Diamond-based nanomachines and micromachines would be vulnerable to combustion. Ones made of sapphire less so, but everything has thermal limits. I would also look at chemicals that gum up particular sensors or surface channels. A whiff of iodine or a spray of certain low viscosity glue might be really troublesome for particular models. The hard part is knowing what works against which design. Nanoswarms are great when you want to have a nanomachine close to every point in a volume, like scouting, spying or fighting other nanoswarms, but they are not very efficient for things like building or destroying stuff - that is where you want your machines to form surfaces or infrastructures that scale effectively.

NewtonPulsifer wrote:

The whole fast moving micro machine idea is pretty dubious to me. You have to have a certain minimum mass to surface area ratio to fly - otherwise the little guy just gets carried away by air currents. Better to crawl. That 16cm per hour land speed isn't exactly hair raisingly suspenseful now is it, though?

A smart swarm of course clumps together into bigger streamlined conglomerates: a mist might turn into a cloud of mosquitoes in one turn, and then into a few swallows the next - upping the speed at the expense of the spread.

Nanomachines and micromachines are vulnerable to EMP. A directed EMP burst aimed at the swarm will help deal with them.

I say that The most effective weapon against nanites, are nanites & Hygiene practices. A comparison to virus and bacteria is fairly close. And would likely need a huge variety of types to achive a perfect response. . That said a cloud of nanites, would be hard to deal with using conventional weaponry, Imagine using a vacum cleaner against a cloud of sawdust. To do so without getting any dust on would be quite a feat. I also think that Igniting a "hazardus" sawdust cloud isnt a good idea either especially it its to preventing sawdust exposure. It would cause a dust explotion, perhaps burn some or most of it, but lots of unurnt dust would be still dispersed -and thus risk pollutin of a even greater area. Also IRL it seems properties of magnetism can disapear on the nanoscale, making a scenario that nanites are imune to emp plausible. http://nanotechweb.org/cws/article/tech/38953 Here is another text I found on this topic.

Quote:

Ridding oneself of a nanobot “infection” could be difficult. If one is carrying helpful nanobots in one’s bloodstream and tissue, and one is “tased”, scanned with a magnetic resonance imaging machine (MRI), or otherwise suffers a strong electrical shock, does one then have to obtain and ingest or inject a whole new batch of the little helpers? Similarly, in a case of nano-warfare where people are afflicted with tiny destructive nanobots, could the little devils be deactivated or purged by a powerful magnetic pulse or electric shock? Electrical shock might not be effective against nanobots. When living organisms are shocked with electricity, the current of electrons – the part that does the damage – follows the paths of lowest resistance. In most cases these paths are the nervous and circulatory systems, continuously-connected networks of relatively more conductive tissue or fluid. Unless a nanobot was directly in the path such that the current flowed through it, it might not be damaged at all. Also, for current to flow there must be a potential (voltage) difference across it. Are nanobots so small that there would rarely ever be enough potential across their longest dimension as to cause enough current flow to damage them? The effectiveness of a magnetic method could be limited. A magnetic pulse or field of sufficient strength might have better potential to harm a nanobot, but that would require the nanobot contain a material susceptible to magnetic fields. Unfortunately, a nanobot comprised of carbon or other non-magnetic materials would be untouched by a magnetic field, just as are most life forms. Biological nanobots might require some different approaches. Of course, nanotechnology experts have been noting for years that the most well developed nanotechnology, living cells and microorganisms, is already in place and doing just fine. Chemical means such as drugs can damage specific microorganisms or encourage the body’s immune system to attack them, so a similar technique might be used to kill nanobots. What if nanobots could be confused or convinced to go away? I remember a radio news story on NPR a few years ago about fruit orchards in Northwest Ohio where lady bugs (predator insects) were used to keep unwanted pests under control. The fruit farmer had bought millions of lady bugs, huge bags of them, and turned them loose in the orchards only to later in the day watch as his entire investment, clouds of the tiny insects, rose into the air and headed for Indiana. Nanobots are expected to have a lot less movement capability, but that doesn’t mean that a suitable agent couldn’t be introduced or ingested that would make them try to abandon ship, exiting the body through every possible means. It might take days or weeks, but eventually the unwanted devices might leave without harm. Of course, it desirable nanobots were also chased off one would have to obtain more when conditions were such that they would stay to do their beneficial jobs. Timothy F. Prosser http://timprosserfuturing.wordpress.com/2008/08/28/how-to-kill-nanobots/

