AttoPilot Users Group

Smarter Cheaper Faster

First questions is about accuracy. I have seen that the v1.8 has 3m resolution with 5 decimal place accuracy. Can this be increased by putting in waypoints on sd card with say 6 decimal place to get approx 10cm accuracy? What is the accuracy of the v3 IMU?

Second about Chips? Could an increase in accuracy (sub-meter) be accomplished with a WAAS enabled GPS receiver? Oh, I see WAAS support is used. Then can more decimal places be used? Is PDOP measured as a tracked parameter? Does the current (1.8) and/or (3.0) have the ability to use the Glonass System for better coverage?

Views: 147

Reply to This

Replies to This Discussion

Hi Erik

I think other issues will come to play then, the airframe you fly better track pretty straight and true.

There probably can't be much wind, or rather gusting, what happens if 15 knots of gust hits your airframe when your 10m from where you want to be??

Perhaps you could explain why you need cm accuracy? Then maybe a different method could be resolved.

Cheers

G
Gary,

Thanks for the response.

You are right about the airframe. Certainly gusting will be a factor, but what i am looking for is real precision when I create a flight plan with regards to the line the plane actually flies. This is because I am working to connect a LIDAR sensor to the airframe for aerial survey capability.

Now, I am not looking for elevations or xy within inches (10cm)....but I would like the accuracy of the GPS to be sub-meter....hence the addition of a sig fig. I do have commercial GPS equipment that has GPS (WAAS) +Glonass to get sub meter real-time (much better post-processed), but all I want is a low PDOP and 1-2 foot accuracy....attopilot will correct and keep it on course as close as possible even with wind gusts correct? So the closer to "precise" based on my line....the better to reduce skew on the ground. I assume? HA! Also I plan to have the ability to fly 2-3 or more times to use the best data based on "those pesky wind gusts" or other "acts of God". My assumptions is that with the capabilities of the new IMU and then the addition of a "commercial GPS chipset" This thing will be commercially viable in this area?

Thoughts?
Well you will struggle making an airframe to lift a Lidar and still come in under 2KG, the first weight restriction point in the USA.

I would have thought the Lidar unit would have known exactly where and how high it was when the scan was taken and would correct. have to find out how its done in aircraft.



Erik Baros said:
Gary,

Thanks for the response.

You are right about the airframe. Certainly gusting will be a factor, but what i am looking for is real precision when I create a flight plan with regards to the line the plane actually flies. This is because I am working to connect a LIDAR sensor to the airframe for aerial survey capability.

Now, I am not looking for elevations or xy within inches (10cm)....but I would like the accuracy of the GPS to be sub-meter....hence the addition of a sig fig. I do have commercial GPS equipment that has GPS (WAAS) +Glonass to get sub meter real-time (much better post-processed), but all I want is a low PDOP and 1-2 foot accuracy....attopilot will correct and keep it on course as close as possible even with wind gusts correct? So the closer to "precise" based on my line....the better to reduce skew on the ground. I assume? HA! Also I plan to have the ability to fly 2-3 or more times to use the best data based on "those pesky wind gusts" or other "acts of God". My assumptions is that with the capabilities of the new IMU and then the addition of a "commercial GPS chipset" This thing will be commercially viable in this area?

Thoughts?
That was a bit sharp, sorry.

But lifting a lidar would be a big mission, I'm sure many people are onto the case though. Legalities in the USA or in fact Europe for that matter would be a bigger mission. Don't expect anyone less than a rated pilot to be flying that!

The Atto holds its line very well but cm resolution, thats one for Dean to answer.

No doubt he will have a very comprehensive answer ;-)
Thanks Gary,

I know it is ambitious. That is why it is fun right?

On ground scanning uses survey stations for their accuracy, planes use GPS. We contract right now with a company that does aerial survey that, because of GPS accuracy, can only get us 2 ft contours on the ground. Which....is about as good as anyone needs for large area civil work.

