Developer Comment by Ryan Beall on October 15, 2009 at 5:49pm

Barometric altitude is good enough for a safe and survivable crash landing. I say crash because you never know exactly where ground is because of elevation change etc unless you use a laser range finder or something of the sort. But either way if you hold a slow airspeed in your control laws or even a pitch attitude slightly below the horizon and kill the motor the plane will land just fine. You can get fancier about things but either way its just going to be a mostly unflared landing which most aircraft can take. Flying wings in a grass field is a joke easy time to accomplish.

Comment by Mike Bakula on October 15, 2009 at 7:41pm

Micropilot and Piccolo both use sonar sensors for precision altimetry during flare and landing. Piccolo can also use a point laser altimeter, but I've never seen one. They are noisy, and take a good deal of power, but they do work. They only work out to 15 feet or so, but that's the critical area for small UAVs. They're not suitable for water, unless you protect the transducer somehow.

Developer Comment by Ryan Beall on October 15, 2009 at 7:54pm

This is great for short distances and no prop noise.

http://www.sparkfun.com/commerce/product_info.php?products_id=8504

Comment by Curt Filipowski on October 15, 2009 at 9:23pm

I was toying with the idea of using a short distance range finder (IR, Sonar, etc.) to detect the ground and initiate a flare for landing. That way I could use GPS and barometic to keep me flying at a shallow dive angle, and then once I get in range of the short range, cut power, and slowly flare until I basically stall it into the ground...figured it would probably work perfectly fine for a flying wing...something delicate may be a different story.

Comment by Recmaster on October 15, 2009 at 10:58pm

As I see the ultra sonic sensor is a best solution.
The problem is how to implement it to Ardoboard (witn IMU) and what algorithm have to use for landing code.
I think it is a possible becouse we have a realisation this schem by BlimpDuino.
Is this question to Jordi and Chris? ;)

Comment by Rana on October 15, 2009 at 11:05pm

James,
Safe autonomous landing is is only possible in your well know flying groung. By well know, I mean you know what kind of surfaces is there on various patches of the ground and in which landing approach direction what kind obstructions like trees are there.

Along with pressure sensor like BM085 (currently used in UAVX Quadrocopter to maintain altitude), you can use Sparkfun's "Ultrasonic Range Finder - Maxbotix LV-EZ4" it can precisely sense the range with a accuracy of an inch, off course you have to use autopilot hardware and firmware.

With continuous high speed video recognition system with stereo cameras, it is possible to land even in the unknown flying fround in that case your plane will become a robotic plane rather than merely a UAV.

Cheers !

Rana

Moderator Comment by Gary Mortimer on October 16, 2009 at 1:03am

Curt, pointing the nose at the ground is not the best way to make an aircraft land, even though its pointing the right way ;-) You should control the attitude with pitch and control the rate of descent with power. Pointing the nose down will just result in an increase of speed that you have to sort out in the flare.

PAT and APT should become phrases you think about when flying, be it full size or RC, applies just the same. Power Attitude Trim and Attitude Power Trim although in the USA something tells me they say Power Pitch Trim or some such ;-)

Comment by Tom in ON on October 16, 2009 at 4:24pm

http://diydrones.com/forum/topics/sharp-ir-distance-sensor

I had much better results when I was testing the Sharp IR sensor in comparison with the MAX EZ4 sonar. However, I am still a relative beginner, and often I get things working by hit and miss. I didn't do much with either yet sensors, and my project is on hold for now until I settle in after moving a couple of continents over.

The IR sensor had much faster and repeatable responses in a room environment. I think with proper conditioning both sensors may work adequately.


There is also a $60 self-contained temperature compensated barometer sensor that one member is offering here (Bosiak?) for sale that supposed to be acccurate enough for plus/minus 10cm (or less) up to 10,000 feet, if I remember correctly.

Comment by Krzysztof Bosak on October 17, 2009 at 2:33am

resolution to 13cm, but precision some 1m after 2min. it can be as good as 5m after 20min if wether changes.
rapid temperature changes discouraged (don't land after 'emergency dive') or loiter at low alt for 30-60s after that.
http://www.aerialrobotics.eu/altigps/altigps-en.pdf

Comment