The Millennial Project 2.0

In previous articles I have discussed my suggestion that the Avalon stage of TMP should represent not just the settlement of the Earth's moon but rather a succession of surface settlements throughout the solar system. My reasoning here is that the basic pattern of development as well as the types of structures used to settle the Moon are very likely to be employed with all other surface locations in space with minor variations for the different environments.

As I've described in previous pieces, the tough economic situation of surface settlement -be it upon our own Moon, other moons, or other planets- calls for a pattern of development very different from what most visions to date have suggested. Historically, colonization on Earth has relied on a kind of dual speculation by both investors at home seeking a return on investment where they were and the settlers seeking a return on investment where they were going. Though the specific financial or social mechanisms varied (and not all settlers in history were willing participants or saw any benefit from their move...), the high cost and risk of intercontinental travel meant that settlers typically relied on investors to pay for their transport and the initial creation of their settlements and had to pay the debt for that sponsorship with export goods of sufficient value to cover that initial cost and provide a reasonable return on investment to those sponsors. However, surface locations in space offer little to no prospect of a return on investment for a sponsor because transit costs are so high -much higher than for asteroids- that one is left with no likely resources on these bodies valuable enough to send anywhere at a profit. Thus only the prospective settler himself has any hope of seeing a return on investment in the place he's going to through the community he creates and, with little hope of outside investor interest, must somehow muster the means alone to pay for his own transit and safe setup in an environment MUCH less benign than any wilderness on Earth. Moons, by virtue of a much lower level of gravity, do offer some better prospects of potential export revenue because export transport costs are so much less than planets. Our own Moon, by virtue of proximity to the first permanent orbital settlement of Asgard, has some small potential to be competitive with asteroids as a source of raw materials and possible bulk beamed energy but the infrastructures required to achieve cost-efficient for this would still tend to be great and take much time to develop. And there is the trade-off of a much smaller spectrum materials likely on moons compared to asteroids and planets.

The common vision of lunar and planetary settlement has been based on the notion of manned science and exploration outposts transitioning into permanent settlements. But this concept is ultimately impractical in light of this difficult economic situation without vast and protracted public (government) investment despite no hope of return on that investment -a fanciful notion at best. Governments have historically sponsored wilderness exploration primarily for strategic military reasons. They have sometimes done it for assessment of resources and -rarely- for sake of science but the agenda is ultimately strategic. They have rarely had interest in actually financing colonization unless it was part of a program of military conquest or the wealth-building of an individual dynastic monarchy. This explains why the agendas of government space agencies to date seem so fuzzy-minded, the actual strategic military advantage of manned space activities difficult to demonstrate and with stated yet nebulous objectives of science and 'exploration' (for no clear strategic purpose) ultimately being rationalized on geopolitical prestige. There is no conception among the members of existing space agencies that their most practical role may be to pave the way for private and commercial development with the intent of colonization. Indeed, most treat independent space programs as annoying 'amateur' competition rather than the community they ultimately should be serving. This leaves the prospective space settler to employ this strategy with little or no outside help, which probably makes it virtually impossible even for the exceptionally wealthy.

What makes this strategy so difficult? Time. The overhead and personal risk of transit across space today is comparable to that for New World settlers of the past. This is a trip the prospective settler can afford maybe once in a lifetime. This means the settler has one shot in a lifetime to get settlement right. If he fails and must return to Earth, that's it. No more chances. And the stuff he leaves behind may offer only limited benefit to future attempts by others -maybe serving no other benefit but some refined material to recycle. So one faces a 'failure is no option' situation. But settlement in Earth history was, logistically, a piece of cake compared to any location in space because it was such a benign and forgiving environment with such an abundance of easily usable resources. The New World settler didn't need to bring everything his survival required to make a go of it because the environment provided at-hand all the resources for basic survival for an indefinite period. Given adequate survival skills, a colonists could 'live off the land' forever wherever the ships dropped him off. Indeed, the incompetent planning of many attempts at New World settlement -especially by those driven more by religious motives than economic ones- was striking and yet many still survived thanks to this benign environment and -at times- the tragically rewarded altruism of natives. This ability to potentially survive indefinitely on local resources gave New World settlers a great deal of time to work at the task of establishing a sustainable community and the infrastructure of farms and industries needed to sustain it. They were not time-limited by a fixed stockpile of supplies brought with them and this was critical as the process of colonization required a lot of experimentation. Settlers had to revise the technologies they brought with them to suit the unique conditions of this new environment. They had to learn by trial-and-error what Old World crops and livestock they could and couldn't cultivate and figure out how to use new ones never seen before. They had to search over vast areas for the dispersed sources of the full spectrum of resources they needed and figure out how to extract, transport, and process them. And talk about your easy living, most of the iron used by early New World colonies was 'bog iron'; big lumps of metal formed by the action of lithophoric bacteria over countless millennia ready to be dredged out of swamps and riverbeds with sticks and ropes and sent straight to the blacksmith! And when settlers got sophisticated enough to mine for iron, they just happened to find it right next to where they settled -large deposits all along the Eastern seaboard -along with wood, coal, and even oil to power its processing. Imagine how different the history of the US would have been if all of North America's iron was west of the Mississippi -largely undiscovered until the the 19th century. But even with all these boons and advantages it took decades to get a New World settlement into a state that was truly self-sufficient and generations to establish the industrial infrastructure that could duplicate the Old World standard of living independent of its import goods.

In space one confronts an environment so harsh that one must literally package a chunk of the Earth and take it with you to survive. No matter how skillful you are in the design and engineering of this bottled bit of the Earth, it is not indefinitely sustainable. It's running on a fixed volume of supplies sent with it or on a limited supply schedule according to how much money and resources the prospective settler(s) saved up for the mission. Even if you could craft the ultimate portable Closed Environment Life Support System recycling perpetually all the staples of life support, it's still running on sophisticated manufactured hardware that will wear out and need to be repaired and replaced, requiring sophisticated industry to make. So for the human inhabited space settlement one faces a very short time limit on establishing a very high degree of industrial self-sufficiency based on how much stuff you can afford to send out into space with you.

Meanwhile, one must deal with local resources that are in such a raw form that it takes a very sophisticated technology to seek them out and a very sophisticated industrial capability to process them for use. On Earth life pre-processed the elemental resources of the planet, refining them into more easily usable forms like trees, plants and animals that one could gather, farm, hunt and herd for food. In space one must rely on machines (and in some cases hybrids of machines hosting life forms as part of their mechanisms) to do all this. Everything right down to the air one breathes must be produced industrially using technology that itself requires very sophisticated industry to produce and maintain.

