Nonist readers may recall a two cents post concerning NASA Institute for Advanced Concepts calling for proposals and wild ideas. The deadline is midnight February 13, 2006. I e-mailed my pdf off to them just now. now did they mean the deadline was forty minutes ago, or eleven hours and twenty minutes from now? no matter; i’m not buckin’ for a contract. i’m just a mailman.

a space mailman. yeeeaaahhhhh! (by the way, that’s a real spacecraft in the picture. but i digress. for your edification, here follows the text of my proposal e-mailed to NIAC some twelve minutes ago.)

Transmittal Letter
In response to: NIAC Phase 1 Call for Proposals, CP 06-01

Good day!
I am not affiliated officially with any entity in the aerospace field; I am a letter carrier with the U.S. Postal Service, that’s all. I am in no position to do anything with any contract awarded in the NIAC process; for this reason you will find no cost proposal attached. I am simply taking advantage of this uncommon opportunity to pass along several ideas I’ve been pondering, in the thought that you may actually find something in them you can use. In at least one case (that of Robotic Docking Tunnels) there is a company which is a logical party to contact about further research, as you will see.

4 Technical Proposals
I propose here four separate concepts for space flight and space industry, from technically simple yet untried, to technically demanding and large in scale.  If one theme runs through these ideas it might be expressed as “Biological design principles in unexpected places.”
1. Robotic Docking Tunnels.
2. Telepresence/Virtual Reality Force-Feedback System with strength augmentation and physical record/playback
3. Orbiting Fuel Farms And Water From Ceres
4. ‘Honeypot Ant’ Manned Spacecraft Design Concept

Yours,
Tom Buckner
(personal info)

1. Robotic Docking Tunnels
I don’t know whether anyone else has given flexible, self-guided docking tunnels much thought, but it seems they may soon be easy and cheap to build. Docking mechanisms could be at least partly flexible and muscle-fiber actuated, rather than rigid and mechanically complex, so that damage might not render mechanism unusable, and could be placed at the end of flexible tunnels (picture a vacuum cleaner with its hoses and rotating couplings; apply this idea to docking bays). Two flexible docking tunnels could have simple guidance aids, such as infrared lights and IR detectors, aligned so that two docked tunnels will have an IR detector directly over an opposed IR light; given the docking command the two tunnels automatically align as the robotic mechanism moves the detector toward the light. The basic idea here is very, very off-the-shelf. Twenty-odd dollars will get you a robot that can see well enough to follow a line:

Line Tracking Mouse
http://www.robotstore.com/catalog/display.asp?pid=686

Eighty dollars will get you a simple light-seeker:
http://www.hobbyengineering.com/H1187.html

A couple of hundred dollars will get you an OctoBot Survivor robot that rolls about and then seeks and docks with its power station:
http://www.robotstore.com/octobot.asp

This demonstrates that the electronic brain of the docking tunnels can be a cheap throwaway unit. If the electronics board goes, you could have ten more in the parts drawer. The hose could bend via segmentation and musclewire. The rotating couplings could be moved mechanically or by musclewire, or couplings might be obviated by using muslewire to torsion the whole tunnel (this gives less than full rotation, but that may be something you can live without; two actively seeking docking mechanisms would not necessarily need to rotate 360 degrees). Alternately, carbon nanofiber could be both muscle and skin (for another such use of nanofiber, see section below on orbiting fuel farms). The tunnel might be configured like an accordion, or like an elephant’s trunk.

In fact, most of the basic design work for such a tunnel has happened, except with milk instead of air:
http://www.nature.com/news/2003/030310/pf/030310-5_pf.html

This 2003 article describes a teat-seeking robot, “modelled on an elephant’s trunk,” developed in Great Britain. Quote: “Dairy farmers of the future may sleep safe in the knowledge that an udder-friendly robot is doing the day’s milking. “The idea is to replace farmers’ hands and allow cows to milk themselves whenever they fancy,” says engineer Bruce Davies at Heriot-Watt University in Edinburgh, UK. Davies’ company IceRobotics has just received a £98,000 (US$157,000) grant from Britain’s National Endowment for Science, Technology and the Arts to develop its rubbery manipulator - the ‘continuum activator’ - into a flexible, teat-seeking robotic arm.”

