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

robotic collective mind cloud

Expand Messages
  • shc101245
    I posted this on another group and got no response. So here is an idea I have been thinking about. What about building a robotic collective mind? I mean
    Message 1 of 8 , Aug 13, 2012
    • 0 Attachment
      I posted this on another group and got no response. So here is an idea I have been thinking about.
      What about building a robotic collective mind? I mean something that would run on
      several computers, either local or distributed on the web. Sort of a cloud
      concept.

      Since the big problem with most robots is not the mechanical it is the
      intelligence. Why not link any type of machine to some extensive computer power
      and see what we could do?

      Since linking any robot to a wifi net is relatively simple why not use their
      sensors to help a collective mind learn and develop the same way people learn.
      I don't know what level learning systems are presently at but we all have some
      ideas about this.
      I haven't thought this through a lot but the concept is very interesting and I
      think that this group might be the place for it. (if anyone is still here)

      What do you think?
    • JustinR
      I d say a lot of people have thought of this. I myself thought of the same thing around 2004 after seeing the movie I,Robot . I could see this happening at
      Message 2 of 8 , Aug 15, 2012
      • 0 Attachment
        I'd say a lot of people have thought of this. I myself thought of the same thing around 2004 after seeing the movie "I,Robot".

        I could see this happening at some point, many scientists and researchers point this as a possibility in the future. For example if in the future all robots run on some standard OS (lets say Android) then knowledge, appliction add ons, and special skill sets could be downloaded to the robot via an "app store" of sorts. Likewise if the robot learns how to do something new it could upload this skill set for other robots to down load.

        The robots could even apply a type of collective voting the the downloads like people do for apps. For example say a robot downloaded a solution to wash a car they could vote on how effect it was.

        As I see it, the problem comes into play with the hardware, OS and code needs to all match up. The market place (or a collection of individuals) would need the same hardware/OS/Code structure in order for something like this to work.

        I'm game for giving this a go if there are some other folks in the club that would want to try something like this on a small scale?

        Happy Roboting,
        -Justin R.


        --- In theroboticsclub@yahoogroups.com, "shc101245" <shc@...> wrote:
        >
        > I posted this on another group and got no response. So here is an idea I have been thinking about.
        > What about building a robotic collective mind? I mean something that would run on
        > several computers, either local or distributed on the web. Sort of a cloud
        > concept.
        >
        > Since the big problem with most robots is not the mechanical it is the
        > intelligence. Why not link any type of machine to some extensive computer power
        > and see what we could do?
        >
        > Since linking any robot to a wifi net is relatively simple why not use their
        > sensors to help a collective mind learn and develop the same way people learn.
        > I don't know what level learning systems are presently at but we all have some
        > ideas about this.
        > I haven't thought this through a lot but the concept is very interesting and I
        > think that this group might be the place for it. (if anyone is still here)
        >
        > What do you think?
        >
      • David
        It is an idea that has been around for a long time. If you don t know how to do Ai then using lots of computers isn t going to help, it s been tried. I take it
        Message 3 of 8 , Aug 16, 2012
        • 0 Attachment
          It is an idea that has been around for a long time. If you don't know how to do Ai then using lots of computers isn't going to help, it's been tried.
          I take it that you haven't built many robots since you think mechanics isn't a problem; from what I have seen of peoples robots the limiting factor is the mechanics. Just because it is easy to buy computing modules, sensor modules, net connection modules et.c. and plug them together then load up some libraries others have written many people seem to think like you that the same applies to mechanics, it doesn't as yet unless you are still using Lego, or have bought a modified vacuum cleaner as complete package.
          So back to a collective mind, if your robot can't process and amalgamate the data from the few sensors it has - sensor fusion - how is asking the hive mind going to help?
          What is lacking is a base level of intelligence from which to reach out for help with the hard problems. Presumably you don't ask Google how to get from your bedroom to the bathroom or how to put on your shoes or even how to stand up, you have some core capabilities which allow you to cope after a fashion with living in the world. Almost no robots to date have core capabilities to build upon. Sure there are neat demonstrations of some function but left to themselves most (all) robots will die because they don't have recharging facilities and their batteries run flat or they get stuck behind a door until their batteries run flat.
          Very few robots have even the core competencies of 'vacuum cleaner' robots.
          When my robots can wander about my house, garden and workshop and know where they are and know how to get to anywhere else in their world including the recharging station then they will be in a state to add facilities to ask about hard problems such as what is the name of this flower which presumes they have already asked about what is the fractaly thing anyway.
          What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
          So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.

