endotherm
First Officer
Flight distance : 503241 ft
Australia
Offline
|
This is a very interesting case. It looks like you were flying on the edge of line-of-sight near buildings which would account for the poor signal and low GPS etc. As labroides notes: "6 inches past right on the edge will do it." However that didn't cause the crash. The fascinating bit is the last track. It is dead straight, but as you note, it is descending. It seems to have changed heading correctly to your approximate home position after initiating a RTH. The constant rate of descent almost seems like it is performing a land-in-place failsafe. That doesn't explain it's forward movement though and I doubt the wind was blowing that accurately to have done it. It would be interesting if you can play back your flight record and show your joystick inputs for that last leg and see if there is any response to your inputs, maybe a stick got stuck or you overrode the automatically set directions?
It is more likely that the straight line doesn't actually represent your path, rather it is joining two points where the data downlink was still working, approx 11 sec at marker "I". For all we know it did go up, then return towards home, and fell for some other reason. Where did you recover the wreckage, on the ground or on someones balcony? The top of the 20 storey building would be at about 80m elevation and you were lower than that. It's possible when the signal was lost it went into a hover, waiting for the RTH to kick in, and the wind blew it closer to the building. Tall buildings tend to have their own micro-weather systems and the wind was likely to be turbulent close to a structure like this (I really miss the estimated wind report on the free version of Healthy Drones ) It may have been a lot closer to the building than you think. All we can calculate from this data is that in 11 seconds, it has moved 84 metres towards the building and fallen 31.5 metres. In the last recorded fraction of a second, it fell around 4m. If it flew straight into the building (go-home without ascending) at 10m/s, that would leave about 2 seconds of fall, which would be about 30m. You really need the detailed flight data from the aircraft for more accuracy and to fill in the blind spots. I'd need to confirm that the go-home speed is 10m/s but it sounds about right, if so, these figures would confirm it has flown straight into the side of the building in the 10.5 seconds the data link was down.
That might explain what happened, but not why it did not ascend to RTH altitude. DJI need to analyze your flight data.
Good luck! |
|