Shop OBEX P1 Docs P2 Docs Learn Events
A robot that flies like a bird — Parallax Forums

A robot that flies like a bird

Ron CzapalaRon Czapala Posts: 2,418
edited 2012-10-30 20:03 in General Discussion
Plenty of robots can fly -- but none can fly like a real bird. That is, until Markus Fischer and his team at Festo built SmartBird, a large, lightweight robot, modeled on a seagull, that flies by flapping its wings. A soaring demo fresh from TEDGlobal 2011.

http://www.ted.com/talks/a_robot_that_flies_like_a_bird.html

Comments

  • ercoerco Posts: 20,259
    edited 2012-10-28 10:41
    Sweet flappin' wings. I just love those dramatic TED flying demos where a crash will take out a spectator!
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-10-29 02:15
    Not all birds fly the same way. Seagulls are a good model as they soar on thermals and cover large distances. Little birds seem to mostly hop in and out of trees, so landing is more critical. And while butterflies and bees so fly, they tend to exhibit a lot of coping with turbulence more than getting from here to there. Not sure where bats and dragonflies fit in - both have we appealing flying skills, but specialized attributes.
  • Mark_TMark_T Posts: 1,981
    edited 2012-10-29 12:25
    Sweet! I can see a new craze in robotics started by that.

    Have to say as soon as I saw the subject I thought "that'll be those Festo guys again" - having seen their wonderful bio-mimetic robots before. For instance http://www.youtube.com/watch?v=F_citFkSNtk
  • ercoerco Posts: 20,259
    edited 2012-10-29 14:43
    Loopy Byteloose wrote: »
    And while butterflies and bees so fly, they tend to exhibit a lot of coping with turbulence more than getting from here to there.

    I am amazed every time I watch a butterfly or moth in flight. Their wing motion is completely different from a bird and provides a very erratic-looking flight path. Imagine it from their point of view. Talk about the ultimate in image stabilization and path planning to fly from A to B.
  • Dr_AculaDr_Acula Posts: 5,484
    edited 2012-10-29 15:11
    That is very impressive.

    I wonder what sort of motors they are using? Would a servo be powerful enough or would they use a brushless motor with a gearbox?

    It looks like each wing has two motors so that would give you directional control.

    Have we enough collective smarts on this forum to build this?
  • skylightskylight Posts: 1,915
    edited 2012-10-29 16:06
    Dr_Acula wrote: »
    That is very impressive.

    I wonder what sort of motors they are using? Would a servo be powerful enough or would they use a brushless motor with a gearbox?

    It looks like each wing has two motors so that would give you directional control.

    Have we enough collective smarts on this forum to build this?
    I could paint the feathers on :blank:
  • prof_brainoprof_braino Posts: 4,313
    edited 2012-10-29 17:45
    Dr_Acula wrote: »
    It looks like each wing has two motors so that would give you directional control.

    Have we enough collective smarts on this forum to build this?

    Sure, but I want to build one, too!

    As far as I can tell, there is only ONE motor that drives both wings, and the folding and rotation of the wings is a function of the linkage, and steers is controlled by the direction of the beak and tail, like in real birds. But I'm just guessing.
  • Dr_AculaDr_Acula Posts: 5,484
    edited 2012-10-29 20:23
    As far as I can tell, there is only ONE motor that drives both wings, and the folding and rotation of the wings is a function of the linkage, and steers is controlled by the direction of the beak and tail, like in real birds. But I'm just guessing.

    I tried looking for some technical details but couldn't find any with a quick search. I agree, if I was designing this I'd have a single powerful brushless motor through a gearbox running to linkages to both wings, and then I'd add and subtract a little from that movement with individual servos. I think you need rotation as well as being able to alter the lift on either side (eg if a gust of wind catches it on one side and tries to flip it over).

    Actually, for full control you want to replicate all the muscles of a bird. Particularly when coming in to land and flapping the wings fast for a hover. I am more and more amazed at the way birds come in to land on my bird feeder, dodging between branches and ducking between narrow gaps.

    ? 6 servos, 3 on each side, - one for amount of wing flap, one for rotation of the wing and one for the tail? It adds up to a fair bit of weight though.
  • prof_brainoprof_braino Posts: 4,313
    edited 2012-10-29 20:37
    Dr_Acula wrote: »
    I think you need rotation as well as being able to alter the lift on either side (eg if a gust of wind catches it on one side and tries to flip it over).

    ? 6 servos, 3 on each side, - one for amount of wing flap, one for rotation of the wing and one for the tail? It adds up to a fair bit of weight though.

    Weight is the enemy of things that fly. This one doesn't take off and land, and looking at the guts in the vid I think its 1 motor and 1 servo. And the battery only lasts for 90 seconds. But I still want one.

    The german bird looks a bit more expensive than the french one.

    http://en.videobuzzy.com/Avitron-RC-Ornithopter:-Remontely-Controlled-bird-341.news
  • msrobotsmsrobots Posts: 3,709
    edited 2012-10-29 20:42
    At the introducion the guy said something about pneumatics, so air I guess

    Enjoy!

    Mike
  • skylightskylight Posts: 1,915
    edited 2012-10-29 23:32
    The robot looked like an inflatable so perhaps was it filled by helium to help with the weight problem?
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2012-10-30 07:29
    I suspect that they use brushless motors and gear according to there specific needs. Do you see the big gears on the interior.

    Directional control is an interesting issue as seagulls do heavily depend of their tail movements to assisting in turning, but it seems the presentation is ignoring what the tail actually does.

    Can we build one in this Forum? The structure looks like carbon fibre, then there are those gears that look custom plastic, and the interior structure of the wings may have been done via some sort of 3D printer. Added to all that is shaping the outer shell from some sort of foam sheet plastic via vacumn molding.

    I think we have the smarts. Shared resources is another thing.
  • prof_brainoprof_braino Posts: 4,313
    edited 2012-10-30 12:25
    head is for steering tail is mostly for up and down, my feeling from the vid.

    if anyone wants to draw a frame, i will print it if i can get it into stl

    i will also buy a brushless air craft motor i have been wanting one for months

    how hard could it be?
  • sylvie369sylvie369 Posts: 1,622
    edited 2012-10-30 20:03
    Excellent! No more cleaning bird cr*p off my car!
Sign In or Register to comment.