Fast moving nanites are not hard sci-fi

45 posts / 0 new
Last post
NewtonPulsifer NewtonPulsifer's picture
Fast moving nanites are not hard sci-fi

So, some people have a vision of a fast, face eating flying nanite cloud.

This wouldn't actually work (in a hard sci-fi sense).

No matter your size, you're going to be operating in a speed of "body lengths per second."

Basically, the smaller you are the lower your top speed.

The top performers tend to be around 30-40 with endurance, with sprinters capable of up to 60-125 body lengths per second (sailfish, tiger beetle, cheetah)

So a swarm of micro-machines 10 microns in length is going to be moving along at say .4mm per second (40 body lengths per second). That's 1.44 meters per hour. Not enough to fight even the lightest breeze or air convection.

Basically, if a nanite cloud wants to move fast in a "hard sci-fi" manner, it needs to form up into an animal sized creature/construct first.

EDIT: Fixed meters per hour. Goofed the calc.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Thampsan Thampsan's picture
Good point, obviously

Good point, obviously favourable (or specifically set up) air currents would allow people to direct nanites. Things like the vortex ring gun.

Also could you use directed magnetic fields to pull nanites toward something at a greater rate?

NewtonPulsifer NewtonPulsifer's picture
Thampsan wrote:Good point,

Thampsan wrote:
Good point, obviously favourable (or specifically set up) air currents would allow people to direct nanites. Things like the vortex ring gun.

Also could you use directed magnetic fields to pull nanites toward something at a greater rate?

Right, control the environment and you can control the nanites.

So if the nanites are ferromagnetic, or ionized, or are partly made of room temperature superconductors you could pull/push them with magnets.

This could be exploited as a strength (to aid them) or a weakness (to hinder them), however.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Thampsan Thampsan's picture
Naturally, but it's something

Naturally, but it's something interesting to note for traps/defenses in a habitat.

OneTrikPony OneTrikPony's picture
I'm not sure I understand

I'm not sure I understand this discussion but seems from the examples in the OP that it's based on some bio-mimic means of propulsion; a system of levers and actuators like animal limbs.

Is it impossible for nanites to use some system of jet propulsion, magnetic inductance, or even something more exotic? It seems to me, if nanites can manufacture materials very fast they could also move very fast by using some of the same cooperative techniques.

edit:
speaking of bio-mimicry in nanite propulsion, a spider's spinnerets come to mind. Spiders make webs at a very fast rate.

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

NewtonPulsifer NewtonPulsifer's picture
An SR-71's body lengths per

An SR-71's body lengths per second is 32.

So I'm not optimistic.

EDIT: Nanites can't manufacture things quickly. That's not hard sci-fi either.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Thampsan Thampsan's picture
What sort of manufacture

What sort of manufacture speed are we talking roughly?

King Shere King Shere's picture
Dust Devil

Howabout Nanites that are on the ground or near ground heat segments of surrounding air or in otherways manipulate it to create the dust devil (or Fire whirl) phenomena.

Quote:
Dust devils form when hot air near the surface rises quickly through a small pocket of cooler, low- pressure air above it. If conditions are just right, the air may begin to rotate.

http://en.wikipedia.org/wiki/Dust_devil



"To find fault is easy; to do better may be difficult."
Plutarch

Jaberwo Jaberwo's picture
What about Nanites that

What about Nanites that condense in a liquid like form, and then flow around on surfaces, maybe even as an invisible film? Or they change form to go forward, perhaps an undulating ribbon of a few centimeters?
How fast can Nanites form millimeter sized structures?

Some thinking on my part:
The Tiger Beetle can move at up to 123 body lengths per second.
A SR-71 may move at 32 bl/s but it still moves 10 times faster than the fastest bird.
So if we build a technological Tiger beetle it might also be ten times faster.

Of course there is a difference in size between most birds and planes, but in general most technological things are more powerful than their biological counterparts.

So if we have a very fast Nanobot with a velocity of say, 2000 bl/s. With a bl of 10^-5 m it moves at 20cm/s, which is still pretty slow.

Could someone explain this by more conventional physics?

EDIT: Oh, did you just ninja me with these calculations or did I miss them?

NewtonPulsifer NewtonPulsifer's picture
Thampsan wrote:What sort of

Thampsan wrote:
What sort of manufacture speed are we talking roughly?

Well, we've established that the time scale doesn't really vary much.

Assume unlimited energy input. They're tapping into the power grid.

It took humans 26 months to build the 324 meter tall Eiffel Tower. If you assume as many nanites are assigned to the task as human workers were, and another unknown number of those constructor nanites are dedicated to providing resources, infrastructure, and transport (probably like 10x).

Assume the nanites work around the clock and don't take weekends off, and they could improve productivity by 4x.

Assume the nanites don't care about energy efficiency and just want to work at the most hyperactive rate possible, another 10x.

And then throw in another 4x by assuming the nanites have a more "evolved" design for this project and are worth 4 workers each.

