Mavic flies into side wall of bridge, uncontrollable whilst 20m away in direct vision. DJI takes no responsibility

[aquinox]

Hi,
My DJI died whilst flying under a bridge. it moved itself to the right instead of to the front, ignoring my commands, also ignoring my correction (Full throttle) to the left. whilst being very close to the drone. In all other cases it did not pose a problem to fly under a bridge in an urban environment nor a narrow tunnel without gps with strong wind effects, as can be seen here.
This is inconsistent behavior at best and crappy programming at worst. Pilot control should be leading at all times, and if GPS loss was such a problem, it could have defered my vector (direction+speed) from just before losing gps. This is my 4th mavic lost to their programming. The 3rd randomly took a plunge whilst stationary (reimbursed), the 2nd was losing connection within 400m range and not returning, the first suicided into a tree without my command right after takeoff.

Are there any alternatives out there with similar camera quality and portability as the m2p which is not DJI?

I have a video. showing my controls at the accident, an earlier, similar accident, and cases where the mavic was functioning normally.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

Attached are the flight logs of my last flight + the flight before where it did not return to home when it was supposed to but went moving further from home.

Zoolz | Share | Revolutionizing business cloud backup

Zoolz is a comprehensive business solution for backing up, archiving, data management and collaboration

cloud3.zoolz.com

 

[Meta4]

There's a lot to flying a drone and reading your story and your flight data, it seems that your understanding is very confused.
That may be related to some of the "bad luck" youve experienced.

Attached are the flight logs of my last flight + the flight before where it did not return to home when it was supposed to but went moving further from home.

Looking at your RTH flight data, your drone didn't mover further away from home at all.
Here's a summary of what happened:
You lost signal at 14:32.9 with the drone 1199 metres away.
RTH initiated and when signal returned at 15:16.2, the drone had started coming home and was making 11.9 metres/sec in the right direction.
It got back to 897 metres away and slowed around 15:35.6, probably because obstacle avoidance was upset from a reflection from snow or the sun shining on the sensors ?
At 15:37.9 you started doing odd things with the joysticks - it's hard to work out what you were doing.
I can't see that you were following any strategy, it looks like you just went crazy with the sticks.
But with the drone facing a different direction, RTH continued to bring the drone closer to home.
RTH brought it back to 538 metres from home before you cancelled RTH and flew the drone home manually.

 

[Meta4]

Hi,
My DJI died whilst flying under a bridge. it moved itself to the right instead of to the front, ignoring my commands, also ignoring my correction (Full throttle) to the left. whilst being very close to the drone. In all other cases it did not pose a problem to fly under a bridge in an urban environment nor a narrow tunnel without gps with strong wind effects, as can be seen here.

Flying without GPS means your drone loses horizontal position holding (unless the lighting is good enough and the ground surface is suitable to allow the VPS to stabilise the drone).
In this incident the VPS was unable to help so the drone could not hold position so the drone was in atti mode.
It couldn't stop when you let go of the sticks and would continue drifting.
Flying in atti mode close to obstacles is always risky and requires very careful piloting and plenty of experience.

Rather than careful slow control inputs, I see sudden, full-stick moves with you making full elevator and aileron inputs.
The pitch, roll and altitude data indicate the crash happened at 8:36.9.
With obstacles very close and inexperience in atti flying, it's no wander that you crashed and are confused about why.
It's also not a surprise the DJI don't recognise this as a warranty claim.

This is inconsistent behavior at best and crappy programming at worst. Pilot control should be leading at all times, and if GPS loss was such a problem, it could have defered my vector (direction+speed) from just before losing gps.

Sorry, but it looks like piloting issues caused this incident and it had nothing to do with DJI's programming.
I'd suggest learning a lot more about what can go wrong and how to make sure it doesn't before risky flying like this.
Without GPS, there's nothing DJI's programming can do to help you.

This is my 4th mavic lost to their programming. The 3rd randomly took a plunge whilst stationary (reimbursed), the 2nd was losing connection within 400m range and not returning, the first suicided into a tree without my command right after takeoff.

Are there any alternatives out there with similar camera quality and portability as the m2p which is not DJI?

Similar quality, but not DJI??
No, but I suspect the issues you've run into probably weren't the fault of the drones.

