Engineers are working on – or, at any rate, some hold out the hypothetical possibility of – nanobots, but what, in theory, would power them? Would they carry miniaturized electric storage batteries, or what?
A little googling turned this up.
Cached version so what you are looking for is highlighted
Further searches for Nanogenerators turned this up directly from Georgia Tech:
http://www.gatech.edu/news-room/release.php?id=932
I guess this is at least one option for what they are looking into.
Yeah, basically they use nano-sized piezo-electric materials, the same thing that makes microphones work. A mechanical action causes an electrical current. Just scale it down really, really, really small. A shit-ton less current, to be sure, but then, these nanobots would be so small that they wouldn’t need a lot.
Depends on what you call nanobots. If we’re talking about the ‘ideal’ nanobot, i.e., a mechanism on the scale of one nanometer, it’s going to be hard. Keep in mind that something one nanometer long is only the length of approximately 10 hydrogen atoms. Hell, many biological enzymes are bigger than that! At that scale, on-board power sources become less a discrete system and more a property of the whole nanobot, and very hard to separate from its task.
Currently, fully-synthetic molecular motors are usually powered by photonic flux; you shine a laser on it, and it spins (or pivots, or walks, or whatever). Biological molecular motors are usually powered by high-energy intermediate molecules dissolved in solution (the triphosphates: ATP, GTP, etc). You could perhaps make a nanobot ‘ecology’ where ‘producers’ (large nanobots that contain light-gathering antennas) synthesize high-energy ‘fuel’ for ‘workers’ (small nanobots that use the high-energy intermediates to carry out whatever tasks you designed them for).
Internally, a nanobot could potentially store power in the form of some sort of spring (torsion of molecular bonds?), or a mechanochemical battery (high-energy steric arrangement vs low-energy). It could perhaps store power in the form of a tiny capacitor constructed from an internal cavity. More speculatively, it could perhaps derive energy from the decay of a radioactive isotope, though it would be a very short-lived power source unless they can scavenge more from someplace. Remember, we’re talking about at most dozens of atoms.
Seems to me, then, that the grey goo scenario could be averted very easily by limiting (intentionally, or simply for lack of other options) the period of time nanobots can operate before running out of power.
A strain of von Neumann replicator that mutated to have a slightly longer battery life would have a reproductive advantage over intentionally power-limited strains. IOW, evolution can trump intelligent design.
I don’t think “nanobot” means that it is literally one nanometer in size; I believe the term nanotech means that nanometers are the most convenient unit to express the size of the devices involved (as opposed to microns, for instance). So I’d call anything smaller than 500nm nanotechnology.
But can it trump physics?
Only if it works at Time Cube or the Palo Alto Caret Laboratory.