You're at 160x faster. About 5 days to build a tower 160x taller than one of the nanites.

So if the nanites were 10 microns, it would take them about 5 days to build a 1.6mm tall Eiffel Tower.

There were 371 workers and engineers according to my source below.

Call it 4081 workers adding in all the infrastructure helpers.

At its inauguration the Eiffel Tower weighed about 11,000 tonnes. Compared to the mass of 4081 humans at 80kg each (326 tonnes total) then the nanites could build something as massive as about 6.75 times their mass per day.

EDIT:

Found a webpage with the number of workers. No idea how good of a source it is.

http://www.eiffel-tower.us/Eiffel-Tower-Facts.html

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

OneTrikPony OneTrikPony's picture
Um... I think I'm even more

Um... I think I'm even more confused now. And by confused I mean not being critical in any way what-so-ever. OK? (also; in a way, I'm playing devils advocate I've argued against the timeframe of nanobabrication as presented in the setting for a while now.)

As someone who was an ironworker for more than a decade I'm not seeing the utility of an analogy between men building the Eiffel Tower with cranes and Hot Rivets and Molecular Nanotechnology in which the Eiffel tower would build itself.

I'm not making an attack here, but I'm confused as to why one would restrict themselves to using 4000 10 micron Assemblers on any project. That hardly even counts as a swarm does it?

The speed of molecular manufacturing is determined by the number of metallic and covalent molecular bonds you can break and connect per time unit. The limit on that speed will be the amount of waste heat you can remove from the "Job Site" before it cooks your assemblers. Molecular assembly of the Eiffel Tower wouldn't be as quick as casting it out of molten steel, but it would probably be a bit quicker than 26 months. And I think that you could do a bit better than 1.6mm in ten days. (this assumes that everyone who argues against Drexler about viscosity, Brownian motion, and sticky fingers is wrong and molecular manufacturing is possible at all)

Also; While I understand the stated limitations of independent sub-micron machines floating in a gassious environment and their mobility I'm not sure that 10 microns is a good benchmark for envisioning the size of molecular assemblers. That's just slightly larger than a red blood cell but the Ribosomes in that cell which assemble proteins are definitely on the nano-scale.

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

Justin Alexander Justin Alexander's picture
OneTrikPony is right: There's

OneTrikPony is right: There's a lot of really awful argument-from-analogy going on in here.

"The highest any animal has flown is # times its body length. Ergo, man will never reach the moon and any science fiction featuring a moon landing is not hard SF."

LatwPIAT LatwPIAT's picture
This website gives a figure

This website gives a figure of 1 gram of product per kg of manufacturing volume per second for the Phonenix nanofactory design.

@-rep +2
C-rep +1

Jaberwo Jaberwo's picture
I tried to factor in the

I tried to factor in the improvements man does to his abilities by using technology, but it still is pretty slow. Also the analogies are not only about animals, but human constructs as well.

While it doesn't mean it's true that nanoswarms are slow, it seems right at first glance, as there is almost no vehicle or robot or similar thing that have a speed of more than a few dozend body lengths per second. (If you don't count rockets, but you wouldn't count bullets either)

Still I would very much like to hear some more traditional explanations to why Nanobots are slow, too. With equations and stuff. :)

And when we're done with that, we should start to look for alternative ways to make swarms move fast, because we all would like to have face eating swarms chasing people.

Concerning the molecular assembler swarm, I belive that Drexler first thought about them and then later said they are inefficient compared to dedicated nano-factories.
I bet someone here can post a link.

NewtonPulsifer NewtonPulsifer's picture
LatwPIAT wrote:This website

LatwPIAT wrote:
This website gives a figure of 1 gram of product per kg of manufacturing volume per second for the Phonenix nanofactory design.

This analysis is based on a 1 micron block size, and is building 86.4 times its mass per day.

If you constrained our 10 micron builders to "finger sized" bricks in hyperactive speed mode (presumably they'd need to slow down quite a bit to thread a needle) you're probably looking at a 1cm cube (finger thickness) for a 200cm tall human. So our "resolution" is 50 nanometers, or approximately 1/20th of this design.

If you go for a longer length, we've already established you can achieve a higher top speed. So if you increase our 6.75 by a factor of 20, we get 135 per day. Actually a larger number.

Also, this supposes a large factory that is much larger than our design size. You'd expect a factory the size of Paris to be able to build Eiffel Towers more efficiently than a "ground up" approach.

This probably means my multipliers were too generous (10x for hyperactive rate, 4x for more efficient design).

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

NewtonPulsifer NewtonPulsifer's picture
OneTrikPony wrote:

OneTrikPony wrote:

I'm not making an attack here, but I'm confused as to why one would restrict themselves to using 4000 10 micron Assemblers on any project. That hardly even counts as a swarm does it?

I'm trying to establish their gross efficiency (worker mass to project mass). Its questionable that even 371 workers is possible - keep in mind they're moving 10x faster relative to human workers. They'd run into each other. Sure you could throw more workers at it, but they'd have to slow down. Your energy efficiency would go way up, but not your work accomplished per worker.

