Autonomous Flights with the UAV v2 Development Platform


The UAV Development Platform has been available on SparkFun's website for almost four years now. In that time, much progress has been made. A group of UAV pilots from across the globe have assembled and put their heads together to develop an extremely powerful control platform for RC airplanes and RC helicopters.

What sets the UAV dev board apart from others is that the control algorithms on the dsPIC controller are very precise even in heavy aerobatic maneuvers. The memory and speed on the dsPIC allows for rotational matrix math to be computed quickly. The method used is an implementation of a direction-cosine-matrix (DCM) algorithm.

An explanation of the DCM method is beyond the scope of this post, but the developers have written a great explanation found here.

To get an idea of how powerful this platform performs, check out the triple autonomous flight using a GentleLady sailplane with an electric motor.

Three autonomous flights were performed through take off and landing, with the closest landing within 6 meters of the target.


For an animated view, click here (you must have Google Earth installed).

The UAV dev board is constantly being revised with better and smaller hardware. Keep an eye on the product feed for new product updates!

If you are interested in working with a very lively and helpful UAV community or want more information about UAVs in general, be sure to check out DIYDrones.com.


Comments 18 comments

  • The late Jef Raskin had a cardboard slope soaring design, the Wester Wind. He had so many bundles of the die cut birds that we used to build them up then have “Cardboard Combat” above Pacifica.
    I’m curious about that landing within 6 meters. Were they volt meters or ameters? I bet they were digital multi-meters. Were they in a circle?

  • You might also be interested in reading about the flights that Rick Kuebler has made with his FunJet flying at 150 miles per hour, including a fully autonomous flight:
    http://www.diydrones.com/profiles/blogs/gt200kmh-funjet-the-fastest

    • The site says the “fastest uavdevboard plane?” Hmmm… question mark….I think this is a challenge:)

  • heh. I initially read the title as “Autonomous Fights with the UAV v2 Development Platform"
    First thought "WOW! Those sparkfun guys are WILD! Autonomous aircraft combat! Battlebots of the skies! I am SO looking forward to this contest!"
    Reread. Oh. Ok. (Dang.) :)

  • It is a nice board - I’m currently working on getting mine installed in a similar aircraft.
    But… Bill’s Gentle Lady has been fitted with an electric propulsion system as is evident from his comments, not to mention what a hell of an arm he would have to have to complete a course like that with one toss. It’s a cool autopilot but it isn’t magic. Yet.

    • Hi OldFar-SeeingArt,
      You are right, my GentleLady has been outfitted with electric, and it was in use for the autonomous flights.
      However, the field where I fly is on top of a hill with some great updrafts. Once I get the GentleLady up a little ways, I can shut off the motor and ride the air currents.
      Best regards,
      Bill

  • I totally support sparkfun UAV projects, but I’m also curious about how many other countries and people are interested in sparkfun’s UAV projects….. Namely Terrorists. Could be a very useful flying Improvised Explosive Device.

    • Because we should limit the availability and spread of knowledge due to fear of it’s misuse…
      Projects like this are breading grounds for future world-class engineers and intelligent individuals in general.

    • I think there’s an element of the human brain’s weakness risk perception here.
      You can mess around with a few grammes of something nasty in a flimsy but complex model aeroplane, or you can put 1.5 tonnes of fertiliser in the back of a pick-up truck and ram into something at 60mph.
      I know which one is the more dangerous to security both at home and abroad, but I bet you’re not leaving blog comments about how we must restrict the export of pick-up trucks.
      I’ve just rad some similarly hysterical comments on Hack-a-day about someone why tried to stabilise a tiny model rocket with some servo-powered fins. The reactions are understandable (if you’re willing to accept that many people can’t see the wood for the trees), but depressing.

    • We’re always concerned about how our products and projects are used (see Bluetooth post: http://www.sparkfun.com/commerce/news.php?id=308). I agree with Andrew on this one - it’s important to me to teach as much as possible. The potential outcomes of education are limitless - it’s up to us to show folks how to use the outcome for good.

    • while that might be possible the payload required for a workable IED payload is far to large for most uavs designed on this site. Further the equipment for a succesful uav would be difficult and expensive to obtain.

  • Bill,
    my main interest in making use of a UAV is for aerial mapping. Based on the plot for these flights and others for ArduPilot line keeping does not yet reach the requirements of an orderly survey with a lot of sag on the lines. Is this because this is not a priority in the navigation code(s) or is it that the systems have yet to mature to that stage?
    Mike

    • Hi Mike,
      The straightness of the lines is a function of the weather, the airframe, and the controls. In calm air, the lines are much straighter. What was happening in the flights that you are looking at is that the wind speed and direction was highly variable with location, and the plane that I was using was flying slow, so it was easily knocked off course. Some folks flying faster planes get much straighter lines. Finally, we are still working on the controls to improve the straightness.
      You might be interested to know that the firmware includes an option to point a camera at a desired location, so it would not be too difficult to arrange for the camera view to sweep along a perfectly straight line.
      Best regards,
      Bill

      • Bill, thanks for the explanation. Do you expect line holding to improve any time soon or is it more likely that it is already as good as it gets?
        Cheers,
        Mike

  • For those interested in the UAV DevBoard, I made a short introductory video of the development environment at this location:-
    http://www.vimeo.com/7120178

  • Slap a cell module in this thing and send it across the country:) My cell phone just flew away. It could be a super hero flying in to let people make a phone call;)


This Week

This Month

Heartbleed

Happy Arduino Day!