All this presents the manned outpost strategy with a virtually impossible challenge; to locate, extract and establish a very high-tech processing infrastructure for a very large spectrum of resources in very raw forms distributed over a large surface area using a mere handful of people over a span of time perhaps no longer than a few years even for the most well-heeled of settlers! Barring the advent of an advanced nanotechnology, this seems like an untenable proposition to me without continuous outside help. To achieve this even in as much time as a decade on a supposedly resource-rich planet like Mars would be a miraculous feat and no small group of settlers are likely to ever be able to afford the supply stockpile for that. Even if government were willing to sponsor colonization, for whatever abstract rationalization space advocacy can conjure-up, the cost of continual supply of a settlement for as long as it is really likely to take to become self-sufficient could strain the budget of a super-power nation. Marshal Savage envisioned the Foundation as taking the place of government as a source of sponsorship for colonization and it is quite likely that, with the beginnings of an orbital civilization established in the Asgard phase, the costs of transit in space could be greatly reduced and the ability to pre-supply a colonization mission with space-sourced materials and in-space manufactured goods becomes practical, thus making the sponsorship of surface settlement an easier prospect. But even an organization and civilization like that would be very hard-pressed to muster protracted support for something offering so little return on so much investment over so much time. One needs a way to radically reduce the up-front cost for settlement activity and overcome those critical time limits for establishing sustainability. The obvious solution is to somehow very cheaply pre-establish a self-sufficient infrastructure for survival BEFORE settlers commit their lives and money to going there and the obvious way to do that is by robotics.

Thus I have arrived at the notion of settlement beginning not with manned outposts but with telerobotic outposts; facilities staffed entirely by relatively simple robots operated more-or-less by remote control from elsewhere. Such suggestions tend to immediately inspire a debate over whether robots can really do 'as good' a job as humans can. The obvious answer is, no, they can't. But it doesn't matter. They don't have to be 'as good' or 'better' than humans. They just have to be 'good enough' to get the essential jobs done, helped by carefully engineering those tasks to suit their limitations. The logistical advantage of robotics in this situation is not in performance but in transport and support. The extremely high overhead of human life support in space is the primary obstacle here. This is what imposes high up-front costs and that critical time limit on what you have to get done before the supplies run out.

On Earth automation has tended to be a very expensive solution in comparison to human labor. But that's only because, thanks to Earth's benign environment, employers don't have to pay for every aspect of their workers' life support down to the very air they breath. And it also 'helps' that on Earth there's a society that has systematically engineered the mass socioeconomic exploitation of large portions of its population. The Total Automation revolution predicted since the middle of the 20th century didn't fail to materialize because of any sort of human superiority. It was stopped by Globalization and the essential economy of human exploitation. It was far cheaper to exploit social and racial underclasses than develop robots. In space robots -as expensive as they individually are- have the definite economic edge because the social class and race of people in space doesn't give you a break on the cost of their life support. You can send robots wherever they need to go and supply their needs for energy and replacement parts for orders of magnitude less money than it costs to transport and support human beings. But most important of all, it doesn't matter how long it takes to get anything done with them. Whether it takes a year or fifty to build a self-sustaining settlement safe and ready for human habitation simply doesn't matter. Robots do have 'life support' needs in the sense that one must supply them with energy and repair parts but these needs can be met much more cheaply and when necessary parts of the settlement and its robots can be put into 'suspended animation' by simply being turned off as a way to reserve duty life while waiting for re-supply. Thus the telerobotic settlement mission never runs out of time no matter how long things take to accomplish or how much trial-and-error experimentation is needed to figure out how to get things working. No one's lives are at stake, no titanic sums of investment and resources are being wagered, and the tele-operation of the settlement can readily change hands if necessary if one or another group runs out of money. In theory, a telerobotic outpost could be operated for many decades for the cost a manned outpost would require to deploy and operate for a single year.

I liken this strategy to colonization of the solar system as a hobby project. It's sort of like caring after a model train layout that assembles itself and eventually develops into something you can actually go and live in. SimCity with real hardware. At the potentially radical cost savings involved and freed of the limitations of time that's something one might accomplish with surplus funds and resources without concern for ROI. Something a small handful of clever and well-heeled people could actually pull off on their own -and possibly make some near-term on-Earth profit by since the robotics technology they must develop -unlike manned spacecraft technology- definitely has an established and strong market value right here on Earth. I think that the over-emphasis on manned space flight in space advocacy has resulted in a missed opportunity because, while manned space flight is extremely difficult to achieve for the lone entrepreneur or space advocacy group, unmanned space flight is quite achievable -especially when you consider how many more options you have to explore in terms of scale, propulsion, and design when you don't have to worry about taking care of fragile human bodies. If you can achieve unmanned interplanetary space flight then you've got the basic tool to open up the solar system to colonization -if you're willing to be a little patient and pragmatic about getting people out there. And one very powerful advantage of this strategy is that one can easily prototype and demonstrate a telerobotic settlement right here on Earth using any number of remote test sights with aircraft replacing rockets as a delivery system and 'planted' simulated resource deposits mimicking anticipated resources in space, working out all the bugs in a fully functional proof-of-concept. Any university or space advocacy group could pull off such a project.

For a community such as Asgard's, employing this tactic for settlement is virtually a no-brainer. It would be a direct extension of the strategies that community would already have employed for the exploitation of asteroids using derivatives of MUOL technology. In fact, the settlement of the Earth's moon is of likely commercial interest to orbital industry as a nearby source of materials if one can effectively exploit its indigenous resources to establish a volume transit system at minimal cost. Thus one can anticipate at least the exploration of lunar and planetary settlement as a normal part of the interplanetary prospecting and mining industry Asgard would cultivate. The Moon is one place where there is some potential for investor ROI on a surface settlement -but only if that investor is seeking profit on a 'parent' market already in space, not one on Earth.

This is not a new idea. During the early planning stages for the Apollo program a number of scientists argued in favor of a robotic settlement of the Moon rather than a manned mission, suggesting that for the same cost of a few flag-planting photo-ops one could open up the whole solar system to development. The difference then was that scientists were anticipating artificial intelligence as the basis of operation and expected the space program to heavily invest in that area of research -with the resulting Total Automation revolution that would have ushered in being the ultimate in world-changing technological spin-offs for the space program. Considering what we know now about the difficulty of achieving AI, this was probably over-reaching. Today we must assume the use of teleoperation, but with AI still a possibility down the road and offering the prospect of extending one's reach farther and making development speed much faster than radio telecom latency would otherwise allow.