It seems you're might be implying you can throw an arbitrary number of workers at a problem to make it get done arbitrarily faster - this isn't true.

If a pregnant woman isn't having a baby fast enough, then throwing 99 more pregnant women at the problem just means you get 100 babies in 9 months.

Also, I'm not assuming assemblers as defined in that link, I'm assuming purpose built micro machines designed for this job. You'd probably have a dozen different varieties easy.

OneTrikPony wrote:

The speed of molecular manufacturing is determined by the number of metallic and covalent molecular bonds you can break and connect per time unit. The limit on that speed will be the amount of waste heat you can remove from the "Job Site" before it cooks your assemblers. Molecular assembly of the Eiffel Tower wouldn't be as quick as casting it out of molten steel, but it would probably be a bit quicker than 26 months. And I think that you could do a bit better than 1.6mm in ten days. (this assumes that everyone who argues against Drexler about viscosity, Brownian motion, and sticky fingers is wrong and molecular manufacturing is possible at all)

Man already does atom by atom construction - electroplating and vapor deposition. It is done at the fastest rate the atoms will bond - and it is really slow because the layers are really thin. The bonding rate is not at all limited by heat.

OneTrikPony wrote:

Also; While I understand the stated limitations of independent sub-micron machines floating in a gassious environment and their mobility I'm not sure that 10 microns is a good benchmark for envisioning the size of molecular assemblers. That's just slightly larger than a red blood cell but the Ribosomes in that cell which assemble proteins are definitely on the nano-scale.

The right size of molecular assemblers is the largest possible one capable of working down to the "resolution" the job requires. Bigger assemblers will get the job done faster (back to the body length to top speed issue) but won't be able to work on finer details.

Re: ribosomes are an example of why molecular manufacturing is going to be very slow when dealing with stronger bonds. They use enzymes with bonding strength on the order of 2600-5000 kJ/mol to unzip DNA that has 30kJ/mol bonds, and due to that difference they're pretty fast.

That's about a 100x difference. To have an analogy to that, to copy what Ribosomes are doing on the stronger bonds out there you'd need something on the order 260,000 kJ/mol bond strength.

I'm not aware of anything with that strong of a bond strength.

So you're stuck with relatively small differences, which means things will operate much more slowly (if at all).

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Eleazar Eleazar's picture
Maybe I don't have enough

Maybe I don't have enough knowledge on this subject, but why is body length per second an important factor here? I mean, I just looked up an surface-to-air missile and calculated it's body length per second. It was 323, which I think pretty definitively shows the limits of applying biomechanical arguments to non-biological things.

Nanomachines aren't biological organisms. They could mobilize by generating electromagnetic fields which only strengthen as more of them clump together. They could build themselves microrockets for quick bursts of speed. There are tons of possibilities! The real limit is manufacturing time. If the swarm is not prepared to move quickly, then it will take time for it to ready itself to do so.

NewtonPulsifer NewtonPulsifer's picture
I'll finish the calculation..

I'll finish the calculation...

A 4 meter rocket might make 120 kilometers before running out of fuel, to deliver a warhead that's 10% of the starting mass.

That's 30,000 body lengths total.

So if the rocket was 20 microns instead it would travel a total of 60cm at a rate of about 5mm per second for 2 minutes.

That's 18 meters per hour (.018 km/hr)

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
So could a swarm then clump

So could a swarm then clump together and use some sort of magnetic power to act like some kind of amoeba? possibly capable of striking out fast over shorter distances like say the T-1000's blade arm from Terminator 2?

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

Eleazar Eleazar's picture
That assessment makes no

NewtonPulsifer wrote:
I'll finish the calculation...

A 4 meter rocket might make 120 kilometers before running out of fuel, to deliver a warhead that's 10% of the starting mass.

That's 30,000 body lengths total.

So if the rocket was 20 microns instead it would travel a total of 60cm at a rate of about 5mm per second for 2 minutes.

That's 18 meters per hour (.018 km/hr)


That assessment makes no sense. You're treating a nanoscale device that probably has to deal with quantum effects as if it is a macroscopic device that has to deal with air resistance, heating due to friction, etc. The things are not at all similar, so why choose to make these arbitrary assumptions?

[edit]
As a further example of how bizarre your analysis is, try a bullet. 9x19mm Parabellum. I'll just use Wikipedia to grab a velocity - 390m/s. Not even +P. Body Length per Second: 13135, roughly.

Again, you are applying assumptions to a situation that is not related to the situation from which those assumptions were derived.
[/edit]

NewtonPulsifer NewtonPulsifer's picture
Yes, it is bit silly to

Yes, it is bit silly to compare objects on a ballistic trajectory to nanites that are not on a ballistic trajectory. Are you proposing nanites being fired out of cannons as some sort of range extending solution? That would change the constraints.

It's also silly to consider a rocket, with about a 2 minute endurance, as a viable comparison to a nanite swarm. 2 minutes just isn't long enough to be useful for nanites.

