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

(possible) Robot Fodder

Expand Messages
  • Your Friendly Hogfather
    I ve never been particularly fond of Tilden s Robot Rules . If you come to think of it, his products over the last four years fail miserably in this area.
    Message 1 of 11 , Jul 29, 2006
    • 0 Attachment
      I've never been particularly fond of "Tilden's Robot Rules". If you come to think of it, his products over the last four years fail miserably in this area. They don't protect themselves in any meaningful way, they don't seek whatever is suppposed to be "good", and they don't attempt to expand their territory.

      However, even in the long distant past, the idea of a "desirable resource", something to be fought for, was light. As we know by now, until someone comes up with a hugely, stupendous development in the area of light->energy conversion that light isn't much of a resource.

      So, I was just wondering lately: what WOULD be decent "robot fodder", that is, something that BEAM robots could vie for and actually use once obtained. What came to me was button battery cells - the relatively thick sort (say, 5mm or so) spread on the good old "Robot Jurassic Park". One difficulty is to make sure that the battery was "upright" when obtained (or that the circuit connecting to such batteries could deal usefully with reversed polarity", but I don't think that is a terrible problem.

      And, of course, once a battery was no longer useful, it would be, um, excreted. Thus there would also need to be some sort of test to prevent "ingesting" useless fodder. Even the thought of "poison capsules" - something button-battery-like that initially produces what appears to be voltage and current but then suddenly shorts out comes to mind.

      In any case, given reasonably sized motors (not so small as pagers, for example), you could obtain a pretty fair amount of energy for having obtained one of these "pelllets".


      Just an idea on a quiet Satuurday afternoon,


      YfH
    • liefmichael
      ... come to think of it, his products over the last four years fail miserably in this area. They don t protect themselves in any meaningful way, they don t
      Message 2 of 11 , Jul 29, 2006
      • 0 Attachment
        --- In beam@yahoogroups.com, Your Friendly Hogfather <hogfather@...>
        wrote:
        >
        > I've never been particularly fond of "Tilden's Robot Rules". If you
        come to think of it, his products over the last four years fail
        miserably in this area. They don't protect themselves in any meaningful
        way, they don't seek whatever is suppposed to be "good", and they don't
        attempt to expand their territory.

        What about the light-seeking behaviour?

        >
        > However, even in the long distant past, the idea of a "desirable
        resource", something to be fought for, was light. As we know by now,
        until someone comes up with a hugely, stupendous development in the area
        of light->energy conversion that light isn't much of a resource.

        However, it is the only one we have naturally and feasibly available. I
        can get my hands on miniature hydrogen fuel cells which appear to have
        enough power to push a small bot, but any power advantage gained over
        solar would be offset by weight & size.

        >
        > So, I was just wondering lately: what WOULD be decent "robot fodder",
        that is, something that BEAM robots could vie for and actually use once
        obtained. What came to me was button battery cells - the relatively
        thick sort (say, 5mm or so) spread on the good old "Robot Jurassic
        Park". One difficulty is to make sure that the battery was "upright"
        when obtained (or that the circuit connecting to such batteries could
        deal usefully with reversed polarity", but I don't think that is a
        terrible problem.
        >
        > And, of course, once a battery was no longer useful, it would be, um,
        excreted. Thus there would also need to be some sort of test to prevent
        "ingesting" useless fodder. Even the thought of "poison capsules" -
        something button-battery-like that initially produces what appears to be
        voltage and current but then suddenly shorts out comes to mind.
        >
        > In any case, given reasonably sized motors (not so small as pagers,
        for example), you could obtain a pretty fair amount of energy for having
        obtained one of these "pelllets".
        >
        >
        > Just an idea on a quiet Satuurday afternoon,
        >
        >
        > YfH
        >

        Great idea, but there was reasoning behind the solar power extraction.
        I believe MT was attempting in a way to create a digital lifeform that
        suited and could survive with minimal/no interference once set loose on
        our current environment. A situation where we not only have to
        construct the lifeform but the world/environment for it to survive in is
        not in keeping with the original idea IMHO.

