3D Robotics Comment by Chris Anderson on April 8, 2008 at 4:11pm

Andrea,

Thanks for the post and news on your project. Can you add a picture of your glider to the post?

As for the idea of sensing thermals, it's a great ambition but VERY hard. Many have tried but few have succeeded;-) Solar panels are also a cool idea, but the ones that are high effeciency and light aren't cheap (that will improve over time). I'd start with a basic autopilot and then add FPV--so it can come back on its own if you lose video contact.

Comment by Neal Clark on April 8, 2008 at 8:35pm

HI Andrea,

I believe that this http://soaring.goosetechnologies.com/ project managed to measure thermals. You may also be able to use an accelerometer, altimeter, gps combination to detect when you are rising. Actually i just had an idea but have how idea weather it would work. Do you think it would be possible to detect thermals using a sensitive thermometer. What we would really need to do is send a digital thermometer on a glider and measure the temperature profile.

There are a number of factors which I think need to be considered:
1) temperature gradient compared to surrounding air
2) rate of cooling or rising air (will it still be warm at flight altitude)
3) sensitivity of sensors
4) is there a correlation between temperature and thermals at flight altitude

The following page seems like it will be a great resource:
http://www.apogeerockets.com/education/detecting_thermals.asp

An alternate method may be to map accent and decent rates to determine areas of greatest lift. Such an approach would work for areas of rising air caused by ridges. You may be able to use a camera to search for areas likely to produce thermals (dark rooftops, parking lots, freshly plowed fields).

Keep me posted as I have been researching this topic myself. I am also interesting in collaborating on a design.

Thanks,

Neal.

Comment by Andrea Salvatore on April 9, 2008 at 3:49am

Thanks you all for your comments.

Chris,

I just added a picture of the Omei 2000 to the post as you suggested. For now it's a "bare" one but it gives an idea of the model.

I'll surely develop the project by steps and probably not all my "dreams" will come true, but it's a try.

For me now it's a long way to the autopilot right now :)


Neal,

Thanks a lot for the links! Probably I'll develop the algorith for "thermal search" using some kind of simulator. Do you know of a simulator that implements thermals?

In the past I've been using a TPA81 thermal array infrared sensor. Maybe it can be usefull in detecting the change of temperature related to thermal. But it's just an idea. And it works exactly as you said. It compares temperature to surrounding air.

I'll do more tests with that. I only had the time to work with it for a couple of days and then left it apart.

Searching for specific areas with the camera seems difficult to me now, but maybe we can implement some algorithm for specific area recognition based on colours for example. A nice camera to do that is probably the CmuCam3.

Regarding the solar panels, I done some calculation in the past and for about 1 hour flight I can gain another 10 minutes more. Not much but it's a beginning. I'll write a post about that soon.

Lots of ideas and lot's of things to do.

Thanks again,

Andrea

Comment by Peter Klemperer on April 9, 2008 at 10:28pm

CRRCsim is a free flider simulator that does thermals. If you look at the Charles River Radio Control (or whatever the CRRC stands for) website, they list some other resources for simulation.

Comment by Andrea Salvatore on April 10, 2008 at 2:02am

Ciao Peter,
thanks for the info. For those interested here's the link to CRRCSim.

I've also found a very good soaring simulator, named Silent Wings that seems to have all I need for my simutations as it also has an output interface that I can use with the software I'll write to develop the soaring "AI".

Comment by dincer hepguler on April 12, 2008 at 2:53am

Hi Andrea,
Nice to see you here posting about your work... :) Following thermals will be a challenge for you, solar charging will also be interesting but you must consider powering only some portion of your electronics with it. Another thing is that the glider in the picture does not seem to lift your equipment. maybe you must consider a plane with larger wing area. A Telemaster perhaps...

dincer
p.s. i tried your Arduino code with Nunchuck and works fine thx again... :)

Comment by Andrea Salvatore on April 12, 2008 at 10:17am

Ciao Dincer,

I've posted the how-to of the pan and tilt camera with the Nunchuck and Arduino. You already got and I'm happy it works for you as well.

Thanks for your suggestions. I was already thinking about that but for now I'll concentrate on the autopilot module and the Omei 2000 is good enough for that.

Let me know what you think of my new post and if you have any suggestion from your experience with the code ;)

Andrea

Comment by Matthew Coleman on March 15, 2010 at 3:09am

Andrea, I seem to have the same project mission as you. it has been a while since you posted on this thread. Are you still actively developing the thermal seeking autopilot?

Comment by GliderUAV on June 10, 2011 at 7:01am

Me too. http://www.diydrones.com/forum/topics/atv-autonomous-thermalling

Any updates?

Comment by Joseph on January 15, 2012 at 3:02pm

I to am useing a motor glider for my first UAV project because of the high lift to drag ratio. I think I will ba able to extend the flight times. I was wondering if I could get a copy of your parm. For the ardupilot so that I have a good starting point. Thanks

Comment