How accurate is Google Earth ?

[Storyline]

I have just got Litchi and am excited by the possibilities offered by waypoint flying using MissionHub.

I have watched the excellent tutorials made by Husky-Flies on youtube so am confident I know how the system works but am a bit concerned about handing over control to the software so would be grateful if experienced users could answer my queries.

Firstly, how accurate is Google Earth ? I realise that the GPS position that the Mavic gets is very accurate but how much can I rely on the position that Google earth gives to each waypoint ? If flying higher than any hazard then any differences caused by projection errors etc are not very relevant but how much safety margin should I build in when flying past a hazard at low level ?

Secondly I don't know where Google earth gets it's elevation data from. If it is from other surveyed mapping then presumably it will be accurate but I don't know how the Mavic gets its height data. I have never really trusted any GPS derived altitude but maybe it uses a barometer but even so that would need to take the current air pressure into account. I have a feeling that it would be prudent to build in a big safety margin but maybe I am wrong. It could be that GPS can now produce accurate height info.

Thanks for any advice from Litchi users and any other tips about planning safe flights would be appreciated.

 

[Kilrah]

The aircraft has a barometer but doesn't care about calibration since everything is taken as relative to takeoff point, which inherently "calibrates" it.

Litchi guide tells you that when using waypoint "above ground" mode the first waypoint should be at your takeoff location, so it looks up the offset in the elevation data.
No idea about the accuracy of said elevation data.

 

[Storyline]

The aircraft has a barometer but doesn't care about calibration since everything is taken as relative to takeoff point, which inherently "calibrates" it.

Litchi guide tells you that when using waypoint "above ground" mode the first waypoint should be at your takeoff location, so it looks up the offset in the elevation data.
No idea about the accuracy of said elevation data.

Thanks Kilrah Now I have thought more about it, the initial elevation data in Litchi must be relative to the take off point. The problem lies when the ground itself changes height during the flight. I had not checked the 'Enable Elevation Features' in Mission Hub. Now I have done that I get the option to make the height at each waypoint relative to the ground. This still means though that the heights on Google Earth need to be accurate so my initial question still stands. There is also still the problem of how accurate the registration of the imagery is. Here in the UK the Ordnance Survey (govt mapping) builds trig points which are used in surveying. It would be interesting to find one of these and it's position and see if it is in exactly the right location on Google Earth.

I am definitely going to err on the side of caution on my first mission. It would be easy to take over if the Mavic is close but as it gets further away it may be difficult to see any potential problems.

 

[bushie]

My understanding

There ate a number of Elevation models available. Presumably Google uses one, or a number of different ones depending on the area, to produce its level info. These Dems are produced from satellites designed to pick up the elevation data using a number of different sensors. They would certainly use ground control points for calobration.
There are a number available for free or more accurate if you pay.

 

[ac0j]

Topo maps have been around for a long time. And tend to be quite accurate. Before GPS, we had to use paper topo maps and rulers when figuring HAAT and location coordinates for radio tower installations.
Unless you are planning a grass cutting mission, I would trust in Google maps data on elevation. With a safe above ground altitude set.

 

[Coyyote]

As mentioned, Litchi has an AGL check box that allows you to run a waypoint mission specifying one or more specific altitudes above ground. I have used it often but am fairly conservatine and dont go nap of the earth. Use 150-200 feet up mtn in areas with large pines and usually not less than 60 feet in flat meadows

 

[Meta4]

. This still means though that the heights on Google Earth need to be accurate so my initial question still stands. There is also still the problem of how accurate the registration of the imagery is. Here in the UK the Ordnance Survey (govt mapping) builds trig points which are used in surveying. It would be interesting to find one of these and it's position and see if it is in exactly the right location on Google Earth.

In my part of the world, Google Earth is quite accurate for horizontal position.
I haven't checked its vertical accuracy but that would be easy enough to do.
I would expect heights to be reasonably accurate on a broad scale but would not trust them on a fine scale for local features.

Some things to be aware of wrt Google Earth accuracy:
Even if Google Earth is accurate for horizontal position, your GPS is not.
GPS is subject to variable inaccuracy and may put you within 1-2 metres most of the time but it can sometimes be out by 10 metres or so.
Do not rely on GPS for flying close to any obstacles.

Like paper maps, Google Earth will display heights of the ground but it won't show the heights of trees and other obstacles.
Allow sufficient clearance to clear trees and obstacles.

 

[Storyline]

.....
Some things to be aware of wrt Google Earth accuracy:
Even if Google Earth is accurate for horizontal position, your GPS is not.
GPS is subject to variable inaccuracy and may put you within 1-2 metres most of the time but it can sometimes be out by 10 metres or so.
Do not rely on GPS for flying close to any obstacles.
.....

I used to have an app on my phone that would pop up a warning if the GPS signal was not up to normal standards. This could be caused by solar flares/storms etc. I have just had a quick look in the Play Store and could not see it. IIRC I ended up deleting it because it used up a lot of the phone's resources.