Let us now consider the design and operation of this robotic settlement. How do we build this ultimate model train layout? I envision this telerobotic phase of development being itself composed of several phases which would tend to be similar in nature in every surface location where this strategy was employed, though for some their very extreme environments will call for much tougher hardware and would likely be exploited much later in history. Let's look at each of these phases and the types of equipment they would employ. For the sake of simplicity, we can assume this example talks about the Earth's Moon or Mars, as these represent the most likely first targets of settlement.

Assay Phase[]

In this phase the body planned for settlement -be it a moon or planet- is assayed by orbital spacecraft for its geographical and geophysical prospects of settlement. Much of the preliminary work involved with this has already been done for much of the solar system by the scientific research missions of the existing space programs and a great deal of data is readily available today, especially for the two bodies of highest interest for manned exploration missions; the Moon and Mars. But these agencies priorities have been very different from those of the serious prospective settler and so it may be necessary to deploy new spacecraft designed more specifically for this task and cable to perform more specific analysis of surface and near-surface geology. Of course, my favored strategy for this would be to use the technology of the MUOL to construct on-orbit very sophisticated orbital assay platforms which can be equipped with a large diversity of remote sensor systems as well as surface and atmospheric probe devices. Surface and atmospheric probes would tend to be fairly simple disposable devices in this stage, based on concepts like sensor webs, impactors, and lighter-than-air structures. The specific choice of technology would depend on the environmental conditions of the body being examined.

The essential mission goal in this phase is to narrow down choices for logistically advantageous initial outpost locations but the data will still -even with these more specialized platforms- be limited. Detailed assay will ultimately rely on fairly large robotic surface explorers with long range and long duty life and more sophisticated tools such as core drilling rigs and on-board chemistry labs. These kinds of systems will be deployed later on.

Perhaps the most important long term role for these orbital platforms would be telecommunications and GPS beacons for the systems on the surface, insuring continuous telecommunications and precision position tracking no matter the relative position of the settlement on the surface. To this end the assay platforms would be equipped with both surface relay link systems and inter-orbit link systems so that they can maintain continuous relay chains from the settlement points to Earth or -in Asgard's case- Earth orbit. Due to the nature of their mission, some may be sent to polar orbits or may deploy separate smaller satellites to polar orbits in order to provide the most comprehensive imaging coverage of the body. Later, telecom platforms specific to that purpose would be deployed in solar synchronous and surface synchronous orbits to facilitate higher bandwidth fixed latency links. If based on MUOL technology as envisioned, these platforms would also have the option to be equipped with on-board robots for their own service and expansion, affording the possibility that the platforms could be expanded into EvoHab based stations which may serve as way-stations for the later transit of goods and people using reusable surface shuttles and interplanetary craft.

Telecommunications is extremely critical to this strategy and so a very flexible highly redundant digital wireless network is going to serve as the foundation of the settlement. One key problem faced in this is that many of the systems the settlement would use have very high bandwidth demands for normal operation yet at the same time may call for a high degree of portability. It's difficult for portable systems to afford such high bandwidth with a link direct to Earth or even to orbit because of the scale of telecom hardware they would need to carry. Thus the settlement's network system must employ a diversity of distributed stationary and mobile telecom systems suited to different tasks and offering high bandwidth relay to systems that can't physically support their own direct long range high bandwidth systems. But for redundancy many systems would still include low bandwidth lightweight communications systems for direct links in the event of emergency. To cope with such a mix of systems I envision the use of an IP based primary communications architecture so that the specifics of individual telecom hardware can be dealt with at a lower level transparent to primary command and status communications. This would increase redundancy in the event of spotty communications, allowing for automatic and transparent transition to different systems as conditions warrant.

Beachhead Phase:

After the likely settlement sites have been sorted out and a specific target site chosen, the settlement program would begin its initial surface systems deployment. There are many criteria for choosing a likely settlement site but perhaps the key ones will be good solar insolation, proximity to a large out-cropping of soft but resilient rock strata suitable for structural excavation as well as nearby locations suited to soft landing vehicles, a largely debris-free drop-zone field, and locations of likely key resource deposits.

The first systems delivered to the surface would be a set of small self-contained beacon and telecom relay units delivered by 'rough' landing systems (which we'll discuss in more detail later) in a rough circular pattern around the settlement site. These units would be similar in design to the ill-fated Beagle Mars probe, consisting of a simple self-contained self-deploying communications systems package. The primary purpose of these units is to serve as a navigational aid for the next series of soft landing vehicles which need to be guided with some precision to the settlement site.

The next wave of systems would be delivered by at least three soft landing vehicles called Beachhead Landers which would arrive more-or-less simultaneously. These would be the most sophisticated surface landers employed in the telerobotic phase of development and they would employ soft landing technology not so much for the protection of their payloads but rather for the precision in landing since they need to take up a very specific position within the midst of the settlement site's features and be placed quite close to each other. This, of course, is also why they need the aid of those initial beacon systems which, though individually deployed in a random pattern, provide ranging signals to establish a precision location grid around them. The Beachhead Lander's role is -as its name implies- to establish an early support infrastructure for a subsequent steady inflow of hardware to the settlement site and would be designed to deploy 4 critical systems;

UpLink Communications Package: A package of telecom systems designed to provide a broadband digital link to Earth relayed to the area immediately around the settlement site by shorter range wireless transponders. This system works in combination with the initial beacon landers, increasing the reach of the local wireless network and providing back-up links to Earth, though many of them may be out of range for this initially due to the manner of their deployment.

Radar Tracking System: A radar ranging and tracking system use to monitor surface robots and to track incoming vehicles from space. It is possible that this system could be replaced by the use of triangulated beacon ranging in conjunction with the initial beacon landers -creating a kind of crude GPS system. This approach would be more economical than deploying a sophisticated radar platform but would be blind to any object that itself lacked a working navigation beacon transponder and/or transmitter.

Assembler Robots: A series of robot arms deployed from the center of the lander which are used for the crude assembly of other systems from largely self-contained components. The chassis of the lander would be designed to provide a workstation for this activity with mounts for modular tool heads, multiple cameras, and lighting systems. The Beachhead Lander Assembler is limited to 'crude' assembly not by performance but by the fact that the lander structure would provide limited shelter from dusts, debris, and radiation making it impractical to work with small and fragile components and exposed electronics.