          DAvid
          [1] HHG

          --- In theroboticsclub@yahoogroups.com, "shc101245" <shc@...> wrote:
          >
          > I posted this on another group and got no response. So here is an idea I have been thinking about.
          > What about building a robotic collective mind? I mean something that would run on
          > several computers, either local or distributed on the web. Sort of a cloud
          > concept.
          >
          > Since the big problem with most robots is not the mechanical it is the
          > intelligence. Why not link any type of machine to some extensive computer power
          > and see what we could do?
          >
          > Since linking any robot to a wifi net is relatively simple why not use their
          > sensors to help a collective mind learn and develop the same way people learn.
          > I don't know what level learning systems are presently at but we all have some
          > ideas about this.
          > I haven't thought this through a lot but the concept is very interesting and I
          > think that this group might be the place for it. (if anyone is still here)
          >
          > What do you think?
          >
        • JustinR
          ... So are you up for that - create a robot which can get about like an ant or a spider, the invention of Income Tax [1] can come later. DAvid [1] HHG
          Message 4 of 8 , Aug 16, 2012
          • 0 Attachment
            >>What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.

            So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.

            DAvid
            [1] HHG>>

            VB.Net! LOL (had to add my language preference in there) As someone one from the back of the room yells "Fortran!" (I know there is someone out there reading this yelling Fortran in their head)

            All joking aside, I think DAvid raised some good points. I would love to see a shared project raise up as it where from the club like the TRaCY robot kit project did back in the day.

            For a collaboration to happen I think it would require a consensus on basic hardware, controller, OS and coding language/coding standards. Would you agree?

            -Justin R.






            --- In theroboticsclub@yahoogroups.com, "David" <david@...> wrote:
            >
            > It is an idea that has been around for a long time. If you don't know how to do Ai then using lots of computers isn't going to help, it's been tried.
            > I take it that you haven't built many robots since you think mechanics isn't a problem; from what I have seen of peoples robots the limiting factor is the mechanics. Just because it is easy to buy computing modules, sensor modules, net connection modules et.c. and plug them together then load up some libraries others have written many people seem to think like you that the same applies to mechanics, it doesn't as yet unless you are still using Lego, or have bought a modified vacuum cleaner as complete package.
            > So back to a collective mind, if your robot can't process and amalgamate the data from the few sensors it has - sensor fusion - how is asking the hive mind going to help?
            > What is lacking is a base level of intelligence from which to reach out for help with the hard problems. Presumably you don't ask Google how to get from your bedroom to the bathroom or how to put on your shoes or even how to stand up, you have some core capabilities which allow you to cope after a fashion with living in the world. Almost no robots to date have core capabilities to build upon. Sure there are neat demonstrations of some function but left to themselves most (all) robots will die because they don't have recharging facilities and their batteries run flat or they get stuck behind a door until their batteries run flat.
            > Very few robots have even the core competencies of 'vacuum cleaner' robots.
            > When my robots can wander about my house, garden and workshop and know where they are and know how to get to anywhere else in their world including the recharging station then they will be in a state to add facilities to ask about hard problems such as what is the name of this flower which presumes they have already asked about what is the fractaly thing anyway.
            > What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
            > So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.
            >
            > DAvid
            > [1] HHG
            >
            > --- In theroboticsclub@yahoogroups.com, "shc101245" <shc@> wrote:
            > >
            > > I posted this on another group and got no response. So here is an idea I have been thinking about.
            > > What about building a robotic collective mind? I mean something that would run on
            > > several computers, either local or distributed on the web. Sort of a cloud
            > > concept.
            > >
            > > Since the big problem with most robots is not the mechanical it is the
            > > intelligence. Why not link any type of machine to some extensive computer power
            > > and see what we could do?
            > >
            > > Since linking any robot to a wifi net is relatively simple why not use their
            > > sensors to help a collective mind learn and develop the same way people learn.
            > > I don't know what level learning systems are presently at but we all have some
            > > ideas about this.
            > > I haven't thought this through a lot but the concept is very interesting and I
            > > think that this group might be the place for it. (if anyone is still here)
            > >
            > > What do you think?
            > >
            >
          • David
            Actually that would be going about it backwards. A little true story. - A good few years ago the British Government had a Robotics Initiative and the Shadow
            Message 5 of 8 , Aug 17, 2012
            • 0 Attachment
              Actually that would be going about it backwards.
              A little true story. - A good few years ago the British Government had a Robotics Initiative and the Shadow Group was a member of the Domestic and Leisure section. After initial feasibility studies the Initiative moved on to implementation. At its first Implementation meeting a bunch of guys decided that the way forward was to decide on connector standards so the different members could build equipment which would plug together. I knew at that point the whole initiative was doomed and after about another two meetings everything fell apart.

