Newbie Crash question

[Opieos]

Hey gang

I sadly joined the crash club 2 weeks ago and I'm trying to wrap my head around how it happened.

I was flying under a pier between pilings on the ocean and was about 5 to 10 feet off the beach with waves coming in. I had already done the same flight 5 or so times without issues. On the "final flight" I was pretty certain I cleared the other side of the pier but was distracted by a person on the beach approaching at which point I lost sight of the drone. As I stepped around one of the pilings to look for the drone I heard a dreaded clank, clank, clank, and then watched as the Mavic Air 2 splashed into about a foot of water. I was able to rescue it and subsequently ship it off to DJI Care Refresh for a replacement.

My question is would the waves have caused crash avoidance to increase the altitude of the drone? I'm 99.9% sure I didn't hit the left stick up though I fully understand more than likely a large % of the crash is user error as I lost sight of the drone.

Any thoughts or suggestions greatly appreciated

Btw for dramatic effect, I was able to save the video

T.I.A.

Opieos

 

[Meta4]

My question is would the waves have caused crash avoidance to increase the altitude of the drone? I'm 99.9% sure I didn't hit the right stick up though I fully understand more than likely a large % of the crash is user error as I lost sight of the drone.

If you want answers, you need to provide flight data.

 

[Opieos]

Not sure if this works or not

 

[slup]

As this flight was conducted with the DJI FLY app with a version above 1.2.1 (it was v1.2.5) we can't use anything else than the Airdata.com conversion that you've linked to in order to analyze this ... which isn't so exact unfortunately.

But seeing your video with all the rolling waves close under the AC (according to the log approx 8,5ft below) ... this from the manual strikes me immediately ...


The GPS lock & quality was good there with 14-16 satellites with a navhealth 5 of 5 ... & you didn't move any of the 2 sticks at all when this happened either.

Have below extracted some data from Airdata's .CSV file & put it into excel ...

The left axis is feet or mph ... the right is degrees for the AC pitch & roll.

As seen by the blue graph with readings from the barometric sensor the height is rather constant ... the orange, the VPS height over ground on the other hand, shows different heights ... most probably readings from the rolling waves. At 402,9sec into the flight the grey graph suddenly starts to show values for the vertical speed ... they are positive, meaning descending. This most probably tricked the flight controller into believe that the AC was descending so it countered by command ascending. At 403,1 the AC hits the pier the first time...

So the take away from this, is that it's no good to stay hovering over rapidly moving water ... close under a pier where a bit of height instability can mean that the AC can hit solid objects.

 

[Opieos]

Thank you. I think with the distraction and loss of sight I left it at a hover. Certainly, lessons learned!!!!

 

[vindibona1]

There is one more aspect that may or may not be part of the equation but I think is worth discussion: Wind reflection from the drone's own wind force.

Have you ever flown your drone down a narrow hallway just trying to keep it straight down the middle? If might be more challenging than you think as the air flow from the drone hits the floors and has nowhere to go but up the walls and comes right back at the drone. I've found that these closed spaces create quite the challenge to keep the drone going straight forward where there the ambient air is static, but the drone's own wind can push it to one side or another. While the bottom sensor may have come into play in the OP's scenario, is it possible that a sudden swell could provide enough resistance to the air flow, providing upward force against that air stream to push the drone up into the top of the pier?

 

[slup]

There is one more aspect that may or may not be part of the equation but I think is worth discussion: Wind reflection from the drone's own wind force.

Have you ever flown your drone down a narrow hallway just trying to keep it straight down the middle? If might be more challenging than you think as the air flow from the drone hits the floors and has nowhere to go but up the walls and comes right back at the drone. I've found that these closed spaces create quite the challenge to keep the drone going straight forward where there the ambient air is static, but the drone's own wind can push it to one side or another. While the bottom sensor may have come into play in the OP's scenario, is it possible that a sudden swell could provide enough resistance to the air flow, providing upward force against that air stream to push the drone up into the top of the pier?

If that have been the case we hadn't seen a logged uncommanded downward vertical speed (which the IMU decided to act on by ascending) together with a constant barometric height measurement.

Have seen similar IMU mistakes before where it goes & acts towards one data point (the vertical downward speed here) even though 2 other data points tells a different story.

If enough space had been available this had most probably only been a slightly surprising height increase ... easily addressed to the waves close beneath ... this time it didn't end well.