MultiRover Robots: A set of multi-function rover robots, with each lander deploying one to a few of them. The MultiRover has several duties; it would clear the area around the Beachhead Lander site of debris. It would deploy stationary equipment about the settlement site. It would recover payloads deposited by 'rough' landers in the settlement drop zone. It would assist the Assemblers in assembly and help recover robots that have failed for repair. The basic features of the robot would be a simple flat chassis with a plow for clearing in the front, one or two robot arms, a cargo carrying bed, and tow attachment points. The MultiRover would be much more self contained than later robots -at a sacrifice in mechanical power- and would feature its own deployable solar power charging systems and back-up low-bandwidth communications links to Earth. The MultiRover -like all robots ultimately deployed by the settlement- would rely on a modular component design where a catalog of interchangeable components with standardized quick-connect/disconnect structural, power, and communications interfaces is used to create all the different specialized robots the settlement deploys. The development of this modular component architecture would be the key R&D challenge of the settlement project and would involve continual refinement indefinitely. Robots based on this technology would tend to be a bit bulkier than the robots typically deployed in space exploration but there's a critical reason for this approach; it allows most systems deployed on the settlement to be shipped as packed components assembled on-site. This allows for the use of 'rough' lander transport that costs only a fraction of the cost of large complex 'soft' lander systems.

In addition to these systems, the Beachhead Lander would also carry some cargo in the form of a small amount of replacement parts, deployable solar power systems, and a number of self-contained communications transponders. These transponders would basically do the same job as the initial beacon landers but would be strategically deployed by the MultiRovers to establish a communications web for the settlement and -most critical at this stage- into the drop zone so that the MultiRovers can cover as much of its area as possible without risk of communications loss.

Upon arrival of the Beachhead Landers the deployed MultiRovers would clear their landing area of debris, deploy solar power systems, and stake-out with their transponders the key geographical features of the settlement site and the easiest paths between them. These would be the key features of the settlement at this stage;

The Beachhead Landing Zone: This is the area immediately around the Beachhead Landers serving as the initial center of the telerobotic outpost.

The Drop Zone: This is a large area of many square miles in a roughly oval shape aligned latitudinally. Chosen in a region of flat terrain with little debris, the Drop Zone serves as a target area for 'rough' lander vehicles. Its large area accommodates the lack of precision associated with this mode of transport. Initially relying on natural flat clear terrain, it is likely to still present some hazards in the form of natural debris and variable surface. Because of the large area, this would not be easily clearable using the small MultiRovers but eventually larger robots would be employed to remove at least the most serious hazards.

The Soft Landing Zone: Located near the Beachhead Zone, this landing area, on terrain similar to that of the Drop Zone but much smaller in area, would be set aside to serve 'soft' lander vehicles. This area would not see much use early on, except in the case of emergency replacement of Beachhead Landers, so clearing of debris would be of lower priority.

The Outpost Zone: Either incorporating the Beachhead Zone or located very near it, the Outpost Zone would be the next center of the telerobotic outpost. As precise as their landings may be, the Beachhead Landers will not likely wind up in exactly the ideal spot for later structures. Thus a true center for the outpost would be established nearby.

The Energy Field: Located optimally for solar insulation or wind exposure as well as to serve both the Outpost Zone and the later Habitat Excavation Zone, the Power Field is basically an area set aside for the mass deployment of solar and/or wind power systems as well as power storage systems.

The Habitat Excavation Zone: This is the site of the eventual human habitat and will ultimately become the third center for the telerobotic outpost as excavations are ultimately used to house the telerobotic systems while being prepared for human habitation.

Resource Zones: This would be mining zones chosen where key resource deposits are already known to exist. It is unlikely, though, that the exact location for many such zones would be pre-identified. The detailed assay work will require a later set of robots. So at this stage the MultiRovers would simply mark transit routes to the general regions.

Outpost Phase:

With a support system for the delivery of cargo established, the settlement would begin to receive and process incoming hardware and begin assembling its stationary facilities in the Outpost Zone and its fleets of larger and more specialized robots. As was noted previously, the primary means of cargo delivery would be 'rough' landers. What exactly are these vehicles? Essentially, they are cargo pallets or containers equipped with small radio tracking beacons designed to be dropped without great precision into a large target drop zone and which use air bags to cushion a relatively rough landing. They would use one of two types of delivery systems; in atmosphere they would mounted within a disposable ablative aeroshell and dropped to the surface by parachutes -a method used by many Mars surface lander systems to date. Where there is no atmosphere a 'rocket chute' system would be used where a disposable thruster pack on a tether is used as a braking system much as a parachute would be used, disconnecting and shutting down as its payload touches the surface. The rocket chute system would be more costly than parachutes but has the advantages of much greater control and indifference to weather conditions leading to a softer landing and a smaller necessary drop zone. The thruster packs would also be more easily recyclable and safer and easier for robots to recover than parachutes, which pose a risk of entanglement. For these reasons the rocket chute may ultimately become the preferred rough lander system even in atmospheric conditions. The pallets and containers used with these systems would be standardized in form to allow for mass production and thus even greater economy.

Initially these pallets and containers will need no pressurization but later on some special pressurized and/or insulated cargo containers may be necessary to store perishable items such as small living organisms used -most likely- in early hydroponics and mariculture systems. The are likely to be sub-containerized, using a kind of pressurized cartridge with modular pressure fittings. A large variety of launch systems may be used with these rough landers. For cargo sent from orbital facilities it is likely that MUOL technology would be used to construct simple disposable mass cargo vehicles hosting arrays of drop capsules on a primary truss structure. Sent from the Earth's surface, the use of individual launchers for each cargo package is likely with each cargo capsule being individually equipped a disposable transit booster. Later on, when there is a sustained human presence, reusable spacecraft will replace these disposable vehicles but at this point this very simple disposable hardware is likely to be the most cost-effective.

Using this delivery technology the outpost's hardware would be delivered in tightly packed modular component form, ready to be assembled into robots and stationary systems. The initial MultiRovers would tend to be too small to pick up and carry these containers and pallets whole so initial cargo payloads would be broken-down on site and their parts individually shuttled to the Beachhead Landers where they would be examined and tested for damage, placed in storage within the landers, and then assembled into other systems. The empty containers and pallets would be left in place until later larger robots can collect them for later recycling.