Yes calculations for bullets aren't the same as rockets, as bullets don't carry their fuel with them.

However, unfavorable scaling laws do apply to anything that carries their own fuel with them.

Take a rocket with 1 unit of frontal area impinged on by the air. Lets suppose its top speed is limited by drag. Halve the length, width, and height of that rocket.

You now have a frontal area 1/4 the size. Excellent, we've cut drag by 3/4!

But your volume (and thus fuel) has now been cut to 1/8 the original amount. Your range has been cut. Would this apply to a chemically launched bullet? No, as the bullet was pushed out of the barrel based on the rearward area of the bullet (which scales the same as the front).

So you can't escape this scaling problem if you carry your fuel with you (Compare the range of a SCUD to a Katyusha, for example). And yes, again, this doesn't apply to 9mm bullets.

Are there other scale issues? Yes, drag changes qualitatively on very small scales. It actually gets worse as you have to expend energy to deal with skin viscosity issues, not just pushing the air out of the way. It just makes flying nanites that much more infeasible. I'd avoided that issue as it is besides the point - flying nanites were infeasible even before considering this scaling issue.

Also, Brownian motion becomes an issue. Once again, this is worse the smaller you get. It's just more reason a flying micro-machine would have to expend more energy. There's no reason to bring it up if bigger problems exist on the macro scale.

There's a reason the smallest wingspan you'll see among flying insects is about 1mm. It is because of these physics.

But here's the crux of the body lengths per second issue - if you shrink a 2 meter tall human to 20 centimeters in perfect scale, and shrink the street he's about to cross on a similar scale, it will take them both the same amount of time to look both ways across the street and then walk across it.

Other things being equal, you reduce body length by 10x, you reduce speed by 10x.

P.S. Objects at the 1 micron scale don't have huge quantum noise issues. Things at the 1 nanometer scale are another story.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

NewtonPulsifer NewtonPulsifer's picture
Prophet710 wrote:So could a

Prophet710 wrote:
So could a swarm then clump together and use some sort of magnetic power to act like some kind of amoeba? possibly capable of striking out fast over shorter distances like say the T-1000's blade arm from Terminator 2?

Electric motors operate on magnetism, right? So I'd say yes, there's no physics reason you wouldn't be able to move a cooperative man-sized blob made of micro-machines using magnetic power. That would be more of an engineering analysis, which admittedly I haven't done.

As for striking out like the T-1000 from Terminator 2, if it was striking out it's .8 meter long arm out to 2 meters at 1/20th of a second, you'd basically be moving 1.5 "body" lengths (.8 meter long arm divided by moving 1.2 meters) in 1/20 second, or 30 body lengths per second.

Fast, but still hard sci-fi fast :)

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
Are there any theories that

Are there any theories that surround getting around these kinds of difficulties? Other than game magic and pseudo-science of course. Arguably there would be an issue coming up with a need for such a fast moving nano swarm other than horror. I really cannot think of one other than rapid dis-assembly of an incoming hostile element like a missile or something and there are already defenses that work quite well against such.

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

NewtonPulsifer NewtonPulsifer's picture
None that I can think of in a

None that I can think of in a hard sci-fi way. Even if you could overcome the speed issue, you'd end up using all of your energy in fractions of a second.

I suppose you could beam in the power, but the swarm would have to stay in range and line of sight of the emitter to keep getting enough energy to fly.

EDIT: my definition of overcoming the speed issue still couldn't catch a guy on a bicycle

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Smokeskin Smokeskin's picture
NewtonPulsifer wrote:I'll

NewtonPulsifer wrote:
I'll finish the calculation...

A 4 meter rocket might make 120 kilometers before running out of fuel, to deliver a warhead that's 10% of the starting mass.

That's 30,000 body lengths total.

So if the rocket was 20 microns instead it would travel a total of 60cm at a rate of about 5mm per second for 2 minutes.

That's 18 meters per hour (.018 km/hr)

You can't just divide everything by size.

A rocket will push reaction mass back at the same velocity regardless of size. If we scale back both mass and reaction mass burned per second down by the same factor, acceleration remains the same for both small and big rockets, and so is velocity and range.

Of course air resistance is a big factor, as that depends on cross sectional area and the cross section:thrust ratio is much higher for small rockets, and micron-scale rocketry may be doubtful. But the principle of just dividing by body length is flawed.

Smokeskin Smokeskin's picture
NewtonPulsifer wrote:

NewtonPulsifer wrote:

But here's the crux of the body lengths per second issue - if you shrink a 2 meter tall human to 20 centimeters in perfect scale, and shrink the street he's about to cross on a similar scale, it will take them both the same amount of time to look both ways across the street and then walk across it.

Other things being equal, you reduce body length by 10x, you reduce speed by 10x.

