Blogs

T3

UAV flight in the rain, automatic landing on the airstrip

Posted by Krzysztof Bosak on October 11, 2009 at 1:00pm

UAV flight in the rain, automatic landing on the airstrip from Krzysztof Bosak on Vimeo.

This time a flight in the rain.Should be a part of 'National UAV Mission Readiness Enhancement Programme'.If there would be any development.Note a few details:-Landing along the 'runway', fully automatic-Nice agreement with preflight mission simulation-No dramatic events due to rain thanks to IMU sensor-Low-tech video reception, almost a trademark. Sorry, my focus is not FPV, but fully autonomous photoshooting-Roll stabilisation failed because the pitot tube got water in it, lowering perceived airspeed and diverging roll angle. Using standard servo didnt helped - 45deg/s is not enough.
Tags: flexipilot, uav, rain
E-mail me when people leave their comments –
Follow

You need to be a member of diydrones to add comments!

Join diydrones

Comments

  • T3
    Krzysztof Bosak October 12, 2009 at 12:53pm
    Jack, I wanted to spare you the looks of this installation, but you asked for it.
    The buildup of water was not dramatic as the underside of the fuselage clearly shows advanced water drain system using patented porous materials Made in Germany.
    The autopilot flew as you see it and I must say putting a foam under it was a good idea: after a hazardous flight, the System Operator might be sweating and a wet sponge would be very handy in those circumstances.

    To be fair, today's flight preps ended the operations quick: the sponge got very wet after the first landing in manual mode, shorted sensors voltage regulator and the Ready LED reported IMU malfunction, requiring drying it. It works fine but the sponge must go. It would be clever to cover the autopilot bay but the reason of not doing so is that I wanted to test the barometer response so I replicated conditions from good weather.
    The regular EasyUAV has nearly rainproof (not fully waterproof) sealing using dense foam - what I am really testing here are technologies from 'Careless Systems Intl'.
  • Jack Crossfire October 12, 2009 at 10:28am
    You haven't photographed any of the weatherproofing or the fuselage but we assume the fuselage is not water tight & only the electronics are wrapped in packing tape. It probably gains a lot of weight if water accumulates in the fuselage.
  • T3
    Krzysztof Bosak October 12, 2009 at 10:01am
    My autopilot has separate analog switches for RC override - all uC and sensors can fail but the analog multiplexers (by luck I got them in mil grade -55C...125C even if this is not that necessary) will always switch back to manual RC. This makes some bonus confidence when flying withing controllable range, as RC receiver is buried deep in the fuselage and well protected - is more likely to fail because of temperature but not due to humidity, unless RC range gets reduced. This protection costs PCB space, however. The scary part is RC transmitter which picks up water, therefore I have kept it under the car until the emergency (which didn't arrived, maybe a little disappointment because something should fail first).
  • T3
    Krzysztof Bosak October 12, 2009 at 9:55am
    the servos are covered, the easystar dervo hinges go to the back.
    the esc is simply wrapped by fiberglass scotch.
    surprisingly this worked for total 40min in the rain build upon 2 flights, plus one short landing demo.
  • T3
    Brian Wolfe October 12, 2009 at 9:47am
    Nice job Krzysztof!. What sort of weather proofing have you done. I've used my RC car in the rain and had the speed controller literally burst into flames (which quickly went out because of the rain). I would think servos would be vulnerable too.

    Brian
  • T3
    Krzysztof Bosak October 12, 2009 at 9:15am
    hawk,
    -the plan was to stay at home when it rains. It requires really solid navigation system
    -EasyUAV is intended to be used even without groundstation, but the weather might change when returning from long range flight. Instead of making 100 flights in mild rain, I prefer to make 20 in heaviest rain possible.
    -I was afraid that the humidity will affect this specific pressure sensor readings too much. It wasn't the case.
    -Also, thermopile-based autopilots are known from having problems among clouds. Reliability makes a difference but must be demonstrated,
  • hawk October 12, 2009 at 6:58am
    good but you do for why?
  • T3
    Krzysztof Bosak October 11, 2009 at 11:33pm
    But this also requires pinpoint parachute landing on tiny deployment sites. And mission-specific platform.
  • T3
    Rory Paul October 11, 2009 at 3:39pm
    Nice. Bad weather flying is the type of thing you would need for search and rescue work!
  • T3
    Krzysztof Bosak October 11, 2009 at 3:26pm
    My IMU has Kalman and filters out gyro bias over entire temprange. It flew many times last winter and in fact was close to flying without any centrifugal compensation at all. The barometer is temp compensated.
    There is an option using temp compensation output of analog gyros, but those make a bulky 3D setup (FLEXIPILOT is flat except servo output pins and I prefer keep it that way, even resigning from airspeed sensor).
This reply was deleted.