              Cargo Cult Robotics, if it plugs together it must work mustn't it.

              So I don't agree "it would require a consensus on basic hardware, controller, OS and coding language/coding standards."

              What we need is for people to build their own robots, or at least make enhancements to existing ones. Note 'enhancements' not just porting another OS or language to them.
              We need people to solve the core problems.
              We need people to discover where to place sensors and what sensors to place.
              We need people to discover what geometries work best.
              What are good designs for getting over door thresholds, and going up and down steps and stairs.
              What are workable architectures for robot minds.
              What are good methods of telling or showing a robot what to do.
              Non of this will be solved by committee, by talking about it or writing papers.
              The core problems will only be solved by doing it and then we can share how we did it.
              "Hey that is a good way, I'll add it to my robot"
              Trouble is at the moment all we have of robots is photos of what they look like from the outside. That may satisfy artists and stylists but they give little or no information about how the robots work. And photos give no information at all about how the software works. Libraries and Listings are not documentation of methodologies.

              On another group there has been a little discussion of the Heathkit Hero-1, probably the most famous hobby robot in the world. So how did it work? Who knows? All we have is some photos, a few of which show the mechanics. But where are the specs, where are the overviews of the electronics and software. Could you if you wanted to reproduce a Hero-1, the answer is no. You could maybe make something which looked like a Hero-1 - sculptures of people are often quite lifelike but they sure don't work like people.

              If you look at what goes on in similar technological hobbies, hobby and educational institute robotics fares poorly. Look at the advances in quadcopters over the last few years. People built them, shared experiences, and built better ones. I would guess that hobby built quadcopters far outnumber hobby built robots.

              We need hobby roboticists to go and build something better than what exists and then share.
              But they need to build and build and build not just discuss concepts.