At this stage the Outpost Zone would be setup to serve as the next center of operations and would develop the following structures;

Primary Telecom Downstation: This is a more robust version of the telecom systems used on the Beachhead Landers, designed for maximum bandwidth, longer surface relay range, use of high bandwidth point-to-point line-of-sight links, use of cabled fiber-optic links, generous data storage capacity, local data processing capability, and perpetual maintenance and upgrade. This is the real heart of the telerobotic outpost and functions both as a communications center supporting a broad spectrum of discrete communications technologies all under the IP umbrella as well as a data processing center storing data logged by other systems and providing 'sequencer' level program support for coordinated operation of the discrete systems. The telerobotic outpost would have an operations architecture exactly that that used by the MUOL; a generally homogeneous network architecture hosting systems individually controlled by web controllers and higher level sequencer programs. The key difference is that it is relying primarily on a wireless network with cabled networking having a supplemental role. While many systems would not need more than a virtual control panel interface, robots are likely to employ a streamed bytecode command and feedback protocol where distant operators compose command streams using telepresence programs that model the environment around the robot and test actions for failure situations before composing and streaming them to the robots. But they would still rely on web controller access to discrete subsystems and status monitoring. The use of AI might also ultimately be supported by the computer systems associated with the Downstation, supplementing and ultimately replacing local sequencer programs and remote command streaming.

In addition to the Primary Telecom Downstation, there would be a large number of telecom transponders deployed. These would be much like the units carried as cargo by the Beachhead Landers and are used to extend the reach of the wireless communications web of the settlement. Each would consist of a solar power system and short and long range high bandwidth communications transponders and low bandwidth Earth uplink transponders. Some may feature point-to-point relay linking between them and the Downstation, requiring a system of auto-stumbling antennas. Small LED lights would also be used to provide visual aids for the control of robots and they may even feature their own video cameras and atmospheric sensors. The primary routes between locations on the surface would initially tend to be mapped out by the placement of these units, the communications web being the most fundamental infrastructure of the settlement.

Power Station: Actually part of the systems deployed in the Energy Field, it consists of a cluster of power storage modules and distribution modules. Power storage may actually be done in various ways. It may be 'direct' in the form of simple electrolytic chemical batteries, self-contained redox systems, gyroscopic storage loops, superconducting storage coils, etc. Or it may be 'hybrid' systems where solar or wind energy is used to convert a resource like atmosphere or water into a stored 'fuel' which then used later to produce power by fuel cell or turbogenerator and can optionally be transported to other locations. There are advantages and trade-offs to both approaches but the end results in terms of scale of deployed hardware will be roughly similar.

The power distribution modules are the basic components of a power utility 'grid' which would employ a simple branched architecture with 'bridged' links to multiple power stations as the outpost grows. Like the MUOL this may be an optical transmission system in order to isolate the grid and the other systems of the outpost from each other as a means of preventing failures from spreading damage along the power grid as well as for EMF isolation, preventing systems from generating signal noise and propagating along power lines. This may be particularly important in locations like Mars where dust storms have the potential to generate much EMF and potentially disastrous static discharges. However, some of the systems used by the telerobotic outpost will be very high energy systems and may exceed the practical limits of optical power communication by fiber cable. The only alternative to conventional power cables then is the use of elaborate point-to-point reflective pipes and enclosed free-space light guides. Most cabling used to link up the systems in the outpost would be designed to be simply laid on the ground and would be ribbon-shaped so as to prevent twisting and to allow for robots to travel over them with ease and without causing them damage. They would, of course, use a standardized modular connector that allows for easy robot handling.

In addition to linking up the stationary structures of the outpost, there would also be a number of Power Terminals setup in strategic locations. Actually a special form of distribution module, these would feature an interface through which the mobile robots of the outpost can be recharged. Some R&D would be needed to determine the most reliable design of this interface and it may take the form of induction loop pads the robot merely rolls over, very short range point-to-point target pads for beamed energy, or some form plug and socket arrangement where the terminal or the robot uses some kind of self-articulated proboscis to link-up. Some of these power terminals would be designed to be moved about as necessary. A few types of very high-energy robots -such as mining excavator systems- would actually need to be continuously cabled to operate and thus would travel point-to-point between them. This is a common approach for large mining systems today, most of which are, in fact, electric powered.

There would also be a series of small remote power terminals setup in strategic locations in the landscape, particularly around the Drop Zone. These remote power terminals would consist of small self-contained power systems with deployable solar panels or wind turbines and their purpose is to provide emergency power supply to remote robots and allow for the long distance transit of robots that lack their own solar power generation capability. A larger variation of the transponder units used for the wireless communication network with added power terminal features might be developed for this purpose.

Service Shed: This is a light prefab shelter structure which houses a special array of Assembler robots consisting of smaller versions of MUOL-like 'inchworm' robot arms which traverse a floor and overhead array of anchor sockets which are also used to mount tool pallets, lighting systems, camera systems, communications patchboards, wireless network transponders, and specialized test and repair equipment modules. The Service Shed uses a cabled link to the Downstation for maximum bandwidth and another cabled link to the Power Station. Most likely employing a simple arch or box shelter form, the Service Shed would feature portals on either end with closable doors and dust sweeping stations which sweep everything sent into it free of dust to keep a clean interior environment. This facility is used to perform more complex assembly and repair tasks than the cruder exterior assemblers and because it sometimes works with exposed electronics it must employ a low-dust interior environment, hence the need for this enclosed shelter. As the outpost grows this facility may grow in length. size, or number to accommodate more work volume, large work items, its own repair component fabrication, and the storage of components that are unsafe to store without shelter from dust.

Robot and Storage Sheds: A series of simple free-standing arch structures without enclosed ends made, perhaps, using corrugated alloy panel or geodesic frame and panel systems. For places with an atmosphere and significant gravity they may take the form of an actual curved arch. On the Moon and similar airless locations they may be simpler low rectangular structures with an non-curved arch frame sections. They would have an option to be buried by loose regolith for meteorid shelter. They would feature internal wireless network transponders, solar charged lamps, and power terminals. Robots don't need much in the way of shelter but on moons and planets like Mars dust is a serious issue. In areas without atmosphere, meteroids are also an issue, though there would not be much one could do about larger impacts at this stage. Thus to extend the life of robots and reduce their rate of servicing it would be useful to provide them with simple shelter, though this may be used mostly for those robots that aren't actually being deployed for a specific task. (with latency in communications so high in some locations, having robots scurry for cover in the event of a sudden change of conditions isn't possible) These sheds would also be used to provide storage for the bulk of the cargo collected by the outpost.