Their website,  http://www.icerobotics.co.uk/ has a bit more on the specs and contact info. According to a PDF report on this site, http://www.automaticmilking.nl/ , at least 700 dairy farms in Europe are using this type of equipment in 2005. You could probably do a consortium with IceRobotics and skunkwork a good docking-tunnel system in a matter of months.

Interestingly, a docking tunnel with manipulators at its end can replace the Canada type arm as well! With such an adaptation, an astronaut in shirt sleeves could use a manual station at the business end of the tunnel, in effect going to the business end of the ‘arm’ without having to suit up.

2. Telepresence/Virtual Reality Force-Feedback System with strength augmentation and physical record/playback
1. Overview
2. Some Design Details
3. Some New Uses

1. Overview
Present force-feedback telepresence systems (those of which I am aware) are still somewhat primitive in terms of immersiveness. Even very expensive and specialized ones consist of a sort of harness for the arm, articulated with but not actually enclosing the arm, hand or fingers. Since I wrote the first version of this essay, the military has come closer to fielding an exoskeleton designed primarily to augment strength, but it does not, in itself, seem to offer the sophisticated utility of the idea I proposed in Spring 2001 (the paper was not widely disseminated, but I did manage to get a copy into Marvin Minsky’s hands in June of that year, at a Discover Magazine awards show in New York; perhaps others read it after he did; he found no obvious flaws in it then).

I propose a novel usage of some old technology everyone has seen, with a layer of new components. This design is simple in essence but must become more complex as full immersion is approached. Early models probably must be custom-fitted and no doubt expensive to build but I am confident that CAD-CAM techniques will bring this situation into line.

The old technology is the articulated suit of armor. As in days of old, a well-fitting suit can follow the contours of the wearer’s body yet bend at each joint. A good approach to fabrication might involve scanning the wearer in a body suit with orientation marks (sequins, anyone?) These readings could then be used as in current laser-hardened resin prototyping, creating a complete set of plates covering the body from fingertip to toe. These may be mounted on a tight-fitting, flexible fabric base (Spandex! Shudders.) The result would be a set of several separate garments forming a whole, as in old armor, typically breastplate and torso, arm pieces, gauntlets, leggings, etc.

Nothing odd so far: now we add muscles. To do this, we need attachments at the ends of all plates so that musclewire (such as Dynalloy, Inc.‘s Flexinol™ wire) or biomimetic/artificial muscle (see http://www.unm.edu/~amri/ for Artificial Muscle Research Institute at the University of New Mexico) may be added to duplicate the actions of the body’s own muscles. The muscle wires may be on the inside of the plate or the outside, as design needs dictate. As in the body both flexors and extensors are needed in all appropriate locations. That is to say that some musclewires will open the fingers, some close them into a fist. Some will bend the arm, some straighten it. It will be seen that an artificial muscle will tend to augment the muscle directly below it or resist the muscle which opposes that one.

2. Some Design Details
If the plates must be custom-designed for each wearer, it will become necessary to standardize attachment points on the suit (Dynalloy’s barrel-crimp-at-both-ends option, for example.) This might be served by equally standard power-supply/control wiring. A wide range of muscle-wires would be needed, in a range of lengths and thicknesses from very short wires at the fingers all the way to large bundles at the thighs.

It may be that a suit made almost entirely of muscles with no articulated skeleton can be made to work; still, one expects that it would need to be fitted closely to the wearer’s dimensions.

Ultimately a fabrication system should be fully automated, proceeding from a body scan to generation of custom plates, their attachment to the flexible base garments, and the addition of appropriate muscles and wiring according to a database of measurements to give the proper range of motion. Obviously, a particular muscle (say, a biceps) from a person four feet tall cannot be the same as a biceps for one six and a half feet tall.

So far, the force-feedback system is different from present models mainly in the way it encloses the body. But with some consideration of how it behaves in use, we can see ways to extend it to realms of employment that go beyond what we have seen in demonstration.

