Crashed P4P While on a Litchi Mission

Joined
Dec 6, 2018
Messages
57
Reaction score
10
Age
72
I was flying my P4P using a mission that I have flown 7 or 8 times. This time it crashed into the tree tops. I was able to retrieve it and the video showed it crashing into a large branch. (see pic). I don't know what changed? The last time I flew the mission was Nov 8, 2020. I live in Pwnnsylvania and do not believe the trees grew that much in that time. The only thing I can think of is that the leaves on the trees allowed the bird to "see them" and avoid them. What I plan to do next is take-off from my usual spot and walk the few hundred yards to the tallest tree and see how tall it is with the bird and then check against the mission specifications. Any other ideas will be appreciated.
 

Attachments

  • Screen Shot 2021-01-11 at 3.54.47 PM.png
    Screen Shot 2021-01-11 at 3.54.47 PM.png
    699.5 KB · Views: 227
Did you take video during the previous 7 runs? If so, I would compare that video to this one and see exactly how the drone acted differently. It's possible your P4P avoided the branches in previous missions by a couple of feet, but this time a gust of wind put the branch momentarily into it's path. And yes, it's always best to conduct field reconasence before actually flying a mission. You cannot easily spot such things as overhead wires, flag poles, etc on the Litchi aerial without having had some experience in doing so.
 
  • Like
Reactions: WV. Rootman
I was flying my P4P using a mission that I have flown 7 or 8 times. This time it crashed into the tree tops.
Programming Litchi to fly close to obstacles is asking for trouble and a comfortable safety margin is necessary.
How much clearance (both horizontal and vertical) do you think your flight should have had with respect to the trees?
 
Programming Litchi to fly close to obstacles is asking for trouble and a comfortable safety margin is necessary.
How much clearance (both horizontal and vertical) do you think your flight should have had with respect to the trees?
I have experienced three very close shaves, where my Phantom 3S just barely cleared treetops that appeared to be much shorter than they really were, when viewed on the satellite map. I also noticed that Lichi does not compute altitude accurately when the drone flies down into a valley from higher ground.

Fortunately, those altitude errors take the form of the drone flying higher above treetops than was programmed, when descending into a valley, causing me to incrementally lower the altitude over several subsequent flights, to arrive at the exact altitude I originally intended for my flight plan
 
Did you take video during the previous 7 runs? If so, I would compare that video to this one and see exactly how the drone acted differently. It's possible your P4P avoided the branches in previous missions by a couple of feet, but this time a gust of wind put the branch momentarily into it's path. And yes, it's always best to conduct field reconasence before actually flying a mission. You cannot easily spot such things as overhead wires, flag poles, etc on the Litchi aerial without having had some experience in doing so.
Yes, I did take videos. I will took to see if there are flight logs. Is my assumption correct that the Litchi software will allow the bird to deviate from the flight plan to avoid an obstacle?
 
I have experienced three very close shaves, where my Phantom 3S just barely cleared treetops that appeared to be much shorter than they really were, when viewed on the satellite map. I also noticed that Lichi does not compute altitude accurately when the drone flies down into a valley from higher ground.

Fortunately, those altitude errors take the form of the drone flying higher above treetops than was programmed, when descending into a valley, causing me to incrementally lower the altitude over several subsequent flights, to arrive at the exact altitude I originally intended for my flight plan
you should consider the "Ground elevation" data that Litchi has access to, to only be accurate to about +/- 10-30 m.........(less accurate for irregular terrain).....
 
  • Like
Reactions: Billie Nelson
Yes, I did take videos. I will took to see if there are flight logs. Is my assumption correct that the Litchi software will allow the bird to deviate from the flight plan to avoid an obstacle?
Good question. I would assume the sensors are programmed to overide any other input -- but I don't know for sure.
 
I am always careful to select altitude above terrain at the drone's current location along its flight path, rather than elevation above the launch point, yet, for some reason Litchi errs on the side of assigning excessive altitude, whenever the day's flight plan takes the drone along the gentle valley of a nearby river whose general elevation is about 100 feet below the launch point.

On higher ground whose elevation is the same, or close to that of the launch point, Litchi's altitude assignment always looks pretty close to the altitude I programmed in the Mission Hub. When overflying terrain say 100 feet below the launch point, however, I have noticed that consistent error of the drone flying quite a bit higher above the trees than my flight plan called for.

By repeating the same flight path at progressively lower altitudes, I can inch closer to the treetops on each successive flight, until the video playback shows the ideal bird's eye view that I seek. All told, Litchi is one amazing work of software engineering, offering such precise control of flight and camera orientation. I can't wait for my Mavic Pro to arrive, so I can exploit its 27 minute battery life on even longer autonomous Litchi forays.
 
Yes, I did take videos. I will took to see if there are flight logs. Is my assumption correct that the Litchi software will allow the bird to deviate from the flight plan to avoid an obstacle?
If you are close enough to hope that obstacle avoidance will ensure a safe flight, you are too close to be doing a waypoint flight.
 
  • Like
Reactions: bsartist
Good question. I would assume the sensors are programmed to overide any other input -- but I don't know for sure.
I watched a video (posted here I believe) of a litchi flight that went out of range and the video shows the P4P expending considerable effort to avoid bushes and even fence posts.
 
If you are close enough to hope that obstacle avoidance will ensure a safe flight, you are too close to be doing a waypoint flight.
Not disputed when it comes to planning a mission. His comment may have been more related to his post mortem analysis though. I.e. can obstacle avoidance possibly explain why he didn’t crash on prior flights.
 
  • Like
Reactions: Aairon

Members online

Forum statistics

Threads
143,066
Messages
1,467,358
Members
104,936
Latest member
hirehackers