              DAvid


              --- In theroboticsclub@yahoogroups.com, "JustinR" <weyoun7@...> wrote:
              >
              >
              > >>What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
              >
              > So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.
              >
              > DAvid
              > [1] HHG>>
              >
              > VB.Net! LOL (had to add my language preference in there) As someone one from the back of the room yells "Fortran!" (I know there is someone out there reading this yelling Fortran in their head)
              >
              > All joking aside, I think DAvid raised some good points. I would love to see a shared project raise up as it where from the club like the TRaCY robot kit project did back in the day.
              >
              > For a collaboration to happen I think it would require a consensus on basic hardware, controller, OS and coding language/coding standards. Would you agree?
              >
              > -Justin R.
              >
              >
              >
              >
              >
              >
              > --- In theroboticsclub@yahoogroups.com, "David" <david@> wrote:
              > >
              > > It is an idea that has been around for a long time. If you don't know how to do Ai then using lots of computers isn't going to help, it's been tried.
              > > I take it that you haven't built many robots since you think mechanics isn't a problem; from what I have seen of peoples robots the limiting factor is the mechanics. Just because it is easy to buy computing modules, sensor modules, net connection modules et.c. and plug them together then load up some libraries others have written many people seem to think like you that the same applies to mechanics, it doesn't as yet unless you are still using Lego, or have bought a modified vacuum cleaner as complete package.
              > > So back to a collective mind, if your robot can't process and amalgamate the data from the few sensors it has - sensor fusion - how is asking the hive mind going to help?
              > > What is lacking is a base level of intelligence from which to reach out for help with the hard problems. Presumably you don't ask Google how to get from your bedroom to the bathroom or how to put on your shoes or even how to stand up, you have some core capabilities which allow you to cope after a fashion with living in the world. Almost no robots to date have core capabilities to build upon. Sure there are neat demonstrations of some function but left to themselves most (all) robots will die because they don't have recharging facilities and their batteries run flat or they get stuck behind a door until their batteries run flat.
              > > Very few robots have even the core competencies of 'vacuum cleaner' robots.
              > > When my robots can wander about my house, garden and workshop and know where they are and know how to get to anywhere else in their world including the recharging station then they will be in a state to add facilities to ask about hard problems such as what is the name of this flower which presumes they have already asked about what is the fractaly thing anyway.
              > > What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
              > > So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.
              > >
              > > DAvid
              > > [1] HHG
              > >
              > > --- In theroboticsclub@yahoogroups.com, "shc101245" <shc@> wrote:
              > > >
              > > > I posted this on another group and got no response. So here is an idea I have been thinking about.
              > > > What about building a robotic collective mind? I mean something that would run on
              > > > several computers, either local or distributed on the web. Sort of a cloud
              > > > concept.
              > > >
              > > > Since the big problem with most robots is not the mechanical it is the
              > > > intelligence. Why not link any type of machine to some extensive computer power
              > > > and see what we could do?
              > > >
              > > > Since linking any robot to a wifi net is relatively simple why not use their
              > > > sensors to help a collective mind learn and develop the same way people learn.
              > > > I don't know what level learning systems are presently at but we all have some
              > > > ideas about this.
              > > > I haven't thought this through a lot but the concept is very interesting and I
              > > > think that this group might be the place for it. (if anyone is still here)
              > > >
              > > > What do you think?
              > > >
              > >
              >
            • JustinR
              I see it all little differently. For example, I tend to use Parallax brand controllers (Basic Stamps programmed in PBASIC) and accessors for servo controllers
              Message 6 of 8 , Aug 17, 2012
              • 0 Attachment
                I see it all little differently.

                For example, I tend to use Parallax brand controllers (Basic Stamps programmed in PBASIC) and accessors for servo controllers and PING sonar sensor as well as their cont. rotation servos on rolling mobile robots. I also tend to use Visual Studio .Net (vb.net) on a windows OS. I also like RoboRealm for vision processing and OpenCV I kinda/sorta understand.

                With that in mind I've seen robot projects where someone made a fantastic moble robot with dedicated PIC chips. With a detailed guide to what made their robot successfully I can't take the PIC code and upload it directly to my Basic Stamp. I'd have to dive into the code, take it apart and figure out what functionality I can remake in PBASIC code.

                Or if someone makes a wonderful speaking robot with facial recognition on a robot with a linux OS and ROS, those are tools I'm not used to using and perhaps don't want to use. Even with a detailed guide on how their robot works, it may not benefit me.

                It would leave me relying on folks who use Basic Stamps, VB.Net and RoboRealm for ideas and tips on how to improve my robot.

                I look at it like the history of old computers and computer clubs. Back in the day when the Commadore64 was around, computer clubs would spring up where folks who had the same interestest and same hardware/software could share ideas and resources. Someone finds a better monitor for the C64, before you know everyone has one like it. Someone writes neat code to managing files, everyone copies it....someone improves it....everyone copies that version and so on. That type of collaboration would not work if someone had a C64 someone else had Apple-IIe and someone else had IBM-XT.

                On a side note, I have a HeatherKit Hero-1 robot with the manuals and schematics. I would guess the robot could be remade with much newer parts. (Not sure if you had a true interest in it or just used it as an example?)