No. Human strength depends on muscle cross section so that is reduced by the square of the length reduction. Weight depends on the cube on the length reduction. Your strength:weight ratio improves as you get smaller and that makes you faster, all other things being equal. I briefly considered running dynamics and a foot stands still on the ground or it is moving forwards in the next step. Obviously your run speed is equal to the average speed of your foot. If we assume that the ability to accelerate the foot is the limit to your speed (and I don't think this is the case because the pushing forward of your body weight seems more important) then an interesting thing occurs because being taller you have a longer gait so you have a larger length and longer time to accelerate your foot to a high speed than if you were shorter. If you do the math on that, then it turns out that the 10x better strength:weight ratio allowing you to accelerate your foot 10 times faster is exactly cancelled out by only having 1/10th the length to accelerate it on before you have to touch the ground again. Just a bit of fun to illustrate how important the scaling laws and which of them that matter regarding speed are.

If you actually look at body length per second speeds http://www.pbrc.hawaii.edu/~petra/animal_olympians.html for running you have horses at 7, humans at 11, cheetahs at 20, cockroaches at 50, ghost crabs (smaller than cockroaches) at 100 and tiger beetles at up to 171.

Aquatic organisms the .1 cm copepod can move at 500 body lengths per second.

For flight, the small animals easily dominate. A fighter jet at 3,000 km/h only does 150 body lengths per second while a humming bird does 215 and the winner is a moth doing 432.

A grasshopper jumps at a speed of 60 body lengts per second while a plant lice can do 1316 body lengths per second.

Small animals seem to consistently beat the larger ones in speed per body length.

Another thing you have to consider is evolution. Muscle is highly energy costly so there's significant selection pressure against unneeded muscle. This research http://www.ncbi.nlm.nih.gov/pubmed/10049474 suggests that absolute speed is not as predictive of prey survivability as body length speed. Animals are not going to develop muscle for unneeded speed. Predators are going to go for the easiest prey, and if the ease of catching prey is determined by body length speed, that's likely to be the number the animals escaping a certain predator will be circling around. Animals are not optimized for speed, they're optimized for survival, so smaller animals are likely to be saving on muscle because they just need to be a bit faster body length wise.

Solar Solar's picture
And given that Nanobots are

And given that Nanobots are smaller than plant lice, one would presume that Nanobots can move at a rather impressive speed for their size. However, on a macro scale it's really hard to actually see how fast that would be.

So nanobots, both in speed and in their ability to deconstruct/construct matter are a lot slower than popular media would suggest, as well as being very power intensive (and, for the paranoid, more easily compromised). That's an interesting thought for Eclipse Phase because it suggests that simpler objects built from basic, easily available materials for low cost aren't nano-fabbed. The construction industry does still exist then, albeit entirely automated (probably run by indentured infomorphs).

NewtonPulsifer NewtonPulsifer's picture
Some of the high numbers in

Some of the high numbers in that animal olympics link need caveats - the flying ones are diving flight, for instance (getting gravity added in, which distorts things more the smaller you are - dropping a 2.5mm ball bearing that reached a terminal velocity of 88m/sec is 35,200 body lengths per second - but isn't that useful to this discussion) - the copepod and plant lice were leaps, and counted only the peak (not average) velocity of the leap (also followed the arc of the leap, not based on lateral distance travelled). I can't take anything from the tiger beetle running for its life on sand though - 100/171. That's just badass.

EDIT: The 215 for hummingbird is indeed level flight, however.

EDIT: Was off by a factor of 10 for BL/s for ball bearing.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

NewtonPulsifer NewtonPulsifer's picture
Here's another scaling

Here's another scaling problem:

Processing power.

A human has a 1200cc brain, with 100 billion neurons (and a bunch of other brain cells too).

EP has something better - the cyberbrain. About 1cc in volume, it can match a biomorph brain, at probably something like less than 1 watt instead of 25 watts of power used.

So lets shrink that cyberbrain down to 1 micron cubed. How many cyberbrain neuron equivalents can we fit into that? Not even a single one.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

NewtonPulsifer NewtonPulsifer's picture
Smokeskin wrote

Smokeskin wrote:
NewtonPulsifer wrote:

But here's the crux of the body lengths per second issue - if you shrink a 2 meter tall human to 20 centimeters in perfect scale, and shrink the street he's about to cross on a similar scale, it will take them both the same amount of time to look both ways across the street and then walk across it.

Other things being equal, you reduce body length by 10x, you reduce speed by 10x.

No. Human strength depends on muscle cross section so that is reduced by the square of the length reduction. Weight depends on the cube on the length reduction. Your strength:weight ratio improves as you get smaller and that makes you faster, all other things being equal. I briefly considered running dynamics and a foot stands still on the ground or it is moving forwards in the next step. Obviously your run speed is equal to the average speed of your foot. If we assume that the ability to accelerate the foot is the limit to your speed (and I don't think this is the case because the pushing forward of your body weight seems more important) then an interesting thing occurs because being taller you have a longer gait so you have a larger length and longer time to accelerate your foot to a high speed than if you were shorter. If you do the math on that, then it turns out that the 10x better strength:weight ratio allowing you to accelerate your foot 10 times faster is exactly cancelled out by only having 1/10th the length to accelerate it on before you have to touch the ground again. Just a bit of fun to illustrate how important the scaling laws and which of them that matter regarding speed are.

