Robot Airplanes Are Here

One of the implications of molecular manufacturing is major advances in avionics. Materials 100 times as strong, motors and computers a million times lighter, and seamless, automated, rapid construction, will enable rapid R&D of highly advanced aerospace systems. Advanced aerospace systems could be used for easier space access (including rapid testing of advanced propulsion concepts), delivery of sensor platforms or weapons, personal transportation, and a number of other beneficial and scary applications. Unmanned aerial vehicles (UAVs) will become far more capable.

We can get a sneak preview of some of the concerns in a recent article at PhysOrg.com: "Flying robot attack 'unstoppable': experts" which describes some of the technologies that are already available, some of the ways they are being used, and some fears about the ways they might be used in the future. "The technology for remote-controlled light aircraft is now highly advanced, widely available -- and, experts say, virtually unstoppable." Small planes don't show up on radar. They can carry hundreds of pounds of payload. They can be guided by GPS. They have already been used by paramilitary and terrorist groups. The concern is that they could deliver explosives to a target with near-pinpoint accuracy.

The article doesn't mention it, but UAV technology is becoming available to hobbyists, and not only for sinister purposes. A group at MIT is building the "Freedom Flies family of UAV's" which can carry fifteen pounds of "video, gps units, pamphlets, water, food and other payloads." It is intended to cost a few thousand dollars, and to be open source (plans freely available). They admit they are still crashing quite often, but that will presumably change soon.

It remains to be seen whether private UAV development will be viewed with as much suspicion as, say, hobbyists blowing up junk for their own amusement (they tend to get called "militias" and prosecuted). There is certainly a contrast, even a tension, between the two websites linked above. A major positive use of UAVs might be to increase accountability by photo-documenting abuses that governments and corporations would prefer to keep hidden. However, from the point of view of those being watched, this is indistinguishable from spying. The article on terrorist uses did not talk about this, but it may turn out to be the biggest issue raised by UAVs.

UAVs are one tiny sub-category of one implication of molecular manufacturing. Even pre-MM technology is cause for concern, and capabilities are already coming online and becoming widely available to individuls and small groups. General-purpose molecular manufacturing will enable hundreds of equally significant advances--many of which no one has thought of yet.

Today's development of UAVs will give the world a chance to get used to at least one kind of robot that can go almost anywhere (that's not too small) and carry out actions remotely. This is probably a good thing. But it also means that software for UAVs will be pretty well understood by the time MM arrives.

The world is going to be a very interesting place.

Tags: nanotechnology nanotech nano science technology ethics weblog blog

Comments

[Tom Craver]

For a full-sized UAV, a million times lighter isn't much better than 10 times lighter. I.e. you might go from a 1000kg engine to a 100kg engine, then to a 1mg engine.

Yeah, it'd result in a different design, but the first 10x reduction gives 9x more reduction in weight than the next 100,000x, reducing lift requirements or increasing cargo or fuel capacity.

To make tiny UAVs, small size is the biggest factor, since most energy will go into overcoming drag rather than providing lift. Tiny weight is pretty much a given for tiny size.

[Chris Phoenix]

Tom, good point about engine weight scaling. But there are other advantages to shrinking the engine. Today, you need space for the engine; it's a big hunk of stuff in the middle of your UAV that you have to design around. Shrink the weight by 10x, you shrink the linear dimension by ~2x. Shrink the weight a millionfold, you shrink the linear dimension by 100x. That's worth doing.

Also, with (lots of) smaller actuators, you can do some anti-drag technologies.

Also, with smaller engines (and different technology, ie fuel cells), you can run them more efficiently. Fuel cells are not Carnot-limited.

Also, an advantage I didn't mention is that full-scale automated manufacturing implies that extra functionality doesn't cost you much.

Chris

[Tom Craver]

I don't want to argue that there's nothing worth doing for UAVs. But again, reducing the engine to 1/10th the weight probably implies 9/10ths of the volume is already eliminated - any further size reductions in the engine give you less than 1/9th the improvement in space.

Anti-drag may be worth doing, but you need to be careful. For example imagine coating the top and bottom of your wings with little wheels that spin at air speed, so that there's essentially no drag. But maybe it also means you've eliminated the Bernoulli effect - the air next to the wheels is effectively at a stand-still relative to the wheel surface.

And then there's a possibility that you just can't make that approach work - the wheels may spin themselves apart at any reasonable flight speed.

[Chris Phoenix, CRN]

Tom, if you just want to build today's UAV's better, then yes, reducing the engine weight by 9/10 is probably enough. But when you get into 2 or 3 or 5 orders of magnitude reduction, you can do new things.

For example, imagine a jet fighter that's 90% fuel. (Maybe 95% but I'll keep it conservative.) Most of the rest of the weight is missiles... *collapsible* missiles... hundreds of them.

The missiles are fueled/inflated from the jet's tanks. They have an airbreathing mode where they fly pretty efficiently for long distances. As they drain their tank, they replace the volume with LOX. Then they turn into a rocket--and whatever propellant is left over, they use as a warhead. I don't know explosives theory, but I'd guess that sub-micron intermixed tanks of LOX and fuel, set off at many closely spaced points simultaneously, would approximate a high explosive.

The jet isn't only a jet. It can wing-warp (small actuators and strong materials) and fly up to 100,000 feet, and go either slow/efficient or fast; in fact, what the hey, it can do the same LOX-gathering trick and go orbital.

Why try to fit all that functionality into one airframe? Well, why not? The engines don't weigh anything, after all...

As to anti-drag wheels on the surface of aircraft, Josh Hall has studied this; last I heard, he was talking about a system that could accelerate at some high number of G's in any direction. But what I was thinking of was turbulence management via microstructures that only had to sense and bend the airstream, not blast it 90 degrees sideways.

Chris

[Tom Craver]

Chris:
I'm not saying there's nothing interesting to be done. Just that saying "a million times lighter" tends to give the false impression "a million times better in some way", when in fact a 10x reduction of some factor like weight often provides 90% of any benefits.

Overall, I'm saying that improvement factors like 2x or 10x or 100x are probably more impressive to experts who might be put off by an apparent implication of "a million times better".

Fuel-LOX warheads could create a powerful concussive shock wave. But I suspect there's not a lot of gain from collecting the LOX in flight vs just launching with it on board, if you know that's how you'll be using the missile. You'll end up expending fuel mass to power the compressors, instead of just carrying LOX mass in lightweight vacuum insulated tanks. Where it might pay off is in flexibilty - do you use your reconfigurable missile mass for fuel-lox, or create an anti-personnel fragmentation bomb, or a penetrator? No point carrying LOX if you end up not using it.

Anti-turbulence wing surfaces are certainly of interest - but at best they'll drop drag to laminar flow levels, and a lot of research has already gone into minimizing turbulence of wings, so my guess is that again the gains will be useful but modest.

It might be interesting to consider a super-long-and-skinny missile with an anti-turbulence surface. For a fixed mass, length would increase as the square of radius reduction. Viscous drag would only increase about linearly with reduction of radius, while laminar flow drag reduced as the square of radius reduction.

So a cruise missile that launches as 5m long with a 30cm diameter might reconfigure to 3cm diameter and 500m long in flight, reducing laminar flow drag by 1/100th while increasing the (generally lower) viscous drag by only 10x.

Assuming a normal cruise missile expends most of it's fuel to overcome drag, that might mean around 100x as much range with the same amount of fuel, or else nearly 99% of fuel could be replaced with payload mass - possibly twice as much payload.