        An alternative and somewhat overarching power source for BEAM bots is
        .... US! We provide their life, their power etc Maybe we should do
        more work into human detection and interaction? Find a human when power
        runs low, flash a 'plug me in' signal. This means the robot is still
        hunting for food and initiating the power exchange?
      • chrisvaughan02
        ... power ... I like the idea of us as power source. when I was young, if I didnt feed my dog he would die, but that didnt make him less real. I agree that
        Message 3 of 11 , Jul 30, 2006
        • 0 Attachment
          > An alternative and somewhat overarching power source for BEAM bots is
          > .... US! We provide their life, their power etc Maybe we should do
          > more work into human detection and interaction? Find a human when
          power
          > runs low, flash a 'plug me in' signal. This means the robot is still
          > hunting for food and initiating the power exchange?
          >


          I like the idea of us as power source. when I was young, if I didnt
          feed my dog he would die, but that didnt make him less real. I agree
          that more work should be done with human detection. At my Job we have
          security doors that are held shut with Magnets which disengage when you
          approach the door. Now I'm pretty sure they work off of IR heat, but
          they are pretty accurate and discriminating. I've tried severyal other
          heat sources to see if it would open them and it wouldnt. they also
          seem to be triggered not only by heat but by motion. if I stand there
          for a minute the lock will re-engage until I move. perhaps that could
          be adopted my a bot. It'd be like a dog that come to see you when he's
          hungry :)

          God Bless,
          Chris
        • wilf_nv
          ... snip ... Hi Chris, That sounds like a job for PIR detectors aka a motion detector which, used together with a compound fresnell lens, can sense variations
          Message 4 of 11 , Jul 30, 2006
          • 0 Attachment
            --- In beam@yahoogroups.com, "chrisvaughan02" <cvaughan1@...> wrote:


            > At my Job we have security doors that are held shut with Magnets
            > which disengage when you approach the door.

            > Now I'm pretty sure they work off of IR heat, but they are
            > pretty accurate and discriminating. I

            snip

            > God Bless,
            > Chris

            Hi Chris,

            That sounds like a job for PIR detectors aka a motion detector which,
            used together with a compound fresnell lens, can sense variations in
            an IR image projected on its surface such as caused by a warm body in
            motion.

            Does the security system disable the automatic opening in case of
            emergency?

            Reason I ask is that it reminds me of design for an office elevator
            that used heat sensing call buttons which had a fatal flaw of
            reacting to a building fire by going to and automatically holding the
            elevator doors open on that floor.

            Speaking of work for food for thought: An important feedstock for
            Beam creations is converting bits of old technology into body and
            brain parts. How about a contest for the most complex beam design
            that is made entirely from recycled materials?

            wilf
          • chrisvaughan02
            ... Well in a fire I dont think the sensor would stand up to it. Its encased in a pvc shell mounted above the door, but I m not sure if it has an actual fire
            Message 5 of 11 , Jul 31, 2006
            • 0 Attachment
              > Does the security system disable the automatic opening in case of
              > emergency?

              Well in a fire I dont think the sensor would stand up to it. Its
              encased in a pvc shell mounted above the door, but I'm not sure if it
              has an actual fire response. We did have some manual switches put in
              next to the door though that when pressed will kill the power to the
              Magnet. I belive these had to be put in at our last fire inspection
              before we reopened after the hurricane.


              > Reason I ask is that it reminds me of design for an office elevator
              > that used heat sensing call buttons which had a fatal flaw of
              > reacting to a building fire by going to and automatically holding the
              > elevator doors open on that floor.