A good criticism.

A 1/100th scale human would have 1 millionth of the mass, and the acceleration of its legs would be 1/100th of the full size human.

force=mass*acceleration

There's a direct relationship between force and energy usage (although different means of locomotion provide different efficiencies of force per unit of energy), so this 1/100th scale human could increase the acceleration of its legs 100 fold and still maintain the same amount of energy endurance as the human (in a spherical animal sense anyways). However, acceleration needs to increase as the square of velocity. So this 1/100th scale human could attain 10x the body lengths/second as the full size human (say 60BL/s vs. 6 BL/s).

So a 2 micron (1 millionth the length) scale human could attain say 6000 BL/s vs. the humans 6 BL/s for the same amount of power output per mass.

That's still only 12 milimeters per second (43.2 meters per hour).

Also, the relatively smooth concrete isn't going to be smooth at all for the 2 micron creature, but rather more like the Rocky Mountains.

In addition, we haven't added in the deleterious effects of viscous skin drag that becomes prevalent at those scales, nor the energy required to offset Brownian motion, nor of friction of its feet etc. It's still a spherical animal solution for the micro-machine and it isn't looking good even then.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Jaberwo Jaberwo's picture
So what's next?

So what's next?

For those who don't think that the problems you mentioned exist, everything is fine. They can continue as usual, it's a game afterall and we can't be sure.

But I am quite convinced now, so I would like to know how we can fix this or how swarms can still be used in an effective way.

RustedPantheress RustedPantheress's picture
Well... I wouldn't call

Well... I wouldn't call transhuman nanotech that fast (although seriously, the advanced can move fast enough to avoid being scattered by a light wind and be able to do something), and the TITAN nanoswarms are not hard science. Then again, what about the TITANs is hard science?

Somebody is using bad science! Snark, facts, snark.
Your body is corrupted: Cool, do more science to it.
Your mind is warped: That's nice, want a cookie?
What do we say to the God of Death? Not today!

Jaberwo Jaberwo's picture
Well uh.. I thought we just

Well uh.. I thought we just established that they are indeed very slow, at least if you draw the same conclusions as NewtonPulsifier and some others do. They do get scattered by a light wind like smoke or a gas. There has never been any official information on how fast they are.

And please don't bring the TITANs into this, this is about 10 AF transhuman technology, nothing else.
No spoilers please, I'm a player not a GM. :)
(I don't even read about psi, I think it's just a conspiracy theory from the mesh, if I even know about it at all)

So under the premise that they are slow
(if we can't agree on that we don't need to talk about it, and everyone does his own thing, no harm done)
What can we do about it?

NewtonPulsifer NewtonPulsifer's picture
Jaberwo wrote:

Jaberwo wrote:

What can we do about it?

Not much with EP tech, unfortunately.

EP is pretty conservative with power/weight ratios.

Take the mobility frame for instance - doubles your long distance range before you get tired (implying it puts out as much energy as a regular human does) and is 11kg. An 80kg human has about 36kg of muscle - 22kg in the legs is believable. So for half the mass, we output the same power.

For burst it is better - it increases top speed by 60%, so that implies 2.56 times the power (1.6x1.6). So +156% for burst rather than +100% for endurance (and for half the mass).

Assume a 2x regular 3.12x burst power output if we replaced the meat muscle with robo-muscle entirely.

Normally a bumble bee can go about 4 meters/sec (movement rate 16 in game, 160 body lengths\sec)

Use that boost to make them faster - about 7 meters/sec (movement rate 28 in game).

I think bumble bee size the smallest you're going to be able to do for face-eaters.

Being bumble bee sized also gives them more potential for processing power (neuron equivalents). If a cyberbrain is 1cc and about 3.5 grams, then a 0.5gram bee-bot could dedicate 5% of its mass to its brain - putting it into the dog range of intelligence.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
I think the development

I think the development behind faster nanites would be then based on a faster delivery system. Have something the size of a cat carrying a sizable swarm for quick release upon contact, or some kind of vehicle behind it for fast fogging like a tear gas grenade, having the nanites ride the chemical compound as they were expelled from a pressurized canister.

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

Jaberwo Jaberwo's picture
I agree, for movement

I agree, for movement purposes micro sized bots won't do it.
Are you considering insect sized robots as the smallest size or as a form that is assembled from smaller bots?
(If the latter is possible I would still think there are swarms with bee sized robots as the smallest unit. It might act as the kind of delivery system Prophet710 suggested)

Like for example when a move command is given they condensate (presumably not without losses) and then evaporate again to act as a smart and stealthy combat chemical.

I could also imagine some other forms of macro objects for movement, (but I'm not sure if they make sense):
A fluid flowing along walls either as a thin film or rolling droplets.
A ribbon undulating at the right frequency to move through air.
Some kind of 'muscle tissue' where several nanobots work together to move others, perhaps as a thin film.
Worm or snake like constructs of varying sizes.
Big thin sheets (perhaps completely transparent at most wavelengths) that use air currents.

