Antarctic work

[KranMc]

This is all too cool. How fascinating. Thank you for sharing. Please post the video.


Sent from my iPad using InspirePilots

 

[southerndoug]

WOWOWOW!!! brilliant use of UAV technology..I take it strong interference was caused by the metal ship? My UAV (it was a phantom at the time) wouldn't boot up on top of my car and on a steel reinforced pier

The metal ship certainly had something to do with it, but I think the biggest deal was the compass being way out. When I did a test on the ship before we left Australia, with a well calibrated machine, the Inspire started without complaint on the helideck of the ship, which really surprised me (like you, I have had trouble starting on reinforced concrete). There was also a "sweet spot" on the deck where starting was relatively easy as well. But the easiest solution was, as I said, to block GPS, and persevere with the CSC in various modes until it started, even after it had thrown up Compass and IMU warnings.

 

[NickU]

I don't think it was the metal of the ship that gave you problems (although ships do have their own magnetic field that needs compensating for). It might sound strange but magnetic compasses can become erratic close to the polar regions (and true north doesn't coincide with the magnetic north either). Gyroscopes can also have problems because of the earth's rotational speed close to the polar regions, and gps coverage can have positional errors due to there not being any overhead satellites (no satellites above 55N or below 55S), and due to signal propagation errors in the atmosphere - put all those together and you've got a headache for any electronic system that uses a magnetometer, gyroscope and gps... it'll have a great time trying to compensate for all the errors it's being hit with at once. It isn't just UAVs that have problems, aircraft and ships do too.

I think compass errors are often compared with and corrected by using gps signal data, so by shielding the GPS receivers, you basically removed one set of calibration 'noise' from being fed back into the system. That probably was enough to allow the aircraft to think the compass had settled for a bit... even if it was actually still wrong! Once the compass had 'set' and you allowed the gps signal again, the aircraft probably gave more credence/weight to the gps signal for Nav positioning rather than the compass. Then there's the gyros to account for...

All in all, it's a testament to the microprocessors in the IMU/Nav systems that the aircraft could even begin to sort out the mess and get (and stay) airborne!

 

[southerndoug]

I don't think it was the metal of the ship that gave you problems (although ships do have their own magnetic field that needs compensating for). It might sound strange but magnetic compasses can become erratic close to the polar regions (and true north doesn't coincide with the magnetic north either). Gyroscopes can also have problems because of the earth's rotational speed close to the polar regions, and gps coverage can have positional errors due to there not being any overhead satellites (no satellites above 55N or below 55S), and due to signal propagation errors in the atmosphere - put all those together and you've got a headache for any electronic system that uses a magnetometer, gyroscope and gps... it'll have a great time trying to compensate for all the errors it's being hit with at once. It isn't just UAVs that have problems, aircraft and ships do too.

I think compass errors are often compared with and corrected by using gps signal data, so by shielding the GPS receivers, you basically removed one set of calibration 'noise' from being fed back into the system. That probably was enough to allow the aircraft to think the compass had settled for a bit... even if it was actually still wrong! Once the compass had 'set' and you allowed the gps signal again, the aircraft probably gave more credence/weight to the gps signal for Nav positioning rather than the compass. Then there's the gyros to account for...

All in all, it's a testament to the microprocessors in the IMU/Nav systems that the aircraft could even begin to sort out the mess and get (and stay) airborne!

Yeah Nick, there's a lot going on in there all right!

As you probably know, the DJI manual mentions that the Inspire (and Phantom) are not to be used in P-GPS modes in Polar latitudes. The compass is not only struggling with magnetic declination, but inclination as well, especially near the magnetic pole. It is not possible to calibrate the compass on the ship, and even if you did, once you left the local magnetic field influence of the ship, your calibration would be out again. When I got to the Antarctic mainland, and had the opportunity to leave the ship and perform a compass calibration, I was able to use P-GPS, and F mode, and even flew a survey using Map Pilot with no problems. However, the same error messages came up every time I wanted to start ("Compass Error. Exit P-GPS Mode"), and the only way around it was to block GPS, as described previously. As you infer, probably mimicking being inside, so the firmware could cope with the compass reading suddenly not being in conflict with the GPS signal.