This paper describes a 40kHz ultrasound system with a 10-20 meter range. http://heim.ifi.uio.no/sverre/papers/05_IEEEUltrasonics_IPS.pdf But our micromachines are too small to hear 40kHz. Their "eardrum" is on the order of 10,000 times the length of a human's (assume 1cm for human). So they're going to be hearing not a human's 20kHz limit but rather more like a 200mHz ultrasound. Unfortunately 5gHz is a hard limit in air, and further a couple of mHz would only have a range of 1-2 centimeters. 200mHz would basically be contact only. http://physics.stackexchange.com/questions/23418/is-there-an-upper-frequ... On top of this the bits per second with ultrasound are quite low (see first link). Even if somehow you could figure out how to make 40kHz transducers for something 100 micrometers in size using applied phlebotinum, you'd have to time slice that between a billion nanomachines to use it. It would make the data rate unusably slow and very very susceptible to "jamming", even from background industrial noise (oh no, the A/C came on! We can't hear each other!). If I have time later I'll detail why any system of gliding nanites would be overcome with turbulence. That might not be until tomorrow. CONCLUSION: Ultrasound is not feasible. EDIT: Further, doppler effects at those ranges limit how fast two communicators can be moving relative to each other when communicating. It would be a max speed of like 6km/h for 40kHz (see the first paper link), and then worse as you change conditions (higher frequency ultrasound). So basically the swarm would have a speed limit while whistling to each other. EDIT2: Without having read Robert Freitas' book, I'm going to hazard a guess that his mesh network idea of ultra-high frequency ultrasound would be operating inside somebody's body. Ultrasound through a solid or liquid is a totally different beast than through the air.

How would one make a 2mHz transducer that small? You're looking at a wavelength of 0.000172 meters. That's .172 milimeters, or 172 microns. If the "eardrum" itself was 172 microns, you could see it (just the eardrum) unaided with the naked eye. http://www.mcsquared.com/wavelength.htm So 2mHz ultrasound is not feasible.

Mechanically, the optimal weapon is a drone with nanoscopic vision and a seeker rocket launcher, armed with plasma rockets. Preferably, multiple drones with nanoscopic vision and seeker rocket launchers, because you're going to need multiple rockets to take down the Titan nanoswarms.

nick012000 wrote:

Mechanically, the optimal weapon is a drone with nanoscopic vision and a seeker rocket launcher, armed with plasma rockets. Preferably, multiple drones with nanoscopic vision and seeker rocket launchers, because you're going to need multiple rockets to take down the Titan nanoswarms.

Nanoscopic vision, is absurdly myopic. You cannot see things of a size smaller than the wavelength of the light you use, so unless you use something in the far UV spectrum (where air is opaque anyway) you will not be able to see nanomachines a few hundred nanometer in size. The solution is near-field imaging that can get around the diffraction limit: however, now you have to be within a wavelength or so: http://www.nanomedicine.com/NMI/4.8.4.htm That seeker will simply fly until it detects that it draws in nanites in its air intake valves, and then detonate. It will not be able to do much more than a flight pattern that randomly changes direction to go uphill in nanite density and detonate at a local maximum. Not very smart, but not stupid either.

NewtonPulsifer wrote:

If you're too small to get turbulence, you're too small to fly.