 

[aquinox]

Flying without GPS means your drone loses horizontal position holding (unless the lighting is good enough and the ground surface is suitable to allow the VPS to stabilise the drone).
In this incident the VPS was unable to help so the drone could not hold position so the drone was in atti mode.
It couldn't stop when you let go of the sticks and would continue drifting.
Flying in atti mode close to obstacles is always risky and requires very careful piloting and plenty of experience.

Rather than careful slow control inputs, I see sudden, full-stick moves with you making full elevator and aileron inputs.
The pitch, roll and altitude data indicate the crash happened at 8:36.9.
With obstacles very close and inexperience in atti flying, it's no wander that you crashed and are confused about why.
It's also not a surprise the DJI don't recognise this as a warranty claim.


Sorry, but it looks like piloting issues caused this incident and it had nothing to do with DJI's programming.
I'd suggest learning a lot more about what can go wrong and how to make sure it doesn't before risky flying like this.
Without GPS, there's nothing DJI's programming can do to help you.

Similar quality, but not DJI??
No, but I suspect the issues you've run into probably weren't the fault of the drones.

Doesn't explain why it went perfect before when it was pitch black in the narrow tunnel in the video. Did you actually check it? In this case I moved slowly forward and as i noticed it moving to the right against my will i moved it to the left slowly then strongly.

 

[Meta4]

Doesn't explain why it went perfect before when it was pitch black in the narrow tunnel in the video. Did you actually check it? I moved slowly forward and as i noticed it moving to the right against my will i moved it to the left slowly then strongly.

I didn't look at the video because video doesn't prove anything.
I look at your joystick inputs and the data from the drone's sensors to work out what was going on.

 

[RayOZ]

You seem to like flying close quarters. Reversing from the tunnel and almost brushing against foliage. Seems like an accident waiting to happen.

 

[aquinox]

I didn't look at the video because video doesn't prove anything.
I look at your joystick inputs and the data from the drone's sensors to work out what was going on.

well two of your assumptions were already wrong so i recommend watching, as the drones sensors were messed and RTH actually continued its path further from home...

 

[aquinox]

You seem to like flying close quarters. Reversing from the tunnel and almost brushing against foliage. Seems like an accident waiting to happen.

Yes. It was just to prove that in the instance it crashed it doesn't behave normally, it is not about my piloting which is slow and secure, as it could handle that close quarters without problems.

 

[Kilrah]

It can handle close quarters well as long as everything is perfect. But things can get confused at anytime requiring you to maually take the appropriate actions, so if you're not ready for that maybe you should skip doing that.

 

[aquinox]

It can handle close quarters well as long as everything is perfect. But things can get confused at anytime requiring you to maually take the appropriate actions, so if you're not ready for that maybe you should skip doing that.

I was ready for it, it wasn't listening as you can see in the video.

 

[StantonZ]

1st time lost...could be a "rookie" mistake.
2nd time lost...some folks learn the hard way.
3rd time lost...there's definitely a pattern developing.
4th time lost...time to look in the mirror.

 

[Aerial Edge]

This thread will be interesting. The logs do not lie, they are data plain as day. Having Hundreds of hours on DJI drones with people I work with having Thousands of hours. They will not do random things uncommanded. Certain responses will be made due to input, or loss of sensors.

I have a feeling I know where this one is heading.

 

[Prismatic]

... I have a feeling I know where this one is heading.

?

 

[Flycaster]

well two of your assumptions were already wrong so i recommend watching, as the drones sensors were messed and RTH actually continued its path further from home...

What was the diagnosis from DJI when they saw the logs you sent to them regarding this "event"?
You did send in the logs to DJI, correct?
What was their actual response?

 

[DanMan32]

"if GPS loss was such a problem, it could have defered my vector (direction+speed) from just before losing gps."

The drone can't predict that it's being flown into an area that has poor GPS signal. GPS signal change tends to be sudden, just like needing to remove your sunglasses would be sudden when going from the open sunny outdoors into a dark tunnel.

 

[aquinox]

What was the diagnosis from DJI when they saw the logs you sent to them regarding this "event"?
You did send in the logs to DJI, correct?
What was their actual response?

