference postion at startup. How a Spark of MA corrects this moving-not moving flight at startup i don`t know, never done this bit agree that every type of drone will correct this is his onw way and speed.

I think you (we all i mean) should never use AutoTake off to lift the drone into hover! The moment its feets are of the ground it will compensate the moving direction etc.
Takeoff manual you will see the drone acts and react with giving corrective inputs. You said that you frantic moved the stick, have to try that out but it looks that during autotakeoff no RC inputs are possible !! [ good to know this, another reason for me to always use Manual Takeoff ]
Pilots error? depends of the definition of a pilot error. I am not here to judge, just to help if i can.

BTW at the end of this short flight VPS heigh is indicating more than before, imo beause of the high bank angle the sensors are seeing more.

EDIT ; just had a check ; you can use RC stick inputs during AutoTakeoff.
EDIT ; oops, think that the pitch up angle is beacuse of the nose up attitude as standard.
cheers
JJB

HereForTheBeer Posted at 2018-7-7 13:31
always hand catch in coming drones on moving vehicals..... always asking for it to take a dump on a moving vehical if your trying to land it on a platform instead of in your hand.. too many physics calculations at work ..   with mavic air and spark all you have to do to stop it is grab it and tilt it 90 degrees sideways and the props will stall out....

Good to know about the 90 degree tilt to stop the rotors.  When I'm talking about a "moving" vehicle here I'm talking about a barely perceptible motion.  When landing on the deck it's almost the same as landing on the ground because the motion is so slight.  That's one of the reasons I was surprised when the Air moved away after take off (Manual takeoff).  I could still control it but had to focus on compensating for the drift constantly.  When I looked away to clear deck space the Air moved sideways and clipped the boat.  I'm assuming that I released the stick and allowed it to drift for the < 2 seconds I looked away...

fans5760103b Posted at 2018-7-8 03:28
Good to know about the 90 degree tilt to stop the rotors.  When I'm talking about a "moving" vehicle here I'm talking about a barely perceptible motion.  When landing on the deck it's almost the same as landing on the ground because the motion is so slight.  That's one of the reasons I was surprised when the Air moved away after take off (Manual takeoff).  I could still control it but had to focus on compensating for the drift constantly.  When I looked away to clear deck space the Air moved sideways and clipped the boat.  I'm assuming that I released the stick and allowed it to drift for the < 2 seconds I looked away...

Hi,

If you upload your flightrecord mayby this forum can explain more what happend.

cheers
JJB

JJBspark Posted at 2018-7-8 03:37
Hi,

If you upload your flightrecord mayby this forum can explain more what happend.

I've been trying to upload the log but it hasn't been successful.  I tried uploading it in a zip file still  no luck.  Is there a different place to upload it or some other trick I need to know?

fans5760103b Posted at 2018-7-8 05:28
I've been trying to upload the log but it hasn't been successful.  I tried uploading it in a zip file still  no luck.  Is there a different place to upload it or some other trick I need to know?

Hi,

did you used this link?  http://www.phantomhelp.com/LogViewer/Upload/

cheers
JJB

JJBspark Posted at 2018-7-8 06:07
Hi,

did you used this link?  http://www.phantomhelp.com/LogViewer/Upload/

Ummm.  I just did!  Thanks.  Now how do others find it?  Do I post the link?
http://www.phantomhelp.com/LogViewer/GWNKCPV6B7GABFFXW44O/

Thanks for the pointer.

fans5760103b Posted at 2018-7-8 07:34
Ummm.  I just did!  Thanks.  Now how do others find it?  Do I post the link?
http://www.phantomhelp.com/LogViewer/GWNKCPV6B7GABFFXW44O/

No need to post a link, the link is active in your post.

JJBspark Posted at 2018-7-8 08:10
No need to post a link, the link is active in your post.

Thanks again professor JJBspark!

Not sure how useful that log is.  I didn't see much in there but maybe someone with a trained eye can spot something.

fans5760103b Posted at 2018-7-8 08:26
Thanks again professor JJBspark!

Not sure how useful that log is.  I didn't see much in there but maybe someone with a trained eye can spot something.

Hi,

can u please edit your post, i am absolutely not a professor so if you don`t mind...just call me a drone-technical-flyer-sort of specialist or something else!

cheers
JJB

fans5760103b Posted at 2018-7-8 08:26
Thanks again professor JJBspark!

Not sure how useful that log is.  I didn't see much in there but maybe someone with a trained eye can spot something.

Hi,

Had a look at your flightrecord, no errors in there.

Try to understand what i see keeping in mind what you wrote what happend.

I do not really understand what i see. The polar chart shows the position during flight, it did not 'land' at your boat, it was 29 feet away.

Is it possible to take of from a moving platform (if i am wrong others will tell their view on this), but the moment the craft is in the air in GPS mode it is 'locked' in the air so the ship is moving away form the drone. Your ship was moored, mayby the movement to one side made you think that the drone drifted away, would it be possible that craft was steady and your boat moved to aside? Did you ship drifted 29 feet away from takeoff time?

Cheers
JJB