              That seems like a major flaw that should have been addressed before
              implementation. Hope no one got hurt as a result of it.

              chris
            • David T
              YFH, We BEAMers should be cheering the development of micro fuel cells, such as the direct methanol fuel cell . See, for example:
              Message 6 of 11 , Jul 31, 2006
              • 0 Attachment
                YFH,

                We BEAMers should be cheering the development of micro fuel cells, such as the "direct methanol fuel cell". See, for example:


                A BEAM butterfly, or similarly proboscis-equipped bot could sip methanol, lighter fluid, or a similar, less-toxic food, as nectar. Butterflies, ingesting only liquid food, have no need to, ummm, excrete anything. This is a somewhat more appealing scenario, and makes for fewer hygiene issues in the Robot Jurassic Park.

                The proboscis might be simply a capillary tube. Surface tension could bring fluid into the stomach, which would be a reservoir stuffed with cotton (like an old Ronson lighter).

                If/when an appropriate fuel cell becomes available, we could easily pack 10 x more energy into a bot than is available from a battery. That would open up many new behaviors, including autonomous, sustained flight.

                As an added bonus, a fuel cell powered bot would depend on atmospheric oxygen, just like we do. (In fairness, a zinc/air battery-powered bot would, too.) Air then becomes a resource the bot must have, just like food.

                Your ideas keep rolling around in my head. If they keep it up, something will eventually shake its way out. :^)

                :--David T.


                On Jul 29, 2006, at 7:44 PM, Your Friendly Hogfather wrote:

                I've never been particularly fond of "Tilden's Robot Rules". If you come to think of it, his products over the last four years fail miserably in this area. They don't protect themselves in any meaningful way, they don't seek whatever is suppposed to be "good", and they don't attempt to expand their territory.

                However, even in the long distant past, the idea of a "desirable resource", something to be fought for, was light. As we know by now, until someone comes up with a hugely, stupendous development in the area of light->energy conversion that light isn't much of a resource.

                So, I was just wondering lately: what WOULD be decent "robot fodder", that is, something that BEAM robots could vie for and actually use once obtained. What came to me was button battery cells - the relatively thick sort (say, 5mm or so) spread on the good old "Robot Jurassic Park". One difficulty is to make sure that the battery was "upright" when obtained (or that the circuit connecting to such batteries could deal usefully with reversed polarity", but I don't think that is a terrible problem.

                And, of course, once a battery was no longer useful, it would be, um, excreted. Thus there would also need to be some sort of test to prevent "ingesting" useless fodder. Even the thought of "poison capsules" - something button-battery-like that initially produces what appears to be voltage and current but then suddenly shorts out comes to mind.

                In any case, given reasonably sized motors (not so small as pagers, for example), you could obtain a pretty fair amount of energy for having obtained one of these "pelllets".

                Just an idea on a quiet Satuurday afternoon,

                YfH

              • J Wolfgang Goerlich
                This begs the question, how well did Tilden s LANL robots demonstrate his rules? Care to give us a history lesson, YFH? I recall Tilden taking a couple of his
                Message 7 of 11 , Jul 31, 2006
                • 0 Attachment
                  This begs the question, how well did Tilden's LANL robots demonstrate
                  his rules? Care to give us a history lesson, YFH?

                  I recall Tilden taking a couple of his robots out to the Yuma desert.
                  Still, most of what I read was of photovores in RJPs. The use of
                  carefully crafted parks seems irreconcilable with robots able to fend
                  for themselves.

                  J Wolfgang Goerlich

                  Your Friendly Hogfather <hogfather@...> wrote:
                  > I've never been particularly fond of "Tilden's Robot Rules". If you
                  > come to think of it, his products over the last four years fail
                  > miserably in this area. They don't protect themselves in any
                  > meaningful way, they don't seek whatever is supposed to be "good",
                  > and they don't attempt to expand their territory.
                • liefmichael
                  http://www.jaycar.com.au/productView.asp?ID=ZM9080&CATID=&keywords=proton&SPECIAL=&form=KEYWORD&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceMin=&priceMa
                  Message 8 of 11 , Jul 31, 2006
                  • 0 Attachment
                    http://www.jaycar.com.au/productView.asp?ID=ZM9080&CATID=&keywords=proton&SPECIAL=&form=KEYWORD&ProdCodeOnly=&Keyword1=&Keyword2=&pageNumber=&priceMin=&priceMax=&SUBCATID=