Concerning the brain/neuron analogy I'm not sure if I agree, but I need to do some searching on it.
It's just that I've never read about computers being a problem in nanomachines.

But anyway, dog is good enough considering that we can network them and specialize the AI for its task.

NewtonPulsifer NewtonPulsifer's picture
How do you propose the

How do you propose the nanites communicate? Should I start a new thread

"Nanite swarm communication is not hard sci-fi?" :)

Or should we leave it in this thread?

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
I would think it would be all

I would think it would be all one network, as in the swarm is one machine, operating with several moving parts to one end. without equipping each nanite to have a communications suite using LIDAR or even radio, you could just have a pre-programmed routine set in that could be adjusted by the user using the containment mechanism as a beacon. Maybe?

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

NewtonPulsifer NewtonPulsifer's picture
Prophet710 wrote:I would

Prophet710 wrote:
I would think it would be all one network, as in the swarm is one machine, operating with several moving parts to one end. without equipping each nanite to have a communications suite using LIDAR or even radio, you could just have a pre-programmed routine set in that could be adjusted by the user using the containment mechanism as a beacon. Maybe?

Are you saying they'd be physically connected?

EDIT: Bringing up the viability of swarm communication was more in response to Jaberwo's topic of condensing/evaporating swarms.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
No not physically connected

No not physically connected no. they could be until released, and then just pre-programmed with a specific task until finished, possibly open to certain physical communication to change the routine such as a burst of air pressure from the containment vessel or even a burst of radio signal. Still individual nanites though.

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

Jaberwo Jaberwo's picture
I don't propose anything :)

I don't propose anything :)
But I guess there is a problem, judging from your earlier comments in the 'Anti-nanite weapon - Flamethrower?' topic.

How about infrared or chemical cues?
Apart from communication, what do you think of these condensed forms?

I want face eating Nanomachines. But I want them to be believable in a hard sci-fi way.

So like always when you are discussing science and engineering that doesn't exist yet, you should work towards things that are fun and bend assumptions for your own good. That may be bad science, bad it's not bad (hard) science-fiction, as long as you don't violate anything fundamental and don't become inconsequential or too implausible. At least thats my approach.
So lets try to make them work.

OneTrikPony OneTrikPony's picture
Putting aside all bio

Putting aside all bio-mechanical analogies to movement limitations. I'm still not certain that Movement is really an issue when it comes to face eating.

It's true that your swarm, in mist or cloud form, isn't going to go chasing people around a habitat. But, in the first place, the swarm doesn't have to be mobile to eat faces because characters are mobile and a swarm is nearly impossible to detect until you're already in it. In the second place, it doesn't need to be a mist it can be a self propelled liquid.

Consider coordinating an engineer or protean swarm with a disassembler, injector, or saboteur swarm. The engineer swarm fills a space with microscopic lattice tubules and vesicles that carry the aggressive nanites. A charter walks or runs through the space and gets infected. It would be nearly impossible to detect because none of the nanites are free floating.

You don't even need an aggressive swarm. Ever walked through a spider web by mistake? Use the same engineer swarm to fill a floatway with microscopic fullerine "spiderwebs" and chase the character into it. Give them 5 - 10 meters before they become entangled by enough of these webs that they have to make a Som test to fight thier way out.

Pack an aggressive swarm inside a condensed engineer or protean swarm with Chameleon capabilities. It can roll or flow anywhere it wants, (especially quickly in lunar or microgravity), because it's composed of nano-particles it will have awesome traction. It could disguise itself as any form that can be programmed. Your characters will NEVER trust another toilet seat.

As for the communication issue; Just have the swarm work like mesh inserts or the recording end of a cortical stack. It can form and break physical connections as quickly as needed to carry out distributed processing and should have processing power at least equivalent to an ecto.

Mea Culpa: My mode of speech can make others feel uninvited to argue or participate. This is the EXACT opposite of what I intend when I post.

NewtonPulsifer NewtonPulsifer's picture
OneTrikPony wrote:

OneTrikPony wrote:

It's true that your swarm, in mist or cloud form, isn't going to go chasing people around a habitat. But, in the first place, the swarm doesn't have to be mobile to eat faces because characters are mobile and a swarm is nearly impossible to detect until you're already in it.

I don't think that a swarm is nearly impossible to detect. 10 cubic centimeters of dust (one swarm) spread over the floor of a 10x10 meter area is going to be very visible. If you propose larger cooperatives, its going to be several 100s of cubic centimeters. On top of that, advanced sensors are very common in Eclipse Phase, and especially among Eclipse Phase characters.

OneTrikPony wrote:
In the second place, it doesn't need to be a mist it can be a self propelled liquid.

I'm not sure what a self propelled liquid is? Could you point to some real world examples?