 

[Meta4]

I used to have an app on my phone that would pop up a warning if the GPS signal was not up to normal standards. This could be caused by solar flares/storms etc.

Deleting it was probably a good idea.
Solar flares etc have never caused any problems for drone flying and the supposed risks of the K-index is a much hyped myth
When your drone is up in the air away from obstructions it's always going to have plenty of satellites to provide optimum GPS signal.
But even with perfect GPS signal, you will still get variable position accuracy.
That's just how GPS is.

 

[Storyline]

K index - yes that was it. I have rarely had any issues with GPS which we sometimes rely on when on our boat. It is usually the charting that is not very accurate especially as some of the surveys in areas with no commercial traffic can date back a hundred years or more. A few years ago there was a large NATO exercise fairly close to where we were sailing and during this the GPS was jammed. Not sure why they did this but it took me back to the time when the only nav aid we had was a radio direction finder.

I don't know where you are based but here in the UK we have just been told by the EU that we will not get (unencrypted) access to the new European Galileo gps system despite having paid our share and provided a lot of the hardware and software. There is now talk of us going alone on our own system or maybe combining forces with Canada and Australia which sounds more sensible. I heard the other day however that there is some question over whether there is enough room on the frequency spectrum used by GPS. Whatever happens us leisure users will benefit from galileo and our Mavics should get an even faster fix presumably.

Drifting further off topic I have just been reading about the Chinese system which is expected to offer millimetre accuracy although I suspect it will not be free for public use.

 

[Meta4]

in the UK we have just been told by the EU that we will not get (unencrypted) access to the new European Galileo gps system despite having paid our share and provided a lot of the hardware and software. There is now talk of us going alone on our own system or maybe combining forces with Canada and Australia which sounds more sensible. I heard the other day however that there is some question over whether there is enough room on the frequency spectrum used by GPS. Whatever happens us leisure users will benefit from galileo and our Mavics should get an even faster fix presumably.

You will still be able to use the existing GPS and Glonass systems just as you do now.
Your current Mavic won't work with any future satellite systems.
Maybe future drones will - but only if they get satellite receivers capable of working with future systems.

 

[Storyline]

Yes that makes sense. The GPS chip can't be expected to receive signals from satellites it hasn't even been told about.

I have just been reading about Galileo and it is not too bad for the UK as we have only contributed £1 billion of the £8 billion total cost. The satellites are apparently all controlled by a centre in Portsmouth so we could always switch them off

 

[gerrymad]

I have found it to be quite accurate. I saw one example of a flight where it came down low between two groves of trees. I don't think I would trust it quite that far. That does bring up one point that elevation is ground elevation. You have to allow for trees and things. I planned one low route across water and then over a little stretch of land. When I got there I found the trees were quite a bit higher than I expected so I needed to adjust.

 

[Storyline]

I have found it to be quite accurate. I saw one example of a flight where it came down low between two groves of trees. I don't think I would trust it quite that far. That does bring up one point that elevation is ground elevation. You have to allow for trees and things. I planned one low route across water and then over a little stretch of land. When I got there I found the trees were quite a bit higher than I expected so I needed to adjust.

I've just learned from another thread how the elevation data the Mavic uses is derived from it's barometer. This means that as long as you tell Litchi to use elevation relative to ground it should fly accurately at that height. You could design a Mission at what you know will be a safe height and then progressively reduce it over a few runs as you know it will fly consistently. Given that the horizontal position is always accurate with GPS it
should be possible to have a safe Litchi mission flying as low as you want through narrow gaps. Throw in some tight turns and flying at high speed should make for some dramatic footage.

Someone else can try first !

 

[Meta4]

Given that the horizontal position is always accurate with GPS it should be possible to have a safe Litchi mission flying as low as you want through narrow gaps. Throw in some tight turns and flying at high speed should make for some dramatic footage.

GPS accuracy is not what you might imagine.
Even in good GPS coverage it is only within +/-2 metres and sometimes quite a bit more than that.
Relying on GPS to precisely position your drone for fast flight close to obstacles may well get some dramatic footage - more than you would really want.
Always allow a comfortable safety margin.

 

[Storyline]

GPS accuracy is not what you might imagine.
Even in good GPS coverage it is only within +/-2 metres and sometimes quite a bit more than that.
Relying on GPS to precisely position your drone for fast flight close to obstacles may well get some dramatic footage - more than you would really want.
Always allow a comfortable safety margin.

Yes, even 2m is quite impressive and as I said it would be good if others tried first
I don't know when Galileo will come online but in 2020 Beidou is supposed to be available for civilian use and they claim it could offer extreme accuracy. So probably another 3 years to wait until we can fly safely within inches of things.

Edit: One of the advantages of DJI being based in Shenzhen is Chinese manufacturers will want to earn brownie points by promoting Beidou and there will be a real push to incorporate it into the navaids on their products. Plus the chips needed and software will be developed rapidly, everyone wanting to help China fly the flag and get one in the eye of the US and what they will say is it's old fashioned and inaccurate GPS system - it's a nailed on certainty