We do not need to recreate every single skeletal muscle in the human body to get reasonably near fully immersive force-feedback. Even so, a full suit will probably incorporate perhaps two hundred discrete muscles (each hand needs about three dozen, half flexors and half extensors.) Each muscle would be a bundle of smaller musclewires (Flexinol™ wires capable of lifting two kilograms are currently available, so a major muscle capable of lifting fifty kilos must be fabricated from twenty-five such wires.) Since a human can only generate about a fifth of a horsepower, the basic suit might be able to run on household AC current; making adequate power portable and safe might be a large portion of the design challenge.
Each muscle can be controlled as if it were an individual ‘note’ in a MIDI instrument (in fact, I believe MIDI files can be created to control the suit!)

Let me illustrate how this might work.

Musical Instrument Digital Interface has been around for about twenty years and is thus mature and well-known technology. In a MIDI file, a synthesizer such as that in your computer’s sound card can play a given number of voices at a time, assigned to a different number of instruments. A typical MIDI file may have sixteen ‘tracks’ with different instruments assigned to them; each note might be assigned to a different frequency to be played by an oscillator or wavetable sample. Each note is played at a certain moment, at a certain volume, with a strong or weak attack, and may be short or long (the Sustain parameter.) More complex events are coded as a series of note events so that a note can bend, be sustained, etc.

Consider instead that each track covers a muscle group, and each note is assigned to a particular muscle. The ‘volume’ parameter may tell the muscle to contract a little, or to its maximum contraction. The ‘attack’ parameter can tell it to contract gradually or suddenly. The ‘sustain’ parameter would tell the muscle whether to release or stay contracted. All the wires in the bundle might act in unison or separately depending on control needs; one good way to get the current to them might be by using printed control circuits to close the contact with a sandwiched conductive layer (this is like many a keyboard or other control pad, except in reverse and with more electrical current.)

One final point: just as a MIDI musical instrument can record the finest nuances of a musical performance, so muscle suit might be equipped with small sensors (such as piezoelectric elements in the muscle attachment sockets) that record the movements of the wearer. A fully orchestrated song a few minutes long can be encoded in a MIDI file of only 200 kilobytes, and I suspect that MIDI code can contain all the information to record and control every movement the human body can make with similarly small files. This should not surprise us since human musical performances are in essence an expression of muscle movement, and in a sense we are simply turning the movement-to-music conversion backward. I will return to this idea under the following section on New Uses.

It can be seen that there are several modes of behavior for the suit, depending on our requirements:
A. Simple force-feedback. In this mode, the wearer moves freely through VR space (using the good old 3D goggles) and meets resistance from objects in that space. This is done by contracting the opposite muscle from that which the wearer is exerting. If I squeeze a virtual lemon, for example, I am applying force mainly through the flexors that close my hand. The suit must therefore contract the extensors on the outer surface of my hand to counteract my motion to the appropriate degree.

B. Force-amplification, the opposite of force-feedback. The same suit can counteract the wearer’s movements or detect and amplify these movements by simply being instructed to contract the corresponding muscle or the opposing one. A suit in this mode will increase the wearer’s strength to a degree dictated by the contractile strength of the musclewires and the rigidity of the suit plates. It may be necessary to use less muscle strength than the suit can generate to avoid injury to the wearer or others, either by using weaker musclewires or by imposing limits in the software signal.

C. Recording movement. This might involve instructing all muscles to contract gently so as to take up slack, and recording the degree, timing, speed of contraction they encounter.

D. Replaying recorded movement. In this mode, the wearer might learn to ‘go limp’ and let the suit move in accord with a performance that was recorded before.

E. Any combination of the above.

3. Some New Uses
Uses we have already seen for force-feedback arms include surgical procedures and other VR and gaming activities. I will close by mentioning some new uses I have considered.

A. Truly immersive VR and gaming, etc. Instead of moving a joystick, one actually climbs cliffs or swings a sword about. Obvious.

B. Strength amplification. This opens up a whole class of uses. One might:
Use for heavy lifting, in jobs that are now backbreaking but where no appropriate equipment exists; or as a replacement for current equipment that is far more cumbersome.