Geology Lab: This is a facility similar to the laboratory modules of the MUOL but consisting of modules designed for on-ground stationary mounting, likely using screw piling foundations. It uses a series of analysis or 'lab modules' which plug together along inter-module access ports and form an internal transport conveyor chain for sample materials supplied from the outside. Like the Service Shed, it relies on a direct cabled connection to the Downstation and Power Station. Its function is to perform very in-depth mineralogical analysis on sample materials brought to it by other rover robots. It can also perform analysis related to the search for signs of life, but geoassay is its primary function.

Factory Zone: Also similar in nature to the MUOL's early factory facilities, the Factory Zone uses a series of linking process modules to manufacture items from the materials about the settlement. Like the Geology Lab, it too would rely on direct connection to the Downstation and Power Station. It would also be open to the possibility of leased space facilities much as with the MUOL. It would later evolve to support structures akin to the Service Shed, which are analogous to the built-up MUOF facilities of Asgard.

In most cases processing of materials from the surrounding environment would be based on initial refinement done at or near the source of extraction, these facilities equipped with independent power and communications, then the product of that transported to the settlement for further use. Initially, though, this Factory Zone would be limited to working with the materials collected in the process of clearing the settlement site of debris and, later, the recycling of the disposable portions of rough lander vehicles, cargo containers/pallets, and broken components. Initial products, though, are likely to be very simple and experimental; processed regolith building blocks or regolete (regolith based concrete) materials, compressed gasses, lower-precision monolithic components made by fabrication derived from rapid prototyping techniques. Ultimately, this external Factory Zone has the potential to produce goods at the same level of sophistication as the MUOF. But operating costs in the external surface environment will be much higher due to the cost of importing a built-up structure, higher failure rates for electronics, and need for radiation shielding for the production of some goods -particularly delicate electronics. So more sophisticated production is more likely with the later underground settlement.

The primary goal of industry on the telerobotic settlement would be sustainability defined in terms of the ability to reproduce most hardware used by the settlement using and recycling indigenous materials. Essentially, the settlement is like a Turing Machine. But sophisticated products require a broad materials spectrum and, until the advent of nanotechnology, the settlement would not immediately have the full spectrum of manufacturing technology for all its needs. This will likely take some time to develop. This portion of the robotic settlement is likely to undergo a great deal of reconfiguration and expansion over time as the facility learns to produce a progressively larger diversity of products from a larger diversity of materials. It will need to incorporate storage facilities as well as transit terminals for a progressively more continuous stream of material delivered by other free-roaming robots and possibly Ultra-Light Monorail. As the Factory Zone expands it is likely to employ its own specialized enclosed conveyor system for short distance travel. This would be akin to a Personal Packet Transit System but enclosed and designed to link the various clusters of systems in the Factory Zone together so they can trade products while also allowing external service robots freedom to move about and service or replace factory modules. This all suggests a tree branched network for the layout of the Factory Zone.

Excavation Zone: This is the ultimate future of the settlement, the future center of the telerobotic activity and the location of the eventual initial human occupied habitats. Thus the work in this site represents the most important work of the telerobotic stage of development. The excavation site would be chosen to provide above or very-near-surface access to rock strata that is soft enough to be easily excavated but strong enough to serve as durable permanent structure. Ideally, it would feature existing caves of generous size or features like lava tubes that can be easily adapted to shelter purposes. Initially, a lot of assay work would be needed to analyze the geological and topological nature of the chosen site with the planned design of the settlement and its pattern of excavation and construction developed in response to that.

Built structures like the Service Shed will tend to be very costly when their components have to be imported from elsewhere while the need for enclosed structures would steadily increase as the sophistication of systems used and the components to service them increases. It is much more cost-effective to exploit indigenous materials for such structure but the manufacturing capability for it will take much time to develop. Thus it makes economic sense to make the most of structure nature herself provides in the form of natural rock. Excavation is a relatively simple process that is relatively easy to design robots to perform and can produce structures of virtually any scale with superior performance in just about every functional characteristic. And in the process of excavation one produces waste material that may have many other potential industrial uses depending on the rock strata composition. The only practical limitation of excavated structure is the inability to put them exactly where one wants them since nature determines the possible building locations according to where the usable rock is. So in the very long term built structures will always be necessary. But initially -and for much of the history of colonization- excavated structure will likely be the primary form of structure out of sheer economy.

Initial work would consist of clearing akin to that done for the Outpost Zone followed by excavation of spaces suited to the uses of most of the features of the Outpost Zone. These would tend to favor a simple arrangement of gridded vaults following the topology of the rock strata akin to that of contemporary mines, creating generic spaces well suited to industrial needs. This space would be outfit with a network of utility modules providing wireless telecom network coverage, cabled network connection, power connections, and power terminals. Human habitats will need more elaborate structures and we will go into that soon. But initially these simple spaces will provide all the structural needs of the telerobotic settlement as well as initial human settlement if need be.

The Outpost Phase would also see the addition of the following new forms of robots to service expanding activities;

Recovery Rover: This is a much larger version of the MultiRover that is designed for the purpose of payload and vehicle waste recovery in the Drop Zone. It would feature its own crane-like robot arms similar to the Assembler Robots, a small plow or digging tool for the excavation of objects partially buried by impact, and a large flat bed which payloads are carried on. Designed to load and carry payload containers or pallets whole, it may also have to deal with disassembly of large payloads, could be designed with a kind of articulated fork-lift manipulator, and may be deployed in pairs for recovery operations. Since it must operate a good distance away from the main parts of the settlement as well as remote power terminals, it would also feature high capacity power storage and possibly a deployable solar power system. This backup solar power system would not be suited to primary use. Only for getting the robot charged enough to get back to the nearest power terminal should it get lost.

Dozer Rover: A key workhorse robot used for most surface debris clearing, especially in the Excavation Zone and the Drop Zone. It would be of the same scale as the Recovery Robot and possibly feature the same chassis (most robots would) but feature much larger bucket excavator and plow tools and have a walled tip-bucket carrier bed much like a dump truck for carrying material.

Mobile Assembler: This is essentially the same kind of system as the Assembler Robot used with the Beachhead Lander only it is mounted on a mobile chassis with deployable stabilizers. It may be self-mobile or may use one of the other rover robots to tow it place to place and it is a cabled power system, meaning that it relies on a continuous connection to power terminals. It's purpose is to perform field assembly work and repair, building things like the shelter structures for the outpost and servicing robots that are too large to be dragged to the Service Shed when they break down. Some may be continuously stationed at storage facilities for the break-down and management of cargo from payload containers/pallets.