Tell that to gnats. As explained on http://www.nanomedicine.com/NMI/9.5.3.1.htm nanobots will not be able to fly using inertial flight, but will be moving in the viscous regime: for all practical purposes they are swimming. Freitas calulates http://www.nanomedicine.com/NMI/9.5.3.4.htm that a 10 micron nanomachine can achieve 10 m/s if it is willing to really burn energy. Furthermore, he writes:

Freitas wrote:

Aerobot power density Dnano scales as ~vnano^2 and ~Rnano^-2 in the viscous regime. In the transitional regime, power density scales as ~vnano^3, and as ~Rnano^-2 at low velocities and ~Rnano^-1 at high velocities. Thus, to minimize power density and therefore conserve energy, circumcorporeal clouds of aerial nanorobots may coalesce into progressively larger but fewer tightly-packed clumps as the collective velocity of the cloud moves to higher airspeeds, assuming that the aeromotive mechanism design is largely scale-invariant over the full size range of the progressive aggregations. This strategy is most effective in the viscous regime. For example, consider a cloudlet consisting of 1 million nanorobots, each of size Rnano = 1 micron. With individual nanorobots traveling at vnano ~ 30 cm/sec, the cloudlet consumes ~0.4 milliwatts and operates at a power density of ~10^8 watts/m^3. Now assume that the cloudlet must speed up to 10 m/sec to track a fast-moving object around which it is stationkeeping, or to compensate for a heavy wind. If the individual nanorobots comprising the cloudlet simply increase their airspeed to 10 m/sec, then power density in each nanorobot increases to ~10^11 watts/m^3 and cloudlet power consumption rises to ~400 milliwatts (a 1000-fold increase). However, if the nanorobots temporarily aggregate into a single collective approximating a single device of Rnano = 100 microns, then the power consumption of the collective can be held to the original 0.4 milliwatts and power density remains constant at ~10^8 watts/m3. Facultative aggregation may permit stationkeeping over a wide range of velocities without significantly increasing power. (Other power-conserving behaviors, such as preferential migration into the downwind slipstream of a rapidly-moving tracked object, are not considered further here but may be useful.)

Can they accelerate well enough? He calculates that they can do tens of thousands of G's if they really burn fuel, and "At a more modest Pnano ~ 4500 pW, the nanoflyer accelerates at anano ~ 1100 g's for taccel ~ 96 microsec, reaching vnano = 1 m/sec after crossing a running distance of Xaccel ~ 48 microns with an efficiency of e% ~ 0.15 (15%) and vair ~ 6 m/sec. "

NewtonPulsifer wrote:

How would one make a 2MHz transducer that small?

Freitas looks at vibrating pistons and spheres changing size, http://www.nanomedicine.com/NMI/7.2.2.1.htm

Freitas wrote:

For example, a vibrating piston radiator of radius r = 1 micron and input power Pin = 10 pW operating at n = 1 MHz in vivo produces Ap = 0.0007 atm radiation pressure, Pout = 0.005 pW (giving e% = Pout / Pin = 0.0005 (0.05%) and an acoustic intensity of Ip = 0.002 watts/m^2 at the surface of the radiator, assuming r = 993.4 kg/m^3, h = 1.1 x 10^-3 kg/m-sec and vsound ~ 1500 m/sec for human interstitial fluid at 310 K. Figure 7.1 summarizes Eqn. 7.6 for various parameter choices.

Some changes in the calculation are needed for air rather than body fluids. The piston is pretty obvious how to build given classic Drexlerian diamond nanosystems (BTW, Eric would like to officially point out - as always - that he only used diamond because it was easy to analyse, and he thinks real nanosystems will use all sorts of different materials, including biological ones). I would assume a pulsating sphere could be done using a graphene-like sheet that is extended and withdrawn using gas on the inside and tension from a holding device. To sum up, everybody wanting to analyse nanodevices ought to read Freitas book. Even when you don't buy his solutions (I have some serious problems with his histonation ideas and the size of the comms networks needed to do brain scanning) he has done his homework and made sure the formulas and physics is in there.

The Doctor wrote:

Arenamontanus wrote:

A smart swarm of course clumps together into bigger streamlined conglomerates: a mist might turn into a cloud of mosquitoes in one turn, and then into a few swallows the next - upping the speed at the expense of the spread.

TITAN nanoswarms, or transhuman nanoswarms? At first scratch, I would guess the latter.