                -Justin R.



                --- In theroboticsclub@yahoogroups.com, "David" <david@...> wrote:
                >
                >
                > Actually that would be going about it backwards.
                > A little true story. - A good few years ago the British Government had a Robotics Initiative and the Shadow Group was a member of the Domestic and Leisure section. After initial feasibility studies the Initiative moved on to implementation. At its first Implementation meeting a bunch of guys decided that the way forward was to decide on connector standards so the different members could build equipment which would plug together. I knew at that point the whole initiative was doomed and after about another two meetings everything fell apart.
                >
                > Cargo Cult Robotics, if it plugs together it must work mustn't it.
                >
                > So I don't agree "it would require a consensus on basic hardware, controller, OS and coding language/coding standards."
                >
                > What we need is for people to build their own robots, or at least make enhancements to existing ones. Note 'enhancements' not just porting another OS or language to them.
                > We need people to solve the core problems.
                > We need people to discover where to place sensors and what sensors to place.
                > We need people to discover what geometries work best.
                > What are good designs for getting over door thresholds, and going up and down steps and stairs.
                > What are workable architectures for robot minds.
                > What are good methods of telling or showing a robot what to do.
                > Non of this will be solved by committee, by talking about it or writing papers.
                > The core problems will only be solved by doing it and then we can share how we did it.
                > "Hey that is a good way, I'll add it to my robot"
                > Trouble is at the moment all we have of robots is photos of what they look like from the outside. That may satisfy artists and stylists but they give little or no information about how the robots work. And photos give no information at all about how the software works. Libraries and Listings are not documentation of methodologies.
                >
                > On another group there has been a little discussion of the Heathkit Hero-1, probably the most famous hobby robot in the world. So how did it work? Who knows? All we have is some photos, a few of which show the mechanics. But where are the specs, where are the overviews of the electronics and software. Could you if you wanted to reproduce a Hero-1, the answer is no. You could maybe make something which looked like a Hero-1 - sculptures of people are often quite lifelike but they sure don't work like people.
                >
                > If you look at what goes on in similar technological hobbies, hobby and educational institute robotics fares poorly. Look at the advances in quadcopters over the last few years. People built them, shared experiences, and built better ones. I would guess that hobby built quadcopters far outnumber hobby built robots.
                >
                > We need hobby roboticists to go and build something better than what exists and then share.
                > But they need to build and build and build not just discuss concepts.
                >
                > DAvid
                >
                >
                > --- In theroboticsclub@yahoogroups.com, "JustinR" <weyoun7@> wrote:
                > >
                > >
                > > >>What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
                > >
                > > So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.
                > >
                > > DAvid
                > > [1] HHG>>
                > >
                > > VB.Net! LOL (had to add my language preference in there) As someone one from the back of the room yells "Fortran!" (I know there is someone out there reading this yelling Fortran in their head)
                > >
                > > All joking aside, I think DAvid raised some good points. I would love to see a shared project raise up as it where from the club like the TRaCY robot kit project did back in the day.
                > >
                > > For a collaboration to happen I think it would require a consensus on basic hardware, controller, OS and coding language/coding standards. Would you agree?
                > >
                > > -Justin R.
                > >
                > >
                > >
                > >
                > >
                > >
                > > --- In theroboticsclub@yahoogroups.com, "David" <david@> wrote:
                > > >
                > > > It is an idea that has been around for a long time. If you don't know how to do Ai then using lots of computers isn't going to help, it's been tried.
                > > > I take it that you haven't built many robots since you think mechanics isn't a problem; from what I have seen of peoples robots the limiting factor is the mechanics. Just because it is easy to buy computing modules, sensor modules, net connection modules et.c. and plug them together then load up some libraries others have written many people seem to think like you that the same applies to mechanics, it doesn't as yet unless you are still using Lego, or have bought a modified vacuum cleaner as complete package.
