Related, yes. But not really the same - a different philosophy. (I'm on the project - though not speaking officially, etc, etc).
Swarm robots are many small independant robots which can work in groups (usually using some kind of behavoural control) to carry out a job. Also I think every swarm robot I recall seeing has been wheeled (though this is a matter of implementation, not a theoretical constraint).
A modular self re-configurable robot is one robot that is assembled out of many small pieces that lock together in some fashion. The control may be distributed, but is likely to be coordinated
and deterministic (rather than emergent behaviour).
Compare a swarm of ants vs the liquid-metal man in _Terminator II_.
Both have common advatages though: robustness (if one fails, then there are many more to replace it); versatility (it's not a fixed monolithic machine); and cheapness (completely composed of many small identical parts, hopefully cheap due to ecconomies of scale).
Yes, sorry about the out-of-date pages (I'm on the project - though speaking unofficially etc). No one has ever been specifically assigned the job of maintaining them - and it's not a job anyone really wants !
The work has in fact been covered in "normal media" for some time (ABC world news last year, and plenty before that too). But yes, they can be very slow in picking things up at times (though slow is better than when they're inaccurate !).
The above stated schedule for G3 has slipped a bit. Partly the nature of research, and partly due to the other parts of the project running concurrently. The last of the mechanical pieces we've had fabricated are arriving pretty much as we speak. So work is continuing (and no, I'm not silly enough to commit to a new date:-).
>What would be real news is an implementation
We have not just one, but several implementations (G3 is yet another). So it certainly is not "just theory". But that being said, you aren't going to find a modular self-reconfiguring robot in the "domestic cleaning" aisle of Safeways for some time yet !
This is in fact one of the medium-term goals stated in our research contract (I'm on of the people on this project - though speaking unofficially etc). But nothing practical "right now" I'm sorry to say - we all wish we did.
Thanks for the kind words ! As you say, we have many challanges (I'm one of the researchers on the project - though speaking unofficially etc, etc).
Purpose built robots (99% of robots out there) are cheaper and do the job they are designed for better...but you need a different robot for each specialised task. If you can buy just one robot and have it do *everything*, then that's a different story.
Not quite true I'm afraid. I'm one of the researchers on the project (though speaking unofficially etc etc).
We have several different prototypes being developed concurrently - including the 15A jobby (which as you rightly point out, drains most batteries fairly fast:-).
But some of them are fully self-contained (completely tetherless) and can travel about 1 Mile on the flat or up several flights of stairs
on one charge.
Slashdot Mirror
Related, yes. But not really the same - a different philosophy. (I'm on the project - though not speaking officially, etc, etc).
Swarm robots are many small independant robots which can work in groups (usually using some kind of behavoural control) to carry out a job. Also I think every swarm robot I recall seeing has been wheeled (though this is a matter of implementation, not a theoretical constraint).
A modular self re-configurable robot is one robot that is assembled out of many small pieces that lock together in some fashion. The control may be distributed, but is likely to be coordinated
and deterministic (rather than emergent behaviour).
Compare a swarm of ants vs the liquid-metal man in _Terminator II_.
Both have common advatages though: robustness (if one fails, then there are many more to replace it); versatility (it's not a fixed monolithic machine); and cheapness (completely composed of many small identical parts, hopefully cheap due to ecconomies of scale).
Yes, sorry about the out-of-date pages (I'm on the project - though speaking unofficially etc). No one has ever been specifically assigned the job of maintaining them - and it's not a job anyone really wants !
:-).
The work has in fact been covered in "normal media" for some time (ABC world news last year, and plenty before that too). But yes, they can be very slow in picking things up at times (though slow is better than when they're inaccurate !).
The above stated schedule for G3 has slipped a bit. Partly the nature of research, and partly due to the other parts of the project running concurrently. The last of the mechanical pieces we've had fabricated are arriving pretty much as we speak. So work is continuing (and no, I'm not silly enough to commit to a new date
>What would be real news is an implementation
We have not just one, but several implementations (G3 is yet another). So it certainly is not "just theory". But that being said, you aren't going to find a modular self-reconfiguring robot in the "domestic cleaning" aisle of Safeways for some time yet !
This is in fact one of the medium-term goals stated in our research contract (I'm on of the people on this project - though speaking unofficially etc). But nothing practical "right now" I'm sorry to say - we all wish we did.
Thanks for the kind words ! As you say, we have many challanges (I'm one of the researchers on the project - though speaking unofficially etc, etc).
Purpose built robots (99% of robots out there) are cheaper and do the job they are designed for better...but you need a different robot for each specialised task. If you can buy just one robot and have it do *everything*, then that's a different story.
Not quite true I'm afraid. I'm one of the researchers on the project (though speaking unofficially etc etc).
:-).
We have several different prototypes being developed concurrently - including the 15A jobby (which as you rightly point out, drains most batteries fairly fast
But some of them are fully self-contained (completely tetherless) and can travel about 1 Mile on the flat or up several flights of stairs
on one charge.