Mobile Transponder: This is a rover on the scale of the MultiRover which is designed for the purpose of providing a mobile equivalent of the stationary wireless telecom transponder units. It would be a self-powered unit. Though the stationary transponder units would be relatively easy to move around, this unit would be necessary in situations where unexpected blind spots in the telecom web occur due to terrain conditions (such as would be the case with caves, craters, or canyons), where transponders fail and cause other robots to become disabled, or where robots have strayed out of the telecom web and need to be searched for by temporarily extending the reach network web to wake them up.

Mobile Power Station: Like the Mobile Transponder, the Mobile Power Station is a contingency system intended to handle situations where robots have become disabled beyond the reach of power terminals and need recharging, where robots without self-charging capability need to be transported some distance, where they must be temporarily deployed in some large number, or where it is necessary to temporarily deploy high power robots requiring cabled power. It is essentially a miniature solar power plant with a large self-deploying photovoltaic array and large capacity power storage. In locations like Mars it might also be based on a reformer plant storing methane fuel in a hybrid power system.

Scout Rover: This robot is charged with the task of concerted long range exploration and preliminary geological assay. It must perform a host of survey, assay, and science missions at the same time for extended periods over vast areas. Consequently, it would be one of the most sophisticated of robots deployed with the telerobotic settlement and probably the most expensive. Using a large vehicle chassis and a drive system designed to handle diverse terrain, the Scout Rover would incorporate many of the basic features of Geology Lab along with small scale tools for excavation and sampling as well as a robust independent power system and Earth uplink capable telecommunications system. It would also have some cargo capacity used for collecting and storing samples to return to the Geology Lab, for carrying some repair parts for self-repair, and more of those deployable telecommunications transponder units for staking out territory and maintaining back-links to the settlement. Scout Rovers would be deployed in teams of two or three units so they can assist each other in extricating themselves from hazards, providing back-up power and communications, performing more complex geoassay tasks (like seismic tomography) and to aid each other in field repair.

Transport Rover and Bucket Rover: These are two of the simplest of robots employed with the telerobotic settlement but critically important ones. Most materials will tend to be dispersed about the surface of the moon or planet being settled and so various satellite mining stations may be deployed to collect them and perform preliminary refinement. And, of course, cargo from the main settlement will need to be transported for the construction and support of these satellite facilities. To move such goods around will require routine transportation and initially that will rely on these two simple robot vehicles. The Transport Rover is little more than a flat bed on wheels, looking rather like some current automated ISO container movers. It has a large capacity power storage system but only a very basic back-up solar charging system to save space and so it will travel between points on very regular routes with strategically placed charging stations. The Bucket Rover is designed for the transport of bulk granular materials and would feature a pair of drive and sensor units at either end of an open top container with a drop door panel underneath, or possibly a container formed of a pair of quarter-cylinder shells that rotate to open and drop loads, or a half-cylinder container that rotates to drop loads to either side. It would otherwise be similar to the Transport Rover. Regular transit routes would likely be marked out with self-powered LED/radio beacons as well as the usual telecom transponder units and be kept clear of debris by the Dozer Rovers. These vehicles are not the most efficient means of long distance transit but will be the primary mode of transit early on as other modes of transport requite much more infrastructure.

Ultra-Light Monorail: This is one relatively simple tracked transit alternative to the Transport Rover for very regular long range material transport and is likely to be employed late in the Outpost phase. It is, quite simply, an updated version of the trusty Banana Monorail system employed on Earth's fruit plantations for most of a century and represents the simplest easiest to deploy railed transport system one could send to an outpost in space in the form of light modular parts. The system consists of simple arched tube supports (easily made self-supporting without much foundation by being employed as a pair with hypoid curves) from which a semirigid wire rail is suspended. 'Cars' are then suspended on the rail on standardized wheeled bogies with simple electric drive units and quick-connect interfaces. They could be quickly hung and lifted off as needed with bogies connected in 'trains' or used individually using field sensors to automate spacing. Initial systems may -like the traditional Banana Monorail- employ tractor units that are self-powered relying on stored energy. But the rail supports could double as mounts for solar panels or wind turbines powering this system along its length and replacing the tractor units with individually powered carriers. A variety of power communication schemes might be employed for this, ranging from inductive charging to the more traditional technologies of spring pantograph power pick-ups. The performance of this system would not be spectacular by contemporary rail transport standards but it would be suited to linking the settlement with relatively nearby mining facilities and providing a steady flow of materials with a much lower operating overhead than the use of discrete Transport Rovers. But even this light and simple a system may be cost-prohibitive to transport entirely from space and so its use may have to wait for local industry to establish the capability to produce its major crude components.

Roadheader: This would be one of the most physically tough robots used in the whole colonization process and second only to the Scout Rover in the sophistication of its engineering. And yet it is a robot based on a very common piece of mining and excavation equipment in use on Earth today; the roadheader excavator. The terrestrial roadheader is a machine that consists primarily of a caterpillar crawler platform supporting a hydraulically articulated boom with a ball or barrel-shaped rotating cutting head at its end. It is used primarily for the digging of tunnels in relatively soft rock and hard compacted earth and has been built in a great variety of sizes and variations, from small one-operator machines used for small scale mining and underground home excavation to huge machines used for cutting rail and road tunnels through mountains. This robotic Roadheader would differ in that it would lean toward the small scale and rely on a multi-wheeled chassis supporting a highly articulate cutter arm as well as service arms carrying tools, lights, and cameras -all highly ruggedized for the tough working environment. The Roadheader would be used in combination with the Dozer Rovers, MultiRovers, Transport Rovers, Bucket Rovers, and Mobile Assemblers (equipped with small tool milling heads) to perform excavation of spaces in the Excavation Zone as well as for tunnel and strip mining, for the clearing of large boulders from the settlement area, and the leveling of terrain when building surface structures. A high power system, it would rely on direct cabled power.

Telerobotic Settlement Phase:

With the excavation work completed to the point where most of the core facilities of the outpost phase can be moved into -or rather 'expanded into'- the new subterranean structures a new phase of development would begin focused on the dual goals of self-sufficiency and preparation for human habitation. Let's consider what this new excavated settlement would be like.

