Shop OBEX P1 Docs P2 Docs Learn Events
Parallax Humanoid - The next generation of robots? — Parallax Forums

Parallax Humanoid - The next generation of robots?

mklrobomklrobo Posts: 420
edited 2014-07-16 17:49 in General Discussion
:nerd: with all this computational power Raspberry Pi and Popeller, (etc), and the advent of a life size robot for sale in September @ wallmart, is
the Humanoid robot platform an inevitability? Will hobbiests start developing the legs, arms, etc, to fit together in a software platform? What will
the purpose be? (engineering assistant, of course!) :lol:

Comments

  • Heater.Heater. Posts: 21,230
    edited 2014-07-15 15:39
    mklrobo
    ...with all this computational power Raspberry Pi and Popeller (etc)...
    Which is like almost none. I doubt that a humanoid robot powered by such a platform could even tie it's own boot laces.
  • mklrobomklrobo Posts: 420
    edited 2014-07-15 15:53
    I have faith in Parallax, coupled with the Raspberry Pi. I have looked at my Robosapien, and it does not have the power of the RasPeller. There are beaglebone robots
    walking around. Surely, a platform could be made. They have made the Quadcopter, that has great balance. Would it not be a matter of programming? More power and
    software capability is avaliable now, more that ever. BUT, are people afraid of life- like robots anyone can build? The future is upon us...........:lol:
  • NWCCTVNWCCTV Posts: 3,629
    edited 2014-07-15 16:11
    I doubt that a humanoid robot powered by such a platform could even tie it's own boot laces.
    That is why man made Velcro!!!!
  • mklrobomklrobo Posts: 420
    edited 2014-07-15 16:15
    mklrobo wrote: »
    BUT, are people afraid of life- like robots anyone can build? The future is upon us...........:lol:
    The Hero was made in 1984, and that was mobile, and had an arm to use. Technology has grown considerably since then. There
    might have to be a Raspeller for the legs, one for each arm, and then one for the head.(comm, sensor database,) Another
    project to build......all right! :lol:
  • mklrobomklrobo Posts: 420
    edited 2014-07-15 16:32
    Heater. wrote: »
    I doubt that a humanoid robot powered by such a platform could even tie it's own boot laces.

    Tying boot laces is a complex operation, so you might be right, in that case. However, as my picture at the left, realistic uses for this type of
    robot were "truely" depicted in the movie, "Silent Running". Playing cards, welding, assembly, storing items logistically, modular system repair,
    and planting trees (surgery, but human supervisor) were some of the duties, and that was back then. To be sure, with the talented people in this
    forum, a life-size robot could be built. The journey of a thousand miles begins with one step..........:smile:
  • mklrobomklrobo Posts: 420
    edited 2014-07-15 17:04
    Heater. wrote: »
    mklrobo

    could even tie it's own boot laces.

    Consider this;
    Parallax made its square bot, and everyone has made software and build parts on that, so that now everyone could
    build whatever sensors/devices they need for their robot, at this point.
    If a "generic" life-like frame was adopted, like the Boe bot, hobbiests could build upon the part that they wanted to,
    then contributed their work on this forum. Even tying boot laces would considered to be one, of many, procedures
    in the humannoid library. Different walks, talks, abilities, would be platform libraries in which the user could
    envoke, from his computer at home which would serve as a library "server".
    Name the Humanoid, Chip, after the man writing the propeller assembly code.:cool:
  • xanaduxanadu Posts: 3,347
    edited 2014-07-15 17:11
    Not only would processing power be demanding, power would as well. Lanky limbs moving around like that it's going to have a short run time, be weak or run on a fuel cell that gives everyone around it cancer.

    I'm not even sure I understand why anyone would build a humanoid robot. For the costs and efforts involved it doesn't seem worth it. Two long legs and two arms with a high center of gravity seems like a bad design for any robot.

    I guess the R&D yields all kinds of cool stuff, and it is fun to look at. Functionality wise, until anti-gravity comes out, I'd rather have four legs and one arm.

    Got a link to the life size robot at Walmart?
  • Heater.Heater. Posts: 21,230
    edited 2014-07-15 17:17
    I have sometimes wondered about "Is it even sensible to build robots in human form?"

    Let's assume we could build a machine in a very human like form and that it has the smarts to take over all the jobs that humans do. After we have built about 7 billion of them we can all retire to the beach and let them provide for all our wants and needs.

    However, I suggest that at least initially and for a long time our engineering skills will not be able to build such a machine that is as energy efficient as an actual human. To be generous lets say the machine takes ten times as much energy as a human to do all that walking and working.

    That means the energy consumption required to use 7 billion of such robots is an order of magnitude greater than our current energy demands.