                > > > So back to a collective mind, if your robot can't process and amalgamate the data from the few sensors it has - sensor fusion - how is asking the hive mind going to help?
                > > > What is lacking is a base level of intelligence from which to reach out for help with the hard problems. Presumably you don't ask Google how to get from your bedroom to the bathroom or how to put on your shoes or even how to stand up, you have some core capabilities which allow you to cope after a fashion with living in the world. Almost no robots to date have core capabilities to build upon. Sure there are neat demonstrations of some function but left to themselves most (all) robots will die because they don't have recharging facilities and their batteries run flat or they get stuck behind a door until their batteries run flat.
                > > > Very few robots have even the core competencies of 'vacuum cleaner' robots.
                > > > When my robots can wander about my house, garden and workshop and know where they are and know how to get to anywhere else in their world including the recharging station then they will be in a state to add facilities to ask about hard problems such as what is the name of this flower which presumes they have already asked about what is the fractaly thing anyway.
                > > > What this group could do usefully is to exchange ideas and information on how to create robots with suitable core competencies. There are lots of other groups - see LinkedIn - which waste endless words on whether C! or C-- is better for robots or whether robots should be banned until we understand how to do Ai.
                > > > So are you up for that - create a robot which can get about like an ant or a spider, the 'invention of Income Tax'[1] can come later.
                > > >
                > > > DAvid
                > > > [1] HHG
                > > >
                > > > --- In theroboticsclub@yahoogroups.com, "shc101245" <shc@> wrote:
                > > > >
                > > > > I posted this on another group and got no response. So here is an idea I have been thinking about.
                > > > > What about building a robotic collective mind? I mean something that would run on
                > > > > several computers, either local or distributed on the web. Sort of a cloud
                > > > > concept.
                > > > >
                > > > > Since the big problem with most robots is not the mechanical it is the
                > > > > intelligence. Why not link any type of machine to some extensive computer power
                > > > > and see what we could do?
                > > > >
                > > > > Since linking any robot to a wifi net is relatively simple why not use their
                > > > > sensors to help a collective mind learn and develop the same way people learn.
                > > > > I don't know what level learning systems are presently at but we all have some
                > > > > ideas about this.
                > > > > I haven't thought this through a lot but the concept is very interesting and I
                > > > > think that this group might be the place for it. (if anyone is still here)
                > > > >
                > > > > What do you think?
                > > > >
                > > >
                > >
                >
              • kensrobots
                ... From: Ken Boone To: theroboticsclub@yahoogroups.com Cc: Date: Sun, 19 Aug 2012 17:26:45 -0400 Subject: Re: robotic collective mind
                Message 7 of 8 , Aug 19, 2012
                • 0 Attachment
                  ---------- Forwarded message ----------
                  From: Ken Boone <kensrobots@...>
                  To: theroboticsclub@yahoogroups.com
                  Cc:
                  Date: Sun, 19 Aug 2012 17:26:45 -0400
                  Subject: Re: robotic collective mind cloud


                  All of the schematics, manuals and robot code is stillavailable for all three of the Heath Kit HERO robots. Also there are improvements (hardware & Software) being made on the HERO 2000 by thegroup.



                  However the HERO 2000's processors (yes it has more than one) and peripheralcards are such old technology you would have to either replace them all or asit has been discussed in the group just use the robot's serial port to control therobot and monitor the robots sensors with a higher level external probram.



                  Ken Boone

                  KensRobots dot com
                • shc101245
                  OK, if you think about it there are only a few OS s out there and it might be possible to have a software driver program that interfaced with each robot s
                  Message 8 of 8 , Jan 22, 2014
                  • 0 Attachment

                    OK,  if you think about it there are only a few OS's out there and it might be possible to have a software  "driver" program that interfaced with each robot's OS. and made it possible to communicate with the cloud.

                     

                    The hardware would have to have a minimal set of capabilities but the OS would interface various hardware to " driver" that in turn talks to the collective brain cloud. 


                    For example: if the robot had a vision capability it could interface with the cloud vision applications.  If it did not then it could not use the cloud apps.

                    Isn't this similar to what is presently done with cell phones, tablets and computers now?




                  Your message has been successfully submitted and would be delivered to recipients shortly.