OneTrikPony wrote:
Consider coordinating an engineer or protean swarm with a disassembler, injector, or saboteur swarm. The engineer swarm fills a space with microscopic lattice tubules and vesicles that carry the aggressive nanites. A charter walks or runs through the space and gets infected. It would be nearly impossible to detect because none of the nanites are free floating.

You don't even need an aggressive swarm. Ever walked through a spider web by mistake? Use the same engineer swarm to fill a floatway with microscopic fullerine "spiderwebs" and chase the character into it. Give them 5 - 10 meters before they become entangled by enough of these webs that they have to make a Som test to fight thier way out.

These would all have to be coordinated by hives, right? So that part is much easier to detect than the micro machines themselves. Micro machines would have to be harder to detect if they're "inside the walls" but here you seem to be proposing they're filling up open space with invisible webs.

Well, most spider webs are actually invisible to the human eye in terms of thickness. What you are seeing when you see a spider web is an effect called fluorescence, and unfortunately carbon nanotubes fluoresce really really well into the infrared. So with a modicum of light (or even not if your vision is extremely sensitive), anybody with Enhanced Vision can see the fluorescence of nanoscale objects even thought the threads/objects would normally be too small to see due to their thickness.

OneTrikPony wrote:
Pack an aggressive swarm inside a condensed engineer or protean swarm with Chameleon capabilities. It can roll or flow anywhere it wants, (especially quickly in lunar or microgravity), because it's composed of nano-particles it will have awesome traction. It could disguise itself as any form that can be programmed. Your characters will NEVER trust another toilet seat.

Chameleon as described in the rulebook isn't going to work with micro-machines. The layer would be too thick to be practical.Invisibility cloak style metamaterials have the same problem.

A cloak a millimeter thick isn't a big deal to something human sized - it's way too thick to be practical for a micro-machine.

And a T-ray emitter would show that faux toilet seat right away.

OneTrikPony wrote:
As for the communication issue; Just have the swarm work like mesh inserts or the recording end of a cortical stack. It can form and break physical connections as quickly as needed to carry out distributed processing and should have processing power at least equivalent to an ecto.

They can not form and break physical connections as quickly as needed. They need real time (as in, this isn't combat speed) to form and modify those links - the nanites aren't that mobile. And even if they were super mobile, their vision doesn't go past beyond say one body length ( except for possible very basic light intensity senses depending on their size) and they're deaf. So they can only coordinate by touch, smell, and by vibrating messages through a solid or liquid. It is possible they can pick up broadcast messages from say a hive via an infrared broadcast, but two-way communication isn't feasible.

And also I have a hard time believing that a network of nanites can add all their processing power together with no overhead.

That's like saying I have 100 million transistors in a single CPU chip. Then I split them 1 million to each of a 100 nodes (1 million transistors each), and then wire them up with network cards - but of course the network cards have hardware of their own which take 100,000 transistors each. I've just increased my total transistor count by 10% for 100 nodes (and we're being kind here -realistically you'd need routers/switches too) *and* you've introduced latency.

I doubt even with EP tech the photon fired out of my optical CPU can directly transmit itself down an optical cavity or cable and end up in the CPU of my neighbor. There is going to be translation overhead.

If a 10 cubic centimeter swarm was composed of 10 micron sized micro-machines, then we would have 10 billion nodes. The added overhead and complexity would dwarf the resulting computing output, and make coordination essentially impossible.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto

Prophet710 Prophet710's picture
So in essence then you would

So in essence then you would need each nanite to be pre programmed to pursue a certain end. This would essentially eliminate any kind of communications bottleneck but, it would inhibit any kind of on the fly re-dedication of your swarm. So much for adaptive nano-swarms.

"And yet, across the gulf of space, minds immeasurably superior to ours regarded this Earth with envious eyes. And slowly, and surely, they drew their plans against us."

NewtonPulsifer NewtonPulsifer's picture
Prophet710 wrote:So in

Prophet710 wrote:
So in essence then you would need each nanite to be pre programmed to pursue a certain end. This would essentially eliminate any kind of communications bottleneck but, it would inhibit any kind of on the fly re-dedication of your swarm. So much for adaptive nano-swarms.

To be fair a micro-machine could implement an infrared transmitter but the power draw would be so great it would really only be for very limited uses (and low bandwidth). Like screaming/shouting to warn the others, or somewhat useful communications (really no faster/better bit rate than verbal communications) if they were tapped into an electricity source and power draw wasn't an issue. This is also more for micro-machines around 100 microns in length, and thus potentially visible at 40 inches to a human's eyes. I suppose if you could build the micro-machine out of translucent materials it would still be unnoticeable at closer ranges, but that does seriously constrain the list of things you can build the micro-machine out of.

Demodex mites that live on humans tend to still be invisible because they're translucent, for instance (lots of water), but also because they freeze in light (and usually only start walking around at night). They camouflage themselves in skin pore sebum. They're around .2mm long so normally would otherwise be visible.

"I fear all we have done is to awaken a sleeping giant and fill him with a terrible resolve."- Isoroku Yamamoto