P4P crashed while Litchi Mission (no relative to terrain altitude)

[rge]

There's a certain thrill about plotting a mission in Google maps which is outside of the range of the remote and having it execute and return. Especially if there are areas which have a lot of interference which causes your remote distance to be limited. For example, I have it fly to the sides of the railroad tracks which is all trees and have it sit there and record for a couple of minutes and then return

 

[JTorsrud]

Wait, let me get this straight . . . your drone is stuck in a THREE?
View attachment 77795

From the looks of the picture, I'd say that DRONE is pretty much Toast, (destroyed) Maybe the camera can be salvaged, but not much else.

Personally, I'd write it off and leave it. Come back in a year and see how the woodland critters have dealt with it. Might be an interesting experiment in Ecology!

 

[Silversand]

Make sure to carefully check the flight path between waypoints. There can be nasty surprises. Got me about 6 months ago. Also elevation data is not very accurate (interpolation between elevation points where the points have a much lower resolution than the image). Give it at least 50 meters of safety hight in addition.

That's the right idea; you are on the right track with this. Google uses a whole host of digital elevation model (DTM / DEM / TM) product to come up with their whole earth elevation numbers...example: for the USA, they extract elevation data from NED (10 meter), SRTM (30meter to 90 meter), and ASTER (30meter) product. And, the best elevation data now provided by NEXTmap (now Europe and parts of Asia and other regions). The accuracy of the elevation data varies widely from product-to-product and with slope range and land cover (shrubs? treed terrain {mixed? evergreen? deciduous?}? urban environment?). Depending on what is sticking up from bare ground, the RMSE can be quite good (1.53 meters, then 2.94 for NEXTmap) to hellacious (41.1 meters off) ! One could expect an ~~ RMSE of about 26~32 meters (and as much as 41 meters from SRTM and NED derived elevations over forest and treed canopy) when in an area of any deciduous trees. I personally would not ever use Google Earth elevation data to plan my flight elevation trajectory unless my flight was over 99% bare ground (and I'd still correct for sum total RMS error). The absolute best and most accurate elevation data in Google is from NEXTmap product (2.94 RMSE from terrain covered by vegetated canopy and urban = quite good). So, the big question is: what terrain in Google uses NEXTmap data, and what terrain uses NED, SRTM and ASTER???

....I think what the OP was hinting at was that he can't reveal himself as the owner of that treed drone, because by default, it may be at an altitude (and possibly on terrain) that he would be fined flying on....so "it" has to be retrieved completely clandestinely, without ANY authorities knowledge.

 

[circpro]

When you make your plan in google earth, draw the path, then add altitude realative to the ground and check extend path to the ground. I always add 50m. Google earth has no idea how tall a tree is, it's also estimating ground alt and can be off, that's why you have to add the safety margin. Litchi is machine code and doesn't think on its own, it does what you told it to do.
Lastly, why didn't your avoidance sensors see the obstacle? Are they turned on?

 

[circpro]

I'm curious. Why do so many folks use Litchi or similar products? I fly with a terrific hood on my iPad Air 2 and fly FPV. I can see the terrain, make adjustments on the fly, circle an interesting point control my altitude etc. Am I missing something? Definitely not knocking any such type of autonomous flight but just curious

Because some of us want to make professional looking videos

 

[Roamer105]

When you make your plan in google earth, draw the path, then add altitude realative to the ground and check extend path to the ground. I always add 50m. Google earth has no idea how tall a tree is, it's also estimating ground alt and can be off, that's why you have to add the safety margin. Litchi is machine code and doesn't think on its own, it does what you told it to do.
Lastly, why didn't your avoidance sensors see the obstacle? Are they turned on?

Most likely in a WP mission, the bird is not flying "forward". So collision detection is not going to do much.