Use as a wheelchair replacement, enabling walking in those who have limited muscle strength; direct neural control might later enable use for para/quadriplegics. This application is expanded on in item F below.

Other uses may open up, depending on the portability of the power supplies that must accompany a mobile suit.

C. Telepresence. Again, obvious. Humanoid rescue robots may be remotely controlled (within time-lag limits, of course). Military and hazardous-duty uses abound. Exploring the surface of an uninviting environment such as Io or Venus via telepresence from a safe, shielded spacecraft in orbit only a fraction of a light-second away would be the next best thing to walking there. Doing the same thing from Earth would be more like looking over an AIBO’s shoulder and making suggestions that won’t be carried out for hours, while it wanders around trying not to fall off a cliff. Still, this seems feasible too.

D. Recording and Replay. This is an intriguing unexplored possibility.
Do you know how it feels to play Rachmaninoff’s infamous Third Piano Concerto or dance like Janet Jackson? If persons of talent consent to be recorded, the exact sequence of movements can be replayed while you wear the suit.

A new surgeon might feel exactly how an experienced surgeon performed an especially challenging maneuver, enhancing present methods. If the new da Vinci system does not record and play back movement, for example, it should; for that system, it would probably be a minor software tweak

More than this, the possibility of such direct learning may be amplified by new memory-enhancing drugs such as those called CREB activators. Such substances may (for a few hours) enable the brain to soak up input like a sponge. If such substances work in humans, this may represent a radical new wrinkle in the learning process. One might take the drug and spend the next two hours in VR tying knots, for example, and afterward be as knot-wise as any old sailor. One might master a musical instrument in a few weeks, a new aircraft in a couple of days. Any task which requires muscle-learning seems a natural for this. They may prevent memory loss as well, and I would think they will be out of the labs and onto the black market in about five minutes. I want some ASAP.

It might be best to use less than full mechanical force in this context so than the limbs are guided rather than forced to the point of possible injury. Since this may result in sloppy results early in the learning curve, it might help to represent the image (via VR software) fed back to the wearer through the VR stereoscopic goggles as if it had performed the action perfectly. That way the wearer sees the correct movements and can try to fall into sync with them. Since this alters proprioception, and since the brain gradually adjusts to such related illusions as image-inverting goggles, it would be unwise to overuse this mode. One might, in extreme overuse, ‘learn’ that motion does not necessarily match vision, and be disoriented in the ‘real’ world.

I can also foresee use of this aspect in conjunction with law enforcement so that (in addition to strength amplification and training) a complete record of an officer’s movements may be available as evidence, providing that records are protected from tampering afterward.

E. Physical Therapy. This would involve programming in various range-of-motion exercises as are presently done by human physical therapists. The suit would need to be programmed not to overextend motion so as to cause further injury, perhaps by attenuating pressure when a preset resistance is met. The same sort of program might be used for basic yoga training, tai chi, etc.

F. Assisted Locomotion. In this usage the suit (or a partial version) acts in the stead of a wheelchair for paralyzed or weakened individuals. It would be best for such a person to acquire a suit while the limbs are still limber and able to move freely with it, as prolonged immobility tends to lead to tightening of muscles and shortening of tendons, combated with painful difficulty by physical therapy. If an injured person uses an assistive suit as soon as possible, normal range of motion should be easier to maintain, along with avoidance of such related problems as pressure sores (decubitus), lower-extremity edema and other circulatory problems, and chronic constipation. The human body is simply not designed to remain immobile. Also, use of an assistive suit obviates most accessibility issues. Furthermore, the suit may be considered cost-effective in comparison with present mobility tools if the cost of treating these other problems is factored in; even a garden-variety wheelchair costs hundreds of dollars.

G. Fitness Exercise. In this usage, the wire muscles are calibrated by the software to offer resistance to the wearer’s movements.

One might own a virtual weight set, virtual swimming pool, etc. The suit might encourage perfect form by resisting much more strongly any motions contrary to the desired form. For example, a virtual golf swing may be completely unimpeded along the optimum path, with pressure applied to any deviation. This might feel the way a gyroscope feels when held in the hand, resisting movements that alter the axis of rotation.