                    'proton exchange fuel cell'

                    micro hydrogen fuel cell (approx 60mm square x 20mm thick) 300mA /
                    30mW @ 0.65V - needs low pressure hydrogen & oxygen and DISTILLED (not
                    purified) water.
                  • chrisvaughan02
                    The use of ... Wolf, I would propose(as a Christian) that we live in a carefully crafted enviroment, which is suited to our unique needs. Without being overly
                    Message 9 of 11 , Jul 31, 2006
                    • 0 Attachment
                      The use of
                      > carefully crafted parks seems irreconcilable with robots able to fend
                      > for themselves.

                      Wolf,

                      I would propose(as a Christian) that we live in a carefully crafted
                      enviroment, which is suited to our unique needs. Without being overly
                      philosophical, I think that as we create "life" the end goal should not
                      be that it survive in our enviroment. I think that bots suited to
                      certain enviroments are great, I don't want it to appear otherwise.
                      But I think that in trying to create a more humanistic(is that a word:)
                      type of bot we may overlook some intermediate "species" or something
                      along those lines :)

                      I was reading an article on beam a while ago and it was about battery
                      opperated bots not being true beam because the bot depends on a human
                      to change the battery. However, isnt it plausible to believe that a
                      Bot could die? Even Humans have a finite life span. Some insects only
                      live for a week or even a day. What are your thoughts on that?

                      chris
                    • wilf
                      Hi Chris, I am also interested in your perspective on beam robotics. I believe that Tilden has a notion about the evolution of robots that mimics the process
                      Message 10 of 11 , Jul 31, 2006
                      • 0 Attachment
                        Hi Chris,
                         
                        I am also interested in your perspective on beam robotics. 
                         
                        I believe that Tilden has a notion about the evolution of robots that mimics the process of natural selection in biological evolution.
                         
                        For example in his early work he suggested, that simple but adaptive mechanisms,  interacting with a complex environment can cause emergent behaviours by which the mechanism survival in that environment is enhanced.   He suggested that swarms of simple mechanisms interacting with each other can lead to self-organization of  complex group behaviour ( the total is greater than the sum of the parts ) .
                         
                        Much of this would have been inspired by his experiments with simple analog robots that exhibited complex behaviour despite of, or perhaps because of, the simplicity of their design (free will?).  Rather than designing  a specific robot with a specific function, he would design a basic building block, circuit and mechanism as a kind of Beam DNA that had the potential to evolve into a more complex mechanism or system. 
                         
                        The non-linearity and chaotic processes of the evolution of Beam DNA would make the outcome of the form and function of the emerging Beam robot unpredictable but having survived the gauntlet of natural selection, its ability to survive would be unmistakable.
                         
                         In the hostile environment of the real world, the chances of survival of such fledgling robots are slim indeed but if millions were tested at a time ..... some may survive and adapt to niches that sustain their existence. One such niche might be to cannibalize other beam robots for their parts and energy source. (may the best tech win)
                         
                        In a simpler more benign controlled environment such as an RJP,  several species of simple  beam robots can survive long enough for observations to be made about relative strengths and weaknesses and some adaptive processes may be observed so that the  basic building block designs may be improved. 
                         
                        Tilden's foray into toy robotics is a kind of extension of his RJP experiments as he gets to test the survival and adaptation of millions of his creations in the environment of human homes and gardens.
                         
                        Now all he needs is a way to gather data to analyze the results of this meta RJP experiment, presumably when future swarms of  his Beam creatures connect themselves to the internet and send data streams from millions of tiny cameras and microphones  back to the hive to show domesticated beam robots which have learned to adapt to live harmoniously in their host's environments and feral beam robots which have had to adapt their host's environments to be more suitable to robots. 
                         