Our experimental airframe is larger and would fit into the Group IV FAA SUAS category....so it would be bound to legalities.

Thanks for the conversation.
Oh. Does Dean have a release date on the IMU set? Ballpark that you may know?

Thanks!!!!
Here is my understanding on the GPS: With PDOP of 1.0 the lat/lon CEP (circular error of probability) is 3.0 meters. This is a measure of accuracy, NOT precision. If you use a Trimble planning software and have a current almanac and input your lat/lon and date and time, you can see the PDOP for your area right then depending on the current satellite coverage. If the PDOP is 2.0 (for example) then you would have 6.0 meters CEP. There are times in almost each week anywhere on the globe when in that area the PDOP will get bad as in 7 or more due to poor view of sats. The sats are not in Geo-sync orbit; they are moving.

Precision (self consistency) is another story. When not in WAAS mode, you can watch the reported precision drift in the 5th and 6th decimals as a random walk. When WAAS locks in then lat/lon are stable to the 7th decimal which means the GPS is self-consistent to the 1cm range. This is seperate from accuracy... there is still a 3+ meter uncertainty in absolute position.

Until the CEP for lat/lon gets < 3m, then there is no reason to pursue navigation beyond the 5th decimal. The 5th decimal is 1.1m resolution on the grid. While the Atto GPS reports lat and lon to the 7th decimal, the math resolution that AttoPilot uses in the navigation trigonometry is only to the 5th decimal (1.1m grid). Compare this to what U-Nav thinks is OK to give their customers: 4th decimal grid (11m navigation!!).

The v3.3 IMU autopilot is only 6 DOF IMU (3 axis accel, 3 axis gyro) without magnetometers. Therefore it isn't intended to do dead-reckoning navigation. It relies on GPS for position and heading like the v1.8 does. Again, 5th decimal is the max needed until a new class of GPS units comes out with better accuracy than 3m. On a practical note, it would not matter (in my opinion) or be very possible to command a fixed wing UAV to < 1m accuracy. Maybe for VTOL UAVs accuracy in cm might help hovering over a fixed spot.

All that said, the Atto v1.8 has no problem at all hitting waypoints and holding to paths on routine basis to 0-4 meter based on the GPS data available.... to state the actual accuracy you'd have to take the 0-4m and combine it with GPS inaccuracy of 3m. This means that you can tust Atto to hit absolute coordinates, on a daily basis, better than 10m.
Dean. Thank you for the thoughtful and thorough answer.

Since GPS is taking time at waypoints, I think I could post-process the data and correct the data based on the differential correction provided by CORS. This would give me as accurate as I am going to get and see what the difference in accuracy is compared with the known data from the source (sats).
With regards to the precision….you are dead on. I really want to know how close the flight plan is to the actual flight….first and foremost. Mainly just for my own giggles. If I take measurements on a vector grid of say 100m….all I would want to see is the lines cross at my grid points….doesn’t matter how it gets between the two( well, minus falling out of the sky).

What I am really excited about is being able to hook up the Attopilot to GIS. I have connected other telemetry devices into my GIS to see it real time and then keep a log. Ultimate goal is to be able to create a flight plan, watch it, log it and then compare theoretical and observed…for analysis.

Just looking to nerd out in my own way.
Well if you can work back from, the exact track then Roberts your fathers brother. You will be able to review the track from the swathes of data that the Atto records for you or in real time via a pair of modems attached.

Good luck with the bigger UAS.

The failsafe from Millswood is another good reason to plump for Atto if your heading this way, IOS is the chap to tell you about that.

I don't think anybody else has a similar product.

I certainly know people that would que up for a LIDAR equiped bird.
Which laser imaging device would you use to record the data? Or are you building your own LiDAR system as well???

Reply to Discussion

RSS


Locations of visitors to this page

© 2017   Created by Gary Mortimer.   Powered by

Badges  |  Report an Issue  |  Terms of Service