Time Point: 08:37 Effective Control: ☒Yes ☐No
Directional Error: ☐Yes ☒No
Control Mode: ☒GPS Mode ☒ATTI Mode ☐OPTI Mode ☐Sport Mode 
Accident Type: I ☐ A. Pilot fault in Positioning Mode ☐ B. Pilot Fault in ATTI Mode ☐ C. Pilot fault when using intelligent flight ☐ D. Pilot fault when using App ☐ E. Aircraft crashed with obstacles when RTH ☐ F. Strong wind caused accident ☐ G. Add third-party component caused accident ☐ H. CSC operation in mid-air ☒ I. Other Aircraft crashed in weak GPS signal environment

I dispute the effective control since it went right while i steered left, also strange that i was hovering near the other end of the bridge and gps was still lost...

 

[Flycaster]

At first blush look at your provided log, looks like a diminishing GPS sat count to me.
One has to remember, even heavy brush/trees can reduce the GPS signal reception...

 

[slup]

Well ... you seems to be a guy that like to take non knowledge based risks without actually knowing or care how close you are/were to disaster ... only in this flight you for instance was up on 225m height over a heavily populated area.

You flew here when you lost the AC ... not much of the sky visible even outside that bridge down there close to the channel.



Already outside the bridge the sat count & the navhealth was dropping dramatically ... the AC fell back shortly to ATTI but then ended up in a "pseudo" GPS mode with a very low sat count & navhelth ... totally unreliable regarding horizontal position hold & certainly with a mushy stick control (common in instances like this). The crash happens where the markers is placed in the chart ... there the navhealth is 0 with 3 satellites... the VPS sensor is reporting height but most probably very unreliable as no switch is made to OPTI mode.

The red is navhealth ... anything from 3 & below is considered by the flight controller as unreliable
The bright green is the sat count
The purple is the measured height above ground by the VPS sensor
The blue is AC pitch (to mark the impact)
The dark green is AC roll (to mark the impact)
The blue background stripe is ATTI mode ... the rest, white is GPS mode with various quality



Furthermore the OD sensors made it even more unreliable, regarding stick control, as they reported braking moments both vertically & horizontally just there in the end


This wasn't anything other than a pilot error ... you pushed your luck one time too many.

And then a short comment (pure speculation as I don't have the log) regarding you vid & what's looking as a uncontrolled flight close to the trees (is it there you claim that the RTH didn't work?) ... really looks like a yaw error coming from a power on in close proximity to magnetic disturbance ... if so then again a pilot error.

 

[Meta4]

well two of your assumptions were already wrong so i recommend watching, as the drones sensors were messed and RTH actually continued its path further from home...

Please point out what assumptions you think I made (I'm not sure what you are referring to) and how they are incorrect.

 

[aquinox]

Well ... you seems to be a guy that like to take non knowledge based risks without actually knowing or care how close you are/were to disaster ... only in this flight you for instance was up on 225m height over a heavily populated area.

You flew here when you lost the AC ... not much of the sky visible even outside that bridge down there close to the channel.

View attachment 125087

Already outside the bridge the sat count & the navhealth was dropping dramatically ... the AC fell back shortly to ATTI but then ended up in a "pseudo" GPS mode with a very low sat count & navhelth ... totally unreliable regarding horizontal position hold & certainly with a mushy stick control (common in instances like this). The crash happens where the markers is placed in the chart ... there the navhealth is 0 with 3 satellites... the VPS sensor is reporting height but most probably very unreliable as no switch is made to OPTI mode.

The red is navhealth ... anything from 3 & below is considered by the flight controller as unreliable
The bright green is the sat count
The purple is the measured height above ground by the VPS sensor
The blue is AC pitch (to mark the impact)
The dark green is AC roll (to mark the impact)
The blue background stripe is ATTI mode ... the rest, white is GPS mode with various quality

View attachment 125089

Furthermore the OD sensors made it even more unreliable, regarding stick control, as they reported braking moments both vertically & horizontally just there in the end

View attachment 125090
This wasn't anything other than a pilot error ... you pushed your luck one time too many.

And then a short comment (pure speculation as I don't have the log) regarding you vid & what's looking as a uncontrolled flight close to the trees (is it there you claim that the RTH didn't work?) ... really looks like a yaw error coming from a power on in close proximity to magnetic disturbance ... if so then again a pilot error.

Lol, DJI reimbursed that tree hugging drone as it had compass failure. I don't know what this toxic vibe here is assuming that technology is perfect But yeah, false obstacle detection probably blocked my controls at the crucial moments, as it sometimes also did in mid air. when there were no birds around