 

[Roamer105]

I have created and flown some pretty spectacular Litchi WP missions, but I also fly quite a bit of FPV. Collision detection works great for FPV flying.
Here is one example of a carefully planned WP mission through mountainous terrain:

 

[Jim abbatoy]

Neon Euc gave the best advice. Don't give up on your bird. I lost my P3P in a tree recently due to low battery RTH issue. It was also hung with a branch through the landing gear. Once I recovered it, the only damage was a broken ribbon cable. Cheap fix and thanks to Amazon sending me the wrong cable for my P3S, I had the bird up and running in an hour.

 

[NRJ]

Thanks for all the replies. I have a friend who also flies drones and we're discussing on how to get it back. It has been pouring rain for the last two days so the P4P will be soaked. The tree is very small in diameter but high and all branches are pointed vertically. Almost impossible to climb. Next problem is the location. It's on a area where a classified monument resides so we just can not do what we want or law enforcement will come. Also the rules flying drones in my country are far more restricted since last year than the rest of Europe. We only can fly drones less than 1 kg on a max height of 10 meters over own property unless we follow lessons, pay taxes, get an exam and pass, to get a licence flying up to 5 kg with max altitude of 45 meters in open space. That alone sets one back 1500$ US while in all other countries one can fly rather free following basic rules. As you see the risk of going there in public doing crazy stuff will set me back a lot more than losing a drone, how hard it might be right now.

As for the tech stuff, what I also don't understand is the collision avoidance. I got the 4 Pro also for having sensors on all 4 sides. I was not speeding, just few m/sec so they should have worked. I've seen YouTube videos where the P4P does a heck of a job avoiding stuff like trees, branches, etc.. and not only walls. Why did mine just flew into the tree without even hesitating? All question I don't find answers for at the moment. But anyhow, I will certainly fly with line in side in the future !!

This may give you some ideas.

 

[SRJ_Phantom_4_Pro]

That's the right idea; you are on the right track with this. Google uses a whole host of digital elevation model (DTM / DEM / TM) product to come up with their whole earth elevation numbers...example: for the USA, they extract elevation data from NED (10 meter), SRTM (30meter to 90 meter), and ASTER (30meter) product. And, the best elevation data now provided by NEXTmap (now Europe and parts of Asia and other regions). The accuracy of the elevation data varies widely from product-to-product and with slope range and land cover (shrubs? treed terrain {mixed? evergreen? deciduous?}? urban environment?). Depending on what is sticking up from bare ground, the RMSE can be quite good (1.53 meters, then 2.94 for NEXTmap) to hellacious (41.1 meters off) ! One could expect an ~~ RMSE of about 26~32 meters (and as much as 41 meters from SRTM and NED derived elevations over forest and treed canopy) when in an area of any deciduous trees. I personally would not ever use Google Earth elevation data to plan my flight elevation trajectory unless my flight was over 99% bare ground (and I'd still correct for sum total RMS error). The absolute best and most accurate elevation data in Google is from NEXTmap product (2.94 RMSE from terrain covered by vegetated canopy and urban = quite good). So, the big question is: what terrain in Google uses NEXTmap data, and what terrain uses NED, SRTM and ASTER???

....I think what the OP was hinting at was that he can't reveal himself as the owner of that treed drone, because by default, it may be at an altitude (and possibly on terrain) that he would be fined flying on....so "it" has to be retrieved completely clandestinely, without ANY authorities knowledge.

Sounds like someone needs to light the bat light!

 

[Carl1770]

May as well leave there mate as an ornament and a warning to would be Litchi lovers. The onboard "alien" technology has little tolerance for water, so the chances are if you do risk more equipment, funds and yourself to retrieve, your success is likely to rewarded by spare parts only. How do I know that? Well that's another (tragic) story :-(

 

[Pharm]

May as well leave there mate as an ornament and a warning to would be Litchi lovers...

who don't take elevations and obstacles into account.

