Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 



Forgot your password?
typodupeerror
×
Input Devices Science

Quadcopter Guided By Thought — Accurately 79

ananyo writes "A toy quadcopter can be steered through an obstacle course by thought alone. The aircraft's pilot operates it remotely using a cap of electrodes to detect brainwaves that are translated into commands. Ultimately, the developers of the mind-controlled copter hope to adapt their technology for directing artificial robotic limbs and other medical devices." From the paper (PDF) abstract: "... we report a novel experiment of BCI controlling a robotic quadcopter in three-dimensional (3D) physical space using noninvasive scalp electroencephalogram (EEG) in human subjects. We then quantify the performance of this system using metrics suitable for asynchronous BCI. Lastly, we examine the impact that the operation of a real world device has on subjects’ control in comparison to a 2D virtual cursor task. Approach. ... Individual subjects were able to accurately acquire up to 90.5% of all valid targets presented while traveling at an average straight-line speed of 0.69 m s^(1)." This also appears to be the first time a Brain-Computer Interface was used to operate a flying device in 3D space. Also, there are several additional videos showing people operating the quadcopter.

This discussion has been archived. No new comments can be posted.

Quadcopter Guided By Thought — Accurately

Comments Filter:
  • by xtal ( 49134 ) on Wednesday June 05, 2013 @08:07AM (#43913645)

    -nt-

  • by cold fjord ( 826450 ) on Wednesday June 05, 2013 @08:10AM (#43913683)

    Do what I say, not what I think!! (At least not that particular thought.)

  • by Anonymous Coward

    http://www.imdb.com/title/tt0083943 :)

    • You laugh but I'd love to see a TV show of teams fighting each other with these in a laser tag arena.

  • Really gotta get myself an EEG headset sometime...I wonder what other signals it could be tuned to pick up? What happens if you think about moving a nonexistant limb, would anything happen? That could free up your hands and feet and let you control other things with your mind.

    • Re:SO COOL (Score:5, Informative)

      by Aguazul2 ( 2591049 ) on Wednesday June 05, 2013 @08:26AM (#43913831)

      Really gotta get myself an EEG headset sometime...I wonder what other signals it could be tuned to pick up?

      Okay, then get ready for lots of goo. You need conductive paste for any kind of serious work. Also, there is the small matter of your skull getting in the way of the signals -- like listening to a symphony through ear protectors.

      (I was involved in the OpenEEG project for a time -- dreams of an efficient BCI (brain-computer interface) to beat my keyboard as an input device were soon shattered.)

      • Aw man I was really hoping the paste problem had been overcome :-(

        Still I'd consider something more invasive if it were useful and futureproof enough.

      • by cffrost ( 885375 )

        Really gotta get myself an EEG headset sometime...I wonder what other signals it could be tuned to pick up?

        Okay, then get ready for lots of goo. [...]

        Up to here I thought we were about to start talking about EEG-controlled fuckbots. :o(

    • The one they were using looked pretty funky, I wonder what the difference in performance is between that one and the cheapie toy ones you can get commercially? The Emotiv Epoc is $300 and apparently (with some hackery) can get you the EEG outputs that you usually have to buy the $700 one for.

      I can't help wondering where exactly this is on the scale between actually picking up specific thoughts (or at least muscle intentions) and just using some global clock signal (like that one that detected the change i
    • Thinking about moving an nonexistent limb is something that would happen naturally.. if that limb was cut off. If you really want a "third hand" controlled by motor nerve impulse (which is probably quite different that an EEG, and which is what I understand by "think about a limb") it would require extensive and long training, I think, like people learning to walk again. Or you would have to do this at a very young age and it would be nearly impossible with an adult brain (dunno).

      • Agreed; easier to link the signal to something else you can control in thought that doesn't distract too much; sub-vocalization of specific words would probably do the trick (although you couldn't talk while doing that) or linking it to imagining yourself floating, falling rising, going left, etc (which would be an extremely weak signal for an EEG to relay in any meaningful way).