                        I could go on but several Robo Sapiens appear to be converging on me just and I should really press the Send button and get out of .....arrrrrgggghhhh .........   
                         
                         
                         
                        ----- Original Message -----
                        Sent: Monday, July 31, 2006 5:19 PM
                        Subject: [beam] Re: (possible) Robot Fodder

                        The use of
                        > carefully crafted parks seems irreconcilable with robots able to fend
                        > for themselves.

                        Wolf,

                        I would propose(as a Christian) that we live in a carefully crafted
                        enviroment, which is suited to our unique needs. Without being overly
                        philosophical, I think that as we create "life" the end goal should not
                        be that it survive in our enviroment. I think that bots suited to
                        certain enviroments are great, I don't want it to appear otherwise.
                        But I think that in trying to create a more humanistic(is that a word:)
                        type of bot we may overlook some intermediate "species" or something
                        along those lines :)

                        I was reading an article on beam a while ago and it was about battery
                        opperated bots not being true beam because the bot depends on a human
                        to change the battery. However, isnt it plausible to believe that a
                        Bot could die? Even Humans have a finite life span. Some insects only
                        live for a week or even a day. What are your thoughts on that?

                        chris

                      • J Wolfgang Goerlich
                        Hello Chris, Interesting perspective and I look forward to hearing more from you. Please allow me, if you will, to expand a bit on lifespans, memory, and
                        Message 11 of 11 , Aug 16, 2006
                        • 0 Attachment
                          Hello Chris,

                          Interesting perspective and I look forward to hearing more from you.
                          Please allow me, if you will, to expand a bit on lifespans, memory,
                          and parks.

                          Given that an animal can only have so many offspring in its lifetime,
                          its rate of biological adaptation is limited by lifespan multiplied
                          by number offspring. Further, to avoid extinction, a species' speed
                          of biological adaptation must be equal to or less than the rate of
                          environmental change. Thus many simpler life-forms (e.g. insects)
                          live for only a few days or even a few hours.

                          Humans skirt this issue in much the same way ants to, on a basic
                          level. I mentioned stigmergy earlier, which I find fascinating and
                          could basically summarize as an insect (or agent) using the
                          environment as the medium of communication and organizational memory.
                          Ants do this by leaving trails and building tunnels. In the same
                          fashion but on a different scale, humans do this with writing,
                          technology, and customs.

                          We can live longer because technological and social adaptation
                          replaces biological adaptation. If the weather suddenly gets colder,
                          we buy and don coats rather than adapt a thicker body hair. As a
                          Christian, you might say that this was how god intended it.

                          I like to get robots out of the park. Take one difference between a
                          person in his house and a robot in his park. In the first, the house
                          and its surrounds have been developed over centuries as part the life-
                          form's (e.g. our) technological and social adaptation. In the second,
                          the Robot Jurassic Park has been designed by an outside agent and its
                          occupant lacks direct impact. In the end, it is too easy to design a
                          park that compensates for flaws in our robots.

                          That, anyways, is my two cents.

                          J Wolfgang Goerlich



                          Chris wrote:
                          >
                          > Wolf,
                          >
                          > I would propose(as a Christian) that we live in a carefully crafted
                          > enviroment, which is suited to our unique needs. Without being
                          > overly philosophical, I think that as we create "life" the end goal
                          > should not be that it survive in our enviroment. I think that bots
                          > suited to certain enviroments are great, I don't want it to appear
                          > otherwise.
                          > But I think that in trying to create a more humanistic(is that a
                          > word:) type of bot we may overlook some intermediate "species" or
                          > something along those lines :)
                          >
                          > I was reading an article on beam a while ago and it was about
                          > battery opperated bots not being true beam because the bot depends
                          > on a human to change the battery. However, isnt it plausible to
                          > believe that a Bot could die? Even Humans have a finite life
                          > span. Some insects only live for a week or even a day. What are
                          > your thoughts on that?
                          >
                          > chris
                          >
                        Your message has been successfully submitted and would be delivered to recipients shortly.