There are plenty of visible satellites down there though... I usually had >20 in view, and all our aircraft happily use GPS no worries at all. The MEMS gyros and accelerometers would probably be more affected by temperature shock than anything to do with the earth rotating. I did an IMU calibration when I was on the ice, and the Inspire started straight away. After that, I just ignored the IMU warning (which was: "WARNING: IMU Initializing. Do not move aircraft until initialization is complete"), and did the GPS-block trick.

 

[NickU]

I'm still amazed the electronics were able to sort themselves out at alll given the environment they were operating in . You'd be seeing a fair number of satellites as you'd pick up ones which would be over the horizon for most people. the gps errors would be in positioning/altitude as the signal path would be longer/shallower and would have propagation errors from the thinner troposphere. I remember reading about some ship based autopilots that had problems with their position suddenly jumping while under cruise, and which resulted in the autopilot commanding full rudder until it worked out its real position again. It caused a fair bit of head scratching as to what had happened

Did you fly during any aurora or solar storms?

 

[southerndoug]

I'm still amazed the electronics were able to sort themselves out at alll given the environment they were operating in . You'd be seeing a fair number of satellites as you'd pick up ones which would be over the horizon for most people. the gps errors would be in positioning/altitude as the signal path would be longer/shallower and would have propagation errors from the thinner troposphere. I remember reading about some ship based autopilots that had problems with their position suddenly jumping while under cruise, and which resulted in the autopilot commanding full rudder until it worked out its real position again. It caused a fair bit of head scratching as to what had happened

Did you fly during any aurora or solar storms?

Not during any storms that I knew of... there was no way of checking the Kp index while I was on the ship. We had some awesome aurora on the way south though. And I think you'll find that the GNSS is accurate worldwide now, and the problems you refer to are a thing of the past.

 

[Jonnyb333]

thanks for posting this. It is really amazing, including the Ground Station.

 

[KranMc]

How did you block the gps? Aluminum foil?


Sent from my iPad using InspirePilots

 

[Rich Sugden MD]

I recently returned from Antarctica, where I was using an Inspire 1 Pro with X3 and X5 cameras, providing short range ice reconnaissance capability for the Australian icebreaker RSV Aurora Australis. It was a very light ice year, so I wasn't really needed, but I did a couple of trial flights to demonstrate what was possible with the technology.
This was a follow-on from a similar trial that was done last year (by another operator), and showed that the RPA could be easily deployed (compared to a full-size helicopter: we had 3 AS350 B3s onboard), and the guys/gals driving the vessel could get immediate info for short range navigation via the live video feed we had from the RPA.
Of course we had issues with compass calibration, and all flights were done in ATTI mode. The greatest challenge was actually starting the Inspire, with the firmware stopping the RPA from starting when it detected strong interference or a compass calibration issue. There were also a number of IMU problems that were thrown up as well. The work-around for all of these was to block all GPS reception, and attempt a start in ATTI, then P-GPS, then ATTI, etc, until it would finally start, and away we went.
The only modifications I made were insulation for the battery compartment, and the addition of a strobe to aid in maintaining VLOS. Oh, and some day-glo orange paint. The coldest we had was -5 deg. C, so not too bad really, and the batteries were fine. Highest wind speed was 22 knots.

View attachment 11193

I had an area approval for operations up to 500m ASL, but only got to 315m on my second flight, which was still pretty good. You may wonder why bother with approvals down there anyway... but I was working for a government department, and all the boxes had to be ticked, including permissions for >400 feet ops from CASA.
Anyway, it was a great experience, and quite challenging too.

Happy to answer any questions anyone may have.

View attachment 11191


View attachment 11192

Great photos!!

I took my Phantom 3 Advanced up to the arctic a couple of years ago and was able to get some great photos ... BUT had a lot of trouble w. unexpected crashes from high hovers which we think was due to problems w. the compass calibration ... due to the high magnetic variation there ...

 

[Kiwi Carl]

Cool.

 

[southerndoug]

How did you block the gps? Aluminum foil?


Sent from my iPad using InspirePilots

Stategically placed hands. Just got my camera guy to cover the GPS antenna until the Go app showed 0 satellites, and viola! It was a bit messy though, and if there's a next time, I'll make a little cover out of lead flashing or something similar to cover the antenna area. Might look a bit more professional...

 

[southerndoug]

Some video from the trip...

<iframe src="Antarctica 2016, from Inspire 1 PRO from Doug Thost on Vimeo" width="640" height="360" frameborder="0" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>

 

[KranMc]

Great! All will want to see more, more, more! Happy New Year!


Sent from my iPad using InspirePilots