 

[Darkside Drone]

Hey gang

I sadly joined the crash club 2 weeks ago and I'm trying to wrap my head around how it happened.

I was flying under a pier between pilings on the ocean and was about 5 to 10 feet off the beach with waves coming in. I had already done the same flight 5 or so times without issues. On the "final flight" I was pretty certain I cleared the other side of the pier but was distracted by a person on the beach approaching at which point I lost sight of the drone. As I stepped around one of the pilings to look for the drone I heard a dreaded clank, clank, clank, and then watched as the Mavic Air 2 splashed into about a foot of water. I was able to rescue it and subsequently ship it off to DJI Care Refresh for a replacement.

My question is would the waves have caused crash avoidance to increase the altitude of the drone? I'm 99.9% sure I didn't hit the left stick up though I fully understand more than likely a large % of the crash is user error as I lost sight of the drone.

Any thoughts or suggestions greatly appreciated

Btw for dramatic effect, I was able to save the video

T.I.A.

Opieos

Poor fella looks like it tried to climb out of the water! I'll be following this. I crash ALL. THE. TIME. (when I try to fly around trees, anyway) and have often wondered if the crash avoidance settings actually caused some of the wrecks. (obviously, pilot error is my biggest problem! but settings can help.)

 

[Darkside Drone]

There is one more aspect that may or may not be part of the equation but I think is worth discussion: Wind reflection from the drone's own wind force.

Have you ever flown your drone down a narrow hallway just trying to keep it straight down the middle? If might be more challenging than you think as the air flow from the drone hits the floors and has nowhere to go but up the walls and comes right back at the drone. I've found that these closed spaces create quite the challenge to keep the drone going straight forward where there the ambient air is static, but the drone's own wind can push it to one side or another. While the bottom sensor may have come into play in the OP's scenario, is it possible that a sudden swell could provide enough resistance to the air flow, providing upward force against that air stream to push the drone up into the top of the pier?

That explains some things! I was trying to follow hubby's '69 Beetle driving down a scenic road that's got this pretty tree canopy overhead. Wondered what on earth was happening because it could NOT fly straight. (No whisky involved)

 

[vindibona1]

That explains some things! I was trying to follow hubby's '69 Beetle driving down a scenic road that's got this pretty tree canopy overhead. Wondered what on earth was happening because it could NOT fly straight. (No whisky involved)

First off, as I read your remarks the first time I was wondering about the flying '69 Beetle as you may have been following him in your Jetta . And then I realized you were chasing him with your drone. No drugs this morning, I promise.

I learned a few things from my hallway exploits. Heavier drones are harder to push around and their stronger motors can help keep them stable. Also going faster provides some consistency in the airflow anomaly that you can counter with the sticks. Can you explain a little more on how the tree canopy might have affected your flight vector (or were you just joking)?

 

[Darkside Drone]

First off, as I read your remarks the first time I was wondering about the flying '69 Beetle as you may have been following him in your Jetta . And then I realized you were chasing him with your drone. No drugs this morning, I promise.

I learned a few things from my hallway exploits. Heavier drones are harder to push around and their stronger motors can help keep them stable. Also going faster provides some consistency in the airflow anomaly that you can counter with the sticks. Can you explain a little more on how the tree canopy might have affected your flight vector (or were you just joking)?

lol yup - drone stalking I'm not sure that the tree canopy impacted it, but here's what happened. I wanted to do the active follow, but the message came up that the drone needed to go much higher than about 10 feet. That would put it in the trees, so nope. I tried doing a manual follow and found that the drone felt out of control. It wouldn't fly straight (side to side swooping!) and seemed to be hyper-sensitive to the tree limbs above it, almost like all my prior crashes had left it traumatized. lol

 

[slup]

lol yup - drone stalking I'm not sure that the tree canopy impacted it, but here's what happened. I wanted to do the active follow, but the message came up that the drone needed to go much higher than about 10 feet. That would put it in the trees, so nope. I tried doing a manual follow and found that the drone felt out of control. It wouldn't fly straight (side to side swooping!) and seemed to be hyper-sensitive to the tree limbs above it, almost like all my prior crashes had left it traumatized. lol

Never thought about that the tree canopy could have shielded out the sky so...

1. the GPS lock quickly went bad & your drone only could...
2. rely on the VPS sensor to keep it steady horizontally...
3. which perhaps also became bad with a bit to far distance to ground or/and a bit to little light...
4. & then leaving your craft in ATTI mode totally without any horizontal hold or breaking at all?