Aside from its nature as an excavated structure, one of the most significant architectural features of the Avalon settlement would be one of its simplest and its something which harkens all the way back to the design of Aquarius and which we have seen influence the design of Asgard as well. This feature is the humble socket grid; a grid of sockets akin to the sockets used in climbing forms which would be installed during the process of excavation to produce a uniform plug-in mounting grid on all surfaces of the spaces. As noted above, the initial excavated structures would consist of an area of simple grid vaults carved out of the rock, using a series of strategically located entrances at the surface. Given the prospect of lowered gravity, the span for these vault spaces can be surprisingly great even when using a generally flat ceiling shape. On Earth such vaults often exceed 10 meters given a stable strata. On the Moon spans ten times as great could be employed. Domed or arched spaces can be far larger still, though those are more likely to be employed later. As precise as the work of robots may be, these spaces would tend to have very rough irregular surfaces which, by themselves, are not well suited to hosting equipment except by simple placement on the floor, which uses this space inefficiently. To make the bare rock of this space easily outfitted using robots it would be helpful to provide it with a regular grid of attachment points for modular equipment and machines, allowing the maximum efficiency of space use. Thus the space would be equipped with a kind of socket grid applied all along the rock surface like the socket grid employed in the interior finishing of Aquarius but designed for robotic use like the node connection system of the MUOL. Because the carved rock surface may vary in physical characteristics, the installation of the socket grid would be performed in several ways; using variable depth socket mounts which are mounted off-set from the rock surface to establish a laser-leveled plane, a regolith concrete or epoxy finish cover which the sockets are formed into, and using a planar space frame anchored to the rock which provides sockets at its node points. In addition, truss columns and walls mounted in these sockets would also be employed to provide vertical equipment mounting. All these approaches may be used to turn the rough space of the excavation into a precision gridded volume with an interspersed utilities network.

This approach allows the telerobotic settlement to undergo the same transformation from the use of small, exposed, self-contained systems to larger sheltered complexes as achieved in the evolution from MUOL to MUOF. Thus the interior of the Telerobotic Settlement becomes a sophisticated complex of integrated systems -a sort of macromachine- with its walls, floors, and ceilings hosting lights and various stationary equipment as well as inch-worm style robots, special tracks for PPT style materials handling, and guideways set aside for larger robots. It would be rather similar to what would be employed in the previous Service Shed, only extended over a vast area and sheltered not only from dust but also from radiation and thermal variations as well and allowing for certain areas to be pressurized with the installation of pressure bulkheads or balloon enclosures. This more comprehensive shelter allows for the use of much less rugged and more intricate systems suited to the fabrication of more sophisticated products as well as the cultivation of plants and other life forms which can be employed both as mechanisms in material processing and as part of the life support for later human habitation. All the core functional elements of the Outpost phase would now be installed within the excavated space and expanded, the exterior systems now limited to solar and wind power systems, radiator arrays, heliostat arrays gathering natural light for fiber optic distribution in the interior, and telecommunications antenna arrays. The stationary external equipment of the Outpost phase would still be employed in support of external activities, particularly with the distant mining facilities and Drop Zone. Some would be moved into the excavated space during its initial outfitting.

The basic organization of the excavated settlement will mirror that of the previous external settlement, with zones specialized for Power, Communications, Service, Science, Industry, and Storage linked by strategically placed transit routes with some incorporating automated packet transit systems or integrating vehicle terminals at surface portals for the transport rovers or Ultra-Light Monorail. Two key new features would be an experimental farming facility and, of course, the new Excavation Zone for the planned human settlement. The Farm would be based on the use of pneumatic pressure hull or bulkhead sealed chambers, hydroponics, and fiber optic lighting using external heliostats to collect light. Initial system would be modest in scale but as experience builds the scale of facilities would as well. The Farm would initially be producing products for research purposes with no specific product production intended. It would then later be used for such purposes as phytomining and the production of complex organic compounds needed for some forms of industrial production. Finally, it would find its way into a prototype CELSS for ultimate human settlement support and preliminary production of food stocks for long term storage in anticipation of human arrival.

The new Excavation Zone would be where work on the major structures of the eventual human habitat would begin and whose architecture we will discuss later. Key transit routes to the surface and external tailing piles must be established for the massive amounts of material removed with various staging facility chambers planned out for the planned excavation sequence -which will be key to the more sophisticated construction process of large domed structures. The massive amounts of material removed here would not go to waste. Some would be processed for industrial materials and some will be used as piled-up shielding material for a new set of covered transit tunnels on the surface. These would be relatively short transit routes with integral vehicle shelter structures made of corrugated alloy arch that provide shielded transit routes to key external facilities. They would be used to provide routed for frequent human access to these systems and may ultimately be pressurized using pneumatic tunnel enclosures. Some may be built in covered trenches to insure the easy surface traverse of vehicles.

In this phase the primary objective of the settlement is the establishment of an industrial infrastructure robust enough to manufacture all replacement components for all the systems used on the settlement. This would be an incremental process and since some of the natural bodies in space are not supplied with a total compliment of the necessary materials the settlement may need to establish its own links to materials sources in space. However, for the major bodies like the Moon and Mars this is less likely a problem. Ultimately, achieving this goal will require crafting a complex ecology of industrial processing among both satellite facilities about the surface and different locations within the settlement. Like the MUOL/MUOF, these different industrial activities may be parceled out to a community of businesses which lease space and operate facilities independently. However, there is no export product economy likely at this stage and so this community of operators would be working in anticipation of a local economy when human settlers arrive. Considering this, it is likely that any such lease space clients will have much closer social ties to the primary sponsors and developers of the settlement than would be typical of Asgard facilities. This is a project by and for prospective settlers -people, companies, and organizations looking to profit primarily in the location of the settlement -from what they make for themselves there- and not on any external market. However, as noted previously, the Earth's Moon at least does offer some possibilities of export products in the form of processed materials, beamed energy, and perhaps even food products owing to its relative proximity to Asgard settlements in Earth orbit. To realize this, though, large transportation systems such as mass launchers and extremely vast solar power and microwave transmitter arrays will be needed to realize these export industries. These facilities can be built telerobotically much like everything else but their very large scale means they will require much time to realize, even with the ultimate increase in construction performance likely when human technicians are available to operate systems locally.

In this phase the key new forms of robots would be those designed to operate exclusively within the sheltered environment of the excavated structures. These would tend to be much lighter, simpler, and often much smaller than the robots intended for external use and also probably much more numerous. But the same strategy of modularity would apply and most of these systems would be designed to integrate into the socket grid and the utilities system it supports.