Both, except that transhuman nanoswarms doesn't seem to be very smart. I am sure one could build in the reflexes to make a nanoswarm that under some conditions gathers itself together into a flock (for example, when putting a reusable swarm back into a cannister), but it would have to be programmed from the start. Most swarms are little more than clever chemicals. TITAN swarms are smart enough to come up with it on their own, plus innovating new tricks: "Ah, the transhuman is opening the airlock. Will not reach it or the walls in time to avoid getting vented - better tell the hive about that trick when I get the chance. Hmm... let's swarm together into microspheres encapsulating air, and when blown out release the pressurized air as vectored thrust to get back to the hull. Enough of me ought to survive to produce a cohesive mind. Either the airlock is still open, or we can cannibalize a bit of hull to make a drill. Here we go!"

Note – for those not up on their greek and/or metric, a µm is a micron (1 millionth of a meter) vs a nanometer (nm) which is 1 billionth of a meter.

Freitas wrote:

...For example, consider a cloudlet consisting of 1 million nanorobots, each of size Rnano = 1 micron. With individual nanorobots traveling at vnano ~ 30 cm/sec, the cloudlet consumes ~0.4 milliwatts and operates at a power density of ~10^8 watts/m^3....

A liter (1000 cubic centimeters) of gasoline has about 40 megajoules (4x10^7). A cubic centimeter has 40,000 joules. A cubic milimeter, 40 joules. A cubic micron (here’s where we are at the aerobot’s size) is at 40 nanojoules. Frietas has a swarm of 1 million nano-aerobots consuming 0.4 milliwatts. That is 0.4 nanowatts per aerobot. 0.4 nanonwatts is the same as saying 0.4 nanojoules per second. The aerobots are likely more like a sphere than a cube, so they lose about half of their volume. Call it 30% of their volume for fuel, and they have about 12 nanojoules of fuel. [b][i]To fly at a velocity of just 1km/h they would consume all of their fuel in 30 seconds.[/b][/i]

NewtonPulsifer wrote:

How would one make a 2MHz transducer that small?

Freitas wrote:

For example, a vibrating piston radiator of radius r = 1 micron and input power Pin = 10 pW operating at n = 1 MHz in vivo produces Ap = 0.0007 atm radiation pressure, Pout = 0.005 pW (giving e% = Pout / Pin = 0.0005 (0.05%) and an acoustic intensity of Ip = 0.002 watts/m^2 at the surface of the radiator, assuming r = 993.4 kg/m^3, h = 1.1 x 10^-3 kg/m-sec and vsound ~ 1500 m/sec for human interstitial fluid at 310 K. Figure 7.1 summarizes Eqn. 7.6 for various parameter choices.

[b][i]The 1Mhz described above is *not* the wavelength, it is the damping rate[/b][/i] (which describes the sensitivity of a transducer’s damper – how fast does it go up and down). 1Mhz for a 2 micron diameter damper is truly awful by the way – that would be similar to a 2cm diameter damper have a damping rate of 100Hz. In this paper we have transducers made at 3 µm < l < 25 µm, 160 nm < w <1400 nm, t=219 nm that do 15MHz to 60MHz. The wavelength of 15MHz ultrasound is 22.93µm, 60mHz is 5.73µm (in agreement with the 3 µm-25 µm length). In this link we see a 9ghz transducer described - the patch is a 240nm x10 micron patch. The wavelength of 9ghz ultrasound is 38.22nm. The transducer is larger than that so we're good. Freitas describes a 2 micron wide transducer. The wavelength it would be most sensitive at would be about [b][i]175MHz plus.[/b][/i] Not even close to the 1MHz wavelength we need. [b][i]The transducer needs to be 200 microns wide, not 2 microns, to be senstive at 1Mhz ultrasound ranges[/b][/i]

Arenamontanus wrote:

To sum up, everybody wanting to analyse nanodevices ought to read Freitas book. Even when you don't buy his solutions (I have some serious problems with his histonation ideas and the size of the comms networks needed to do brain scanning) he has done his homework and made sure the formulas and physics is in there.

Uhm.....er....*sound of crickets chirping*...maybe after Freitas consults with an engineer on his rewrite?