One last usage which I find intriguing is for astronauts on long-term missions to wear the suit (minus gauntlet) more or less full-time while in flight. Instead of being forced to spend time on exercise bikes and other equipment, the astronauts would go about their normal routine wearing suits calibrated so as to resist movement in a way which simulates gravity. The software settings would treat an imaginary axis perpendicular to the hips as vertical and offer resistance to all movements in accord with an imaginary gravity source below. This could even be set higher than normal part of the time. Possibly the suit can combat the pooling of fluids in lower extremities as well, by rhythmic squeezing, and the suit might prove more effective in conserving bone mass and muscle tone than current exercise regimes have shown themselves to be.

In time it may be possible for the suit to be somewhat self-adjusting. Improved proprioceptive sensors might detect the body’s contours while the plates themselves (in smaller segments perhaps joined by musclewires) might expand or contract to fit, then become rigid. It may even be possible for a nonrigid version to achieve the same ends through properly controlled contraction, inflatable sleeves, or other means not yet devised.

I will add only one more observation: every use I have mentioned above is applicable to the same basic suit design. If sensors and muscle movement achieve adequate accuracy, one need only change the software to serve every single one of these roles.

3. Orbiting Fuel Farms And Water From Ceres
I started thinking about this architecture only recently, when I heard how much water there is on Ceres. You could mine Cerean water using solar power grids, extraction equipment and mass drivers. First off, it’s my understanding that Ceres, in the asteroid belt, is just close enough to the sun for solar power to work reasonably well; solar collectors could be augmented by very large mirrors made from advanced films as used in solar sails. The energy requirement might be so massive as to necessitate shipping impractical amounts of powerplant, but I remain optimistic that very large and lightweight photovoltaics can be made, considering that it appears possible to manufacture extremely light carbon nanofiber that might prove suitable for such large-scale solar power needs and perform other valuable functions in the same architecture, as mentioned in the following quote from an Eurekalert press release on work done at U.T. Dallas at this URL:

http://www.eurekalert.org/pub_releases/2005-08/uota-utd081505.php

Quote: ” The nanotube sheets can be made so thin that a square kilometer of solar sail would weigh only 30 kilograms…. The nanotube sheets combine high transparency with high electronic conductivity, are highly flexible and provide giant gravimetric surface areas, which has enabled the team to demonstrate their use as electrodes for bright organic light emitting diodes for displays and as solar cells for light harvesting. Electrodes that can be reversibly deformed over 100 percent without losing electrical conductivity are needed for high stroke artificial muscles, and the Science article describes a simple method that makes this possible for the nanotube sheets.”

Mining equipment such as drills and conveyors might be run non-electrically by solar-heated steam or by steam-generated electricity if this works better than photovoltaics or microwave power transmission. Once extracted, water or other materials would be melted, poured into containers for space flight (or perhaps they could survive vacuum as plain chunks of ice?) It would then be fired from mass drivers to distant locations where needed. Obviously, this is an adaptation of Gerard K. O’Neill’s plan for using lunar mass drivers to loft moon rock for building colonies, but adapted to a different resource from a different location, sent to a greater variety of locations.

It’s well known that our moon (Luna) has about 1/6 Earth gravity on its surface; if you weigh 150 pounds here, on the moon you weigh 25 pounds on the moon, with a diameter of 2160 miles. The surface gravity of the 580 mile wide asteroid Ceres is a bit less than 1/6th that of the moon. On Ceres you’d weigh four pounds. A gallon of water here weighs 8.31 pounds; on the moon, about twenty-two ounces; on Ceres, less than 3/12 ounces. A mass driver on Ceres need accelerate its cargo only 1431 mph to achieve escape velocity; for an electric mass driver, this is a trivial speed; and a fifty gallon drum of water would weigh just 10.8 pounds there, or a bit more than what a gallon weighs here.

Although the escape velocity of Ceres is low, even payloads accelerated to the limits of a mass driver’s capability will take far, far longer to get somewhere useful than the original O’Neill proposal, which only involved Earth-Moon distances; nevertheless, this may not be reason to abandon the idea altogether. By analogy, a Ceres mass-driver ‘pipeline’ is very long, and there’s a great time lag before anything comes out the other end; but once started, it could be steady. How long would Moon colonies take before using up all the estimated 6.6 billion tons of water in the polar regions? How long would Mars terraforming take, and might it be accelerated by Cerean methane?