    This is just not workable. It's much more efficient, i.e. cheaper, to use actual humans than humanoid robots.
  • mklrobomklrobo Posts: 420
    edited 2014-07-15 19:18
    Heater. wrote: »
    I have sometimes wondered about "Is it even sensible to build robots in human form?"

    Let's assume we could build a machine in a very human like form and that it has the smarts to take over all the jobs that humans do. After we have built about 7 billion of them we can all retire to the beach and let them provide for all our wants and needs.

    However, I suggest that at least initially and for a long time our engineering skills will not be able to build such a machine that is as energy efficient as an actual human. To be generous lets say the machine takes ten times as much energy as a human to do all that walking and working.

    That means the energy consumption required to use 7 billion of such robots is an order of magnitude greater than our current energy demands.

    This is just not workable. It's much more efficient, i.e. cheaper, to use actual humans than humanoid robots.

    True, True. it all comes down to money. I guess the robots will stick to the hazardous and repetative jobs they are doing now. The only exception would be alot of
    people not wanting to work. People who have money and need work done, will need robots.(AND the techs to repair them) The energy needed would be astronomical.
    Or, a more efficent robot not of human form, but of a different design.:innocent:
    I will have to find the link to the Wallmart advertisement, for the female robot.:frown:
  • PJAllenPJAllen Banned Posts: 5,065
    edited 2014-07-15 19:51
    mklrobo wrote: »
    I will have to find the link to the Wallmart advertisement, for the female robot.:frown:

    Oh my.
  • PropGuy2PropGuy2 Posts: 360
    edited 2014-07-15 20:28
    A humanoid robot would need a new type of muscle actuator. I have been working on a silicone membrane low voltage (5 to 10 volt) actuator for awhile now and progress is slow. Not a trivial engineering problem. Maybe someone else has some insight on how to do this.
  • mklrobomklrobo Posts: 420
    edited 2014-07-16 04:06
    PropGuy2 wrote: »
    A humanoid robot would need a new type of muscle actuator. I have been working on a silicone membrane low voltage (5 to 10 volt) actuator for awhile now and progress is slow. Not a trivial engineering problem. Maybe someone else has some insight on how to do this.
    :cool: Cool! I will have to look into that. Different materials react, expansion and contraction, to the input you are using. The muscle wires heat up quickly,
    giving a phase change to the metal. In this phase change, the metal constricts. Can you provide any link to the silicone membrane documentation? I have
    been playing (on calculations) with the concept of an organic CPU. The two might be related.
  • PropGuy2PropGuy2 Posts: 360
    edited 2014-07-16 04:50
    From an Internet search I found a very high voltage (4000 VDC) "muscle" project, but it is probably not practicable unless you can get the voltage down to a few hundred volts or so. And the range of contraction is also small. The other option is to use a salt solution thru a membrane something like an insect eating plant (Venus flytrap) So far the best option seems to be to use this method. Transferring an concentrated ion salt solution thru a membrane dependent on the polarity of the electric charge. Reversing the reaction is troublesome though. I suppose a bio engineer and a dedicated engineering team could come up with something, but so far nothing practicable. I can imagine hundreds of applications, so it would definitely be a game changer for anyone who could figure this out.
  • mklrobomklrobo Posts: 420
    edited 2014-07-16 05:19
    PropGuy2 wrote: »
    From an Internet search I found a very high voltage (4000 VDC) "muscle" project, but it is probably not practicable unless you can get the voltage down to a few hundred volts or so. And the range of contraction is also small. The other option is to use a salt solution thru a membrane something like an insect eating plant (Venus flytrap) So far the best option seems to be to use this method. Transferring an concentrated ion salt solution thru a membrane dependent on the polarity of the electric charge. Reversing the reaction is troublesome though. I suppose a bio engineer and a dedicated engineering team could come up with something, but so far nothing practicable. I can imagine hundreds of applications, so it would definitely be a game changer for anyone who could figure this out.
    No doubt, this is a game changer. I am a student of chemical engineering, so I am inclined to work on the project, along with mine. This will take some time. The science your are
    targeting to use, is called material science. A material scientist examines the nessesary elements to be used, develops a dedicated formula focused on the Equation of State of
    the process. I have been playing with light on Jellyfish for memory processing.(materials that is) Movement could also be used, with control. I am working on trying to print out a model of
    a neural net on my printer paper, to set up a matrix. (Toner conducts, providing) After the neural net is proven, I can move to the next process.:smile:
  • kwinnkwinn Posts: 8,697
    edited 2014-07-16 05:51
    Heater. wrote: »
    mklrobo

    Which is like almost none. I doubt that a humanoid robot powered by such a platform could even tie it's own boot laces.

    +1 It's surprising how under appreciated the human (and animal) brain and nervous system is.