    • by t4ng* ( 1092951 )
      These guys are a bit late to the party. The military is already providing thought controlled artificial limbs [thedianerehmshow.org] that are far more complex than a quad-copter.
  • Better idea... (Score:3, Interesting)

    by Anonymous Coward on Wednesday June 05, 2013 @08:18AM (#43913753)

    Screw the drone, hook him up to Quake 3 and lets see him pull off some sweet strafejumps...that'll be a better benchmark :D

    Also i'm thinking that perhaps they should focus more on feedback. If visual feedback is the only thing he gets, especially with the delay, it'll take him a lot longer than if he had say, rumblepack belts attached to him which gave feedback on direction, tilt, speed etc.

  • I can see this as a natural progression.
    First we have direct brain control for very basic, then it gets more detailed. Finally we can get to less invasive methods of monitoring the brain activity.
    Once we get that good, I could see getting more details for more finer type of movements. After we get a good map of the movement sectors of the brain, we may want to scan other sections such a vision we we can record what we see, or hearing. Perhaps we could scan our language processing areas and make real time

    • by wagnerrp ( 1305589 ) on Wednesday June 05, 2013 @09:54AM (#43914621)
      You're thinking about this the wrong way. Most of these systems are like training your brain how to walk again after a serious trauma. You just hook up electrodes to a patch of cortex and come up with some unique brain pattern that you will use as a trigger for a certain task. You program the computer to respond to that pattern, providing a feedback response. Using the feedback, you train your brain to more accurately produce that pattern, while you simultaneously tune the computer to pick up on the evolving pattern. These are not the kinds of things you just plug in and go, they take considerable time and effort to make work for each individual.
      • The fact that you have to learn how to produce these patterns makes me wonder what would happen if you "attached" these to a very young infant. From what I have read, their brains are wiring themselves based off of responses to their actions. I wonder if the thought controlled robot could become an extension of their own body if they are introduced to it at a very young age. I imagine it's harder to learn to control these actions once your brain has a basic understanding of what does what, hence the lear
        • From what I have read, their brains are wiring themselves based off of responses to their actions.

          So is yours. This idea that infants are unique in the way they learn doesn't do the adult brain justice. My daughter has been teaching herself English via total immersion for 2 years (with of course, help from mom and dad and everyone else around her) I could match her progress in some other language in a matter of weeks (thanks largely to the enormously complex framework that my mind already contains).

          The point is, you have a lot of tools at your disposal that the infant doesn't. You know what your goal

        • by JTsyo ( 1338447 )
          Trouble with that is that the infant wouldn't know to try to use it. So you would have to just pick arbitrary commands for the computer to watch for. The infant might never think them or even if he does, he might not realize the link. Think of it as recording a marco, you need to do it the first time before you can use it.
      • Yes. As is visible in the video, the amount of control the user actually has is deplorable.
        The whole obstacle course looks nice, but the copter is never actually shown to do anything in it for more than 0.5s. And that is with a thoroughly trained system.

      • Thankfully my brain hasn't yet learned to walk, that sounds like it could cause some serious problems.
    • by t4ng* ( 1092951 ) on Wednesday June 05, 2013 @11:01AM (#43915303)
      The Modular Prosthetic Limb [jhuapl.edu], developed by the Johns Hopkins University Applied Physics Laboratory with funding from the Defense Advanced Research Projects Agency provides 26 degrees of motion, including independent movement of each finger, in a package that weighs about nine pounds and has the dexterity of a natural limb. In 2012, a patient at the University of Pittsburgh Medical Center successfully demonstrated that the arm could be controlled by the user’s thoughts. Several patients, including a decorated Afghanistan war hero, are helping researchers further develop the prosthesis. In 2013, the MPL will continue to be tested and refined in a clinical trial at the California Institute of Technology.
  • Personally I think the hand is still the best tool we have to do physical manipulation. That is what it was designed for and it is excellent. Since we don't use our 'minds' to move objects normally, I think the hand, or a hand-guided tool, will always be the best.

    • Luckily there are people who are not as narrow-minded as you to move us forward.

      1) Not everybody has hands or control over them. Technology like this will be priceless for amputees and people with muscular or other degenerative diseases.

      2) This is a first generation technology, as systems like this are refined, they can potentially exceed the range of our physical control through cutting response time and scaling up the "granularity" of control while eliminating "interference" (e.g. the involuntary shak
    • You know, you use your "mind" to control your hands, right?
    • Ten thousand years ago, I bet you would've been that caveman skeptic crapping all over Grog's new fandangled "wheel."

      • Ten thousand years ago, I bet you would've been that caveman skeptic crapping all over Grog's new fandangled "wheel."