It would be interesting to study exactly how much a mass driver on Ceres could fire away, with momentum to carry it along a precisely calculated course to destinations around the solar system where water is needed.  That fifty gallon drum of water weighs as much as light bowling ball, and gets its kick from a technology that moves trains. Perhaps much larger containers can be moved. As each water container passes out of sight, it joins a stream of containers which will be intercepted by catchers in space or, conceivably, one could simply pepper a chosen spot on the moon with ice chunks to be dug out and used there. Intriguingly, ‘Ceres Evolution and Current State: A Summary,’ by Tom McCord and Christophe Sotin, suggests enough methane and other volatiles and useful compounds to consider separating them and shipping them across the solar system as raw materials, or as fuel; or, as just mentioned, one might even entertain raining methane down on Mars to employ greenhouse gases in the service of terraforming.

http://www-spc.igpp.ucla.edu/dawn/newsletter/html/20030822/ceres_evolution.html

But here’s another destination for ice and volatiles: orbiting fuel farms for interplanetary missions. Using solar power to electrolyze water gives oxygen and hydrogen which may be used together or separately to power space craft. Or one might have a robotic industry on Ceres to refine deuterium and tritium from Cerean water, and send that out when there was enough to use. Hydrogen, oxygen, water, or breathable air might all be contained in ‘balloons’ of the same nanotube material mentioned before. These could be very simple structures almost reminiscent of honeypot ants in overall design functions: very large, strong bags filled with gas or liquid, with docking attachments so that they could then be strapped to a spaceship assembled in orbit, and pumping systems that be as simple as squeezing the envelope via another useful property mentioned in the U.T. Dallas article, the potential of such nanotube sheets to contract as artificial muscles: the skin could itself squeeze its contents out on command like a bladder. The filled balloon tanks might just as easily be sent elsewhere in the solar system if needed, for instance to ferry fuel to satellites or colonies needing fuel to stay in position.

The major components of this overall architecture would include:
On Ceres: 1. Solar power collectors and storage.
2. Mining, processing, and packaging equipment.
3. Mass driver, able to be aimed with extreme accuracy.

In Earth orbit: 1. Catchers.
2. Solar power collectors and storage.
3. Electrolysis and separation equipment.
4. Balloon tanks for storage and transport.
5. Orbital fuel depots/shipyards where balloon tanks are mated to otherwise completed Honeypot-Ant ships.

4. ‘Honeypot Ant’ Manned Spacecraft Design Concept
Like the Cerean water mining concept, this is so recent I haven’t had much time to consider it in detail, so simply consider it as a starting point for discussion: if we assume sources for fuel and water outside Earth’s gravity well (or a space elevator ferrying up such materials cheaply) and widespread of superstrong carbon-nanotube materials, we can consider a radically different way of building manned long-mission spacecraft.

It is not only the fuel tanks which can be made from nanotube sheets; why not let the crew quarters also be gigantic balloons, rather than tiny metal capsules? It might prove possible using very high-strength materials to dispense with metal or moon rock as radiation shielding, instead using water. Such a design might resemble O’Neill colony designs, though initially at least on a smaller scale, with the crew quarters spinning for artificial gravity. Inside the nanofiber skin would be water at least several feet deep, and living quarters within (‘above’) that. If a small meteor punctures the skin, one would plug it from inside just as one would put a drain stopper into a sink. In addition, mirrors could gather sunlight onto the exterior of the crew quarters for seaweed/algae in the water (the nanotube sheets made at U.T. Dallas being transparent, at least in some wavelengths) and thus providing a natural air-cleaning system. Sewage might be treated the same way; the key is to make the water/photosynthesis biomass large enough to handle the human crew’s needs easily.

The ship itself might begin as little more than a frame, with engines, fuel tanks, instrumentation and crew quarters added, and much of its structural strength derived from a minimal mass of carbon. This implies that very little of its total mass need necessarily be built and hauled up from Earth.