    Can you even imagine a robot that would be capable of visually identify a small airborne object, estimating it's trajectory, moving to the location it is expected to land, and catching it before it hits the ground? Yet young children, dogs, and various other animals can catch balls, frisbee's, and other objects with ease.
  • kwinnkwinn Posts: 8,697
    edited 2014-07-16 05:55
    xanadu wrote: »
    Not only would processing power be demanding, power would as well. Lanky limbs moving around like that it's going to have a short run time, be weak or run on a fuel cell that gives everyone around it cancer.

    I'm not even sure I understand why anyone would build a humanoid robot. For the costs and efforts involved it doesn't seem worth it. Two long legs and two arms with a high center of gravity seems like a bad design for any robot.

    I guess the R&D yields all kinds of cool stuff, and it is fun to look at. Functionality wise, until anti-gravity comes out, I'd rather have four legs and one arm.

    Got a link to the life size robot at Walmart?

    Why would a fuel cell give anyone cancer? It is a device that converts a hydrocarbon or hydrogen fuel into water and carbon dioxide to produce electricity.
  • kwinnkwinn Posts: 8,697
    edited 2014-07-16 05:57
    PropGuy2 wrote: »
    A humanoid robot would need a new type of muscle actuator. I have been working on a silicone membrane low voltage (5 to 10 volt) actuator for awhile now and progress is slow. Not a trivial engineering problem. Maybe someone else has some insight on how to do this.

    Start by looking at how animals do it and try to do the same or something similar?
  • GadgetmanGadgetman Posts: 2,436
    edited 2014-07-16 06:04
  • kwinnkwinn Posts: 8,697
    edited 2014-07-16 06:08
    Heater. wrote: »
    I have sometimes wondered about "Is it even sensible to build robots in human form?"

    Let's assume we could build a machine in a very human like form and that it has the smarts to take over all the jobs that humans do. After we have built about 7 billion of them we can all retire to the beach and let them provide for all our wants and needs.
    ...............................................................................

    Short of all out nuclear war, a comet, or an asteroid ending all life on earth this is the worst calamity that could happen. Can you even imagine what 7 billion bored beyond description people could come up with to entertain themselves? The end of civilization for sure.
  • Heater.Heater. Posts: 21,230
    edited 2014-07-16 06:29
    kwin,
    Can you even imagine a robot that would be capable of visually identify a small airborne object, estimating it's trajectory, moving to the location it is expected to land, and catching it before it hits the ground?
    No need to imagine. It's been done. Well sort of.
    Back in the 1980's I worked for a company involved in the development of the Seawolf missile. That thing could detect, track and hit an incoming artillery shell!

    Or then there is the ping-pong playing robots: http://www.youtube.com/watch?v=t_qN3dgYGqE

    But, In general yes, I agree with you.
  • ercoerco Posts: 20,256
    edited 2014-07-16 06:37
    See also:

    And
  • kwinnkwinn Posts: 8,697
    edited 2014-07-16 08:08
    @ Heater & erco

    Impressive achievements, but a far cry from what an athletic 10 year old can do. Missiles accomplish their targeting with equipment specifically designed for that one task, and the robots are performing under controlled conditions on top of having specialized hardware.

    When a bipedal robot can run down a field and catch a ball that was hit in a semi-random trajectory I'll be more impressed. IMHO we have a long way to go before humanoid robots can replace people in all but the most repetitive tasks.
  • prof_brainoprof_braino Posts: 4,313
    edited 2014-07-16 16:06
    PropGuy2 wrote: »
    A humanoid robot would need a new type of muscle actuator. I have been working on a silicone membrane low voltage (5 to 10 volt) actuator for awhile now and progress is slow. Not a trivial engineering problem. Maybe someone else has some insight on how to do this.

    I'm doing pneumatic. The micro valves are a bit large and heavy, but the accutators themselves are very light weight. They don't have a very large stroke.
  • prof_brainoprof_braino Posts: 4,313
    edited 2014-07-16 16:08
    Heater. wrote: »
    I doubt that a humanoid robot powered by such a platform could even tie it's own boot laces.

    If I say its impossible for a humanoid robot to tie boot laces, does that mean someone has to build one to prove me wrong? This actually sounds like a good place to start.
  • jazzedjazzed Posts: 11,803
    edited 2014-07-16 17:41
    mklrobo wrote: »
    Will hobbiests start developing the legs, arms, etc, to fit together in a software platform?


    And end up with something like this?


    Multi-armed-god-in-India.jpg&w=300&h=200&zc=1&a=c
  • mklrobomklrobo Posts: 420
    edited 2014-07-16 17:49
    PJ Allen wrote: »
    Oh my.

    :frown: I could not find that link yet, but I am working on it.:frown:
    somewhere on youtube.......
Sign In or Register to comment.