        Grog wheel no good! Look, it go 'round and 'round; it start here and then come back here! How that work? Thog no want go 'round; Thog want go places! Grog wheel stupid.

        And what good Grog wheel? Thog want go somewhere, Thog walk. Thog want go somewhere faster, Thog run!

        Grog say wheel help carry heavy thing, but Grog wheel made of stone; Grog wheel /is/ heavy! Thog need wheel just to carry whee

  • Now to hook this up to my flying car...

  • Comment removed based on user account deletion
    • Cats will then control the world. Be careful what you wish for.

    • That sounds excellent.
    • by tgd ( 2822 )

      All this is fine, but can they make a version controllable by cat's thought?

      I saw the subject of your post and thought you were looking for one you can use to control a cat.

      At 4am, that is something I'd find very useful.

    • Oh! We're all going to die when that happens...
    • All this is fine, but can they make a version controllable by cat's thought?

      Shouldn't be too hard; they can modify the already-existing thought-control mechanism* used by felines to control their bipedal drones.

      Cat glances meaningfully at food bowl. "Hungry"
      Human gets up and puts food in bowl.
      Cat rolls over on back and thinks "Rub"
      Human stops what he is doing and dutifully scrubs the inviting tummy.

      * may not be compatible with all drones.

  • by Anonymous Coward

    The roughly 500 msec phase delay from having to average the very noisy electrical signals detectable on the skin into a valid, usable motor controller was the failing point of the Boston Arm, and seems to be the problem of this system as well. It's just not going to be as quick as something that's tied directly to the physical neural structure and signal processing built into normal nerve-muscle interfaces, which are specific, complex, adaptive, and generated in a physical 3-D structure with its own process

  • As can be seen, the drone used is a Parrot.
    The Parrot has pretty insane automatic correction so that it always stays stable and make smooth turns.

    Try moving like that while controlling the drone directly.

  • by pla ( 258480 )
    Now make it available to "pilot" quadruplegic people.

    The military already has basic mech-suits. Combine it with a functional BCI, and we've effectively made one of the most horrific of human injuries a mere nuisance.

    That said, as an able-bodied person, I'll still gladly take a few BCI-enabled toys - When can I get one of these?
  • you know, electrode caps really scare the crap outta me, because its not hard to see how a device used to read your brain signals could be modified to INSERT them too. as governments and take-over-the-world types learn more and more about how the brain works, wouldnt it be a great plan to use brain-caps to control peoples actions and turn them into the true mind-numbed robots that pop culture like to report on so..

    it would an interesting career to experiment with sending signals INTO the brain...for the ki

    • because its not hard to see how a device used to read your brain signals could be modified to INSERT them too

      Yes. Yes, it really is. This isn't the computer reading your mind, so much as it is your mind learning how to talk to the computer. There is a steep learning curve as your mind rewires itself to be able to produce the kinds of signals the EEG is set up to trigger off of. There is no way this could operate in reverse with our current level of technology and understanding.

  • Wow this is impressive. Just what I needed for Christmas, a brain cancer. lol
  • They want their "Biofeedback" fad back.

    Though it could have been newsworthy, if they used Raspberry Pi in 3D-printed box to implement the control algorithms.

  • "If thoughts could ^H^H^H^H^H^H^H^H^H^H^H^H^H^H^HWhen thoughts can kill" this may not be seen as such a great idea.
  • Except for the obvious disabled folks can more stuff, why is thought control better then just using you thumbs on a joystick? Seems to me it's much harder to put on a scull cap, try to think of just one thing for a long time and so on verses picking up a controller and flying the thing. Or, do we intend to have an army amputees piloting drones in the future?
  • Far fewer idiotic public appearances by polititians and celebs - at least not without covering ack-ack.
  • The person doing the controlling and the robot don't need to be in the same location...it could be that a robot could effect repairs in space controlled by a person on the ground by thought, or a surgeon could similarly control a robot to work on battlefield casualties.

  • ... I'm going to hook this up to my dog.

  • The paper [iop.org] (PDF, 1.7MB)
  • ... tomorrow it's Iron Man suits.

    The future looks bright! And also a bit shiny.

I had the rare misfortune of being one of the first people to try and implement a PL/1 compiler. -- T. Cheatham

Working...