DISABLE "Smart go home."

Won't it do that exact thing with safe go home selected when the battery reaches return point reserve?

Seems when it initiates it uses the RTH altitude settings once the 20m from home point is satisfied, if homepoint is approached again will it just autoland? Other than that, normally the next alarm is safe go home(if set) and then critical battery land.
 
These are extremely sophisticated gadgets. Yes, technology isn't perfect and does fail. But usually it isn't the cause though it can be. So unless proven otherwise I would think the OP was over 65' from home point when the low battery RTH triggered. If say that was the case, being over 65' away from Home point, then all the pieces fit together (no pun intended) and it behaved as it should.
More likely the drone calculated it was greater than 65' from the homepoint when it was actually not. There is a lot of assumption that GPS accuracy is always great, especially if tracking a lot of satellites, but this is not always the case. GPS is notoriously inaccurate under trees and +/- 65 feet for absolute position accuracy is not unreasonable assumption. I work with survey grade GPS and I can tell you it is common for a GPS receiver to calculate a position that is 65' off from the actual position when the GPS reception is poor, especially under trees.

And its not so much the quantity of satellites being tracked as it is poor satellite geometry. The more spread out they are the longer the baselines for the math to figure out the position of the GPS receiver. Tracking 8 satellites spread out evenly over the sky will give better positioning accuracy than 16 satellites all in one portion of the sky, like you might encounter standing along the edge of a forest. If you google dilution of precision for GPS you'll find more information, but here are a couple with decent explanations: Basic GPS and Dilution of precision (GPS) - Wikipedia, the free encyclopedia

So in a nutshell its entirely possible to be landing within 65' of the home point but have the drone think its not that close and go into RTH mode instead of auto-land.
 
Last edited:
More likely the drone calculated it was greater than 65' from the homepoint when it was actually not. There is a lot of assumption that GPS accuracy is always great, especially if tracking a lot of satellites, but this is not always the case. GPS is notoriously inaccurate under trees and +/- 65 feet for absolute position accuracy is not unreasonable assumption. I work with survey grade GPS and I can tell you it is common for a GPS receiver to calculate a position that is 65' off from the actual position when the GPS reception is poor, especially under trees.

And its not so much the quantity of satellites being tracked as it is poor satellite geometry. The more spread out they are the longer the baselines for the math to figure out the position of the GPS receiver. Tracking 8 satellites spread out evenly over the sky will give better positioning accuracy than 16 satellites all in one portion of the sky, like you might encounter standing along the edge of a forest. If you google dilution of precision for GPS you'll find more information, but here are a couple with decent explanations: Basic GPS and Dilution of precision (GPS) - Wikipedia, the free encyclopedia

So in a nutshell its entirely possible to be landing within 65' of the home point but have the drone think its not that close and go into RTH mode instead of auto-land.
Great insight. Thanks for posting.
 
I agree, that was good, reminds me of looking for geocaches, if there hidden under trees, many are, you have to pick your target area while walking in clear area.
Yes the Flight data, should have the answers!
 
  • Like
Reactions: ParsnipHysorter
More likely the drone calculated it was greater than 65' from the homepoint when it was actually not. There is a lot of assumption that GPS accuracy is always great, especially if tracking a lot of satellites, but this is not always the case. GPS is notoriously inaccurate under trees and +/- 65 feet for absolute position accuracy is not unreasonable assumption. I work with survey grade GPS and I can tell you it is common for a GPS receiver to calculate a position that is 65' off from the actual position when the GPS reception is poor, especially under trees.

And its not so much the quantity of satellites being tracked as it is poor satellite geometry. The more spread out they are the longer the baselines for the math to figure out the position of the GPS receiver. Tracking 8 satellites spread out evenly over the sky will give better positioning accuracy than 16 satellites all in one portion of the sky, like you might encounter standing along the edge of a forest. If you google dilution of precision for GPS you'll find more information, but here are a couple with decent explanations: Basic GPS and Dilution of precision (GPS) - Wikipedia, the free encyclopedia

So in a nutshell its entirely possible to be landing within 65' of the home point but have the drone think its not that close and go into RTH mode instead of auto-land.
Yes, totally agree. It might not have actually been 65' away but given the description of what happened I truly believe it thought it was 65' or more away. Whether it actually was, or only perceived because of the GPS satellite info I think that is what happened, it believed it was 65' or more away.
 
  • Like
Reactions: ParsnipHysorter
More likely the drone calculated it was greater than 65' from the homepoint when it was actually not. There is a lot of assumption that GPS accuracy is always great, especially if tracking a lot of satellites, but this is not always the case. GPS is notoriously inaccurate under trees and +/- 65 feet for absolute position accuracy is not unreasonable assumption. I work with survey grade GPS and I can tell you it is common for a GPS receiver to calculate a position that is 65' off from the actual position when the GPS reception is poor, especially under trees.

And its not so much the quantity of satellites being tracked as it is poor satellite geometry. The more spread out they are the longer the baselines for the math to figure out the position of the GPS receiver. Tracking 8 satellites spread out evenly over the sky will give better positioning accuracy than 16 satellites all in one portion of the sky, like you might encounter standing along the edge of a forest. If you google dilution of precision for GPS you'll find more information, but here are a couple with decent explanations: Basic GPS and Dilution of precision (GPS) - Wikipedia, the free encyclopedia

So in a nutshell its entirely possible to be landing within 65' of the home point but have the drone think its not that close and go into RTH mode instead of auto-land.
Yes. We see that a lot in newbie boaters. They have the nice GPS autopilot that happily runs them into the channel marker and they;re all bent out of shape (literally). But. But. But. It says that it's accurate to three feet! I was just getting a beer!
 
Good one also, Channel Marker (Bent out of shape):)
 
Would be nice if the OP posts the log so we can see what happened :)

Could it be the VPS which caused this?

The lack of GPS signal under the tree seems likely though.


Sent from my iPhone using Tapatalk
 
I don't know if I'm blind, but I can't see this option in Settings. Or is it available only in iOS verion of DJI GO?
It should be in both, but I can only confirm in iOS. It's in the Fail Safe menu.
 
I completely am not following what Smart RTH has to do with this problem, LOL. ???
 
I completely am not following what Smart RTH has to do with this problem, LOL. ???
Read the attachment. It covers a bit about the "Smart Go Home".. hope it will lighten you up.. :)
 

Attachments

  • Compass calibration warning 11-6-15.pdf
    447.4 KB · Views: 404
Thanx for that !! Nice jib


Sent from my iPhone using PhantomPilots mobile app
 
Yes, That was good, Don't know what mine setting was, option not on tablet, Phantom in for repair, probably won't ever know unless it is some data somewhere.
 
Read the attachment. It covers a bit about the "Smart Go Home".. hope it will lighten you up.. :)
I don't know who wrote that pdf, but it contains misleading information.
Especially sentence "when the battery hits 10% a DJI Inspire 1 or Phantom 3 will automatically initiate it’s “Smart Go-Home” feature and head for its home point" is ********.
We all know, that if battery hits 10 %, Phantom will start to land immediately, no matter what.
Smart Go Home means, that during the flight, Phantom calculates, how long would it take to flight to Home point and also, how much flight time can battery give. And if Phantom finds out there is less battery flight time than time needed to get home, Smart Go Home is initialized.
 
I disabled smart go home because of this reason like what happened to him..
Hi Aquar. From what the op said. He did not envoke RtH. That happened bc of low batt. So have smart rth off would not have helped. Correct?
 

Recent Posts

Members online

Forum statistics

Threads
143,094
Messages
1,467,591
Members
104,979
Latest member
jrl