Loading ...
Sorry, an error occurred while loading the content.

Re: Are Hobby Servos Stunting Robotics?

Expand Messages
  • dan michaels
    ... the table top scale. The real question is what is the answer? What is there for an easily scaleable, extendable robotics system. Maybe there needs to be
    Message 1 of 401 , Nov 3, 2007
    • 0 Attachment
      --- In SeattleRobotics@yahoogroups.com, "Naphtali Moore"
      <nrmstudios@...> wrote:
      >

      >
      > The existing hobby systems like lego mind storm only enforce
      the "table top" scale. The real question is what is the answer? What
      is there for an easily scaleable, extendable robotics system. Maybe
      there needs to be some sort of open yet consolidated standardization.
      >


      Most of these arguments totally dissipate if one puts things into the
      perspective of the "learning curve". Lego kits, servos, etc, make it
      very easy for people to become involved in an area [robotics] that
      requires learned aptitude in many different subareas - namely,
      electronics, mechanics, sensors, motors, programming, basics of AI,
      etc. It makes no sense to throw the baby out with the bath water
      [dpa].

      Plus, I think, there are many moves being made towards
      standardization, although not all are viewed in that way at present.
      First, you have the Microsoft Robotics Studio, plus several efforts
      going on in sourceforge. Then, you have all of the various kits - the
      general ones like Lego, Vex, Bioloid, and the more specific ones,
      like from Lynxmotion, like Aibo, and all of the humanoid robots
      currently coming out of Japan.

      There must be an enormous amount of $$$$$ being spent, by people we
      don't know about, to support those markets. My guess is that "most"
      of the people buying those things are not true roboticists, but
      rather dapplers with more money than general skills in all the areas
      necessary to produce sophisticated robots from scratch. So, they buy
      a prefab kit, and concentrate on mainly one area - like programming.

      Then, there are many people who are taking the "next step", like
      Robert with his robomagellans, George and his wifi router-equipped
      tank, and even dpa with jBot [nothing is stunting dpa's efforts ;-)].

      So, what's all the stew about? All progress happens according to
      pyramid-shaped demographics. A lot of people do simpler sorts of
      things along the bottom of the pyramid, and a few people push the
      envelope near the top.
    • dan michaels
      ... satisfaction problem. The constraints include keeping the CoG inside the triangles - calculate where the CoG is in the triangle by measuring the force on
      Message 401 of 401 , Dec 29, 2007
      • 0 Attachment
        --- In SeattleRobotics@yahoogroups.com, PeterBalch <PeterBalch@...>
        wrote:
        >

        >
        > IMHO, you're on the right track. Treat walking as a constraint-
        satisfaction problem. The constraints include keeping the CoG inside
        the triangles - calculate where the CoG is in the triangle by
        measuring the force on each foot then adjust the overall posture.
        >Don't pick up a foot if the force on it is large.
        >

        Yes.


        >
        >Keep theCoG moving forward. Don't let the CoG aproach the fron
        > of the current triangle unless the Flight foot is approaching
        touchdown. Maintain roughly the correct phase relationships between
        feet.
        >


        Yes. This is what you get with the Gait 6 step sequence, and which is
        similar to the diagonal-walk of Jake on my page, and also the trot.
        This is especially clear with Jake, where the sequence is essentially
        one of falling-forward onto the new tripod. With the trot, the
        forward inertia of the CoG performs the same function.

        These are all variations on the same thing, which is the back-to-
        front "wave". You [Peter] might not see it as a "wave", but it's
        really the same step sequence as goes on in hexapods and also
        octopods, except here there are fewer feet.


        >
        > The problem is, I don't see any easy way to reconcile all those
        disparate goals. In my imagination, the constraint satisfaction
        algorithm that made it walk on the flat happens to work equally well
        on rough ground.
        >


        As I see it, this is because creep, diagonal-walk [Jake's], and trot
        are all just variations on the same sequence, just like metachronal-
        wave, ripple, and tripod are for hexapods.
      Your message has been successfully submitted and would be delivered to recipients shortly.