 

[chris digata]

Loop string over the branch and knock it down.. last resort.. we dropped a weight over

 

[Mavician]

Whoow, a lot of replies, some more interesting than others ;-)

I regret to tell you that we had made up a plan to recover it but it pored rain for a whole day. The next day the P4P was vanished. We did not discover any tracks or evidence of retrieving it. I'm now watching the second hand websites carefully. Did someone beat me and got it that rainy day, maybe. Did the towns administration got it ? maybe. After a few days of feeling bad I didn't look back... up for new adventures with new knowledge in pilot error and training.

 

[RyanK]

From the looks of the picture, I'd say that DRONE is pretty much Toast, (destroyed) Maybe the camera can be salvaged, but not much else.

Personally, I'd write it off and leave it. Come back in a year and see how the woodland critters have dealt with it. Might be an interesting experiment in Ecology!

The photo was just a random photo I found on Google images. It was just to give me a starting point to make the "3" photo. It wasn't the actual drone the OP crashed.

 

[RodPad]

Well I've been lying in bed wondering I might set up a rescue mission. I believe the P4P can lift about 3 times it's own weight. I could hook up some fishing wire that will snap if to much tension is applied. Attach a rather wide angled hook and try to hover over the tree top keeping a safe distance. With the camera facing down I should have a pretty good view and when I choose a day without wind I might get an opportunity to lift the P4P out of the tree. It sits only 1 meter from the top and the branches are thin and flexible and also pointing vertical so the odds could be in my favor. Any thoughts on my plan ?

Is this true, I was under the impression, that a Phantom would not have much more power then carrying its own payload. I figured if you hooked on to a Phantom, with a Phantom, they would both probably crash.

Even though your Phantom is gone / disappeared, how did you come up with this 3 times its own weight.

Rod

 

[RodPad]

Whoow, a lot of replies, some more interesting than others ;-)

I regret to tell you that we had made up a plan to recover it but it pored rain for a whole day. The next day the P4P was vanished. We did not discover any tracks or evidence of retrieving it. I'm now watching the second hand websites carefully. Did someone beat me and got it that rainy day, maybe. Did the towns administration got it ? maybe. After a few days of feeling bad I didn't look back... up for new adventures with new knowledge in pilot error and training.

Hum, It must of fell out if you don't think anybody could climb the tree. How much time had passed between it was there, then it was gone?
Was it easy to see, in the tree?

Rod

 

[Mavician]

Is this true, I was under the impression, that a Phantom would not have much more power then carrying its own payload. I figured if you hooked on to a Phantom, with a Phantom, they would both probably crash.

Even though your Phantom is gone / disappeared, how did you come up with this 3 times its own weight.

Rod

Hum, It must of fell out if you don't think anybody could climb the tree. How much time had passed between it was there, then it was gone?
Was it easy to see, in the tree?

Rod

Yes, no leaves and pretty visible. Only two days, one poring rain. About the weight pulling, I once saw a youtube video of a phantom lifting a wire with 3 phantoms hanging on the wire. Did not own a drone at that time and now understand I must have been a hoax video ?

 

[RodPad]

Yes, no leaves and pretty visible. Only two days, one poring rain. About the weight pulling, I once saw a youtube video of a phantom lifting a wire with 3 phantoms hanging on the wire. Did not own a drone at that time and now understand I must have been a hoax video ?

Hey, thanks for your reply,
Hum, If I'm correct its a good thing you didn't try lifting it with another Phantom.

I was thinking of my own rescue mission, and thought why would they build something that could lift more than its own weight? Then I thought it doesn't seem possible, no matter what the craft, would have to bigger and more powerful than the payload.

Rod

 

[RodPad]

Your thoughts now, If no leaves, would somebody climb it, you didn't think the top would be stable enough?
Did you take any photos when you saw it in the tree?
I'm kind of wondering how far up was in the tree?
I also wonder about somebody reading this thread and coming after you, giving you a fine as stated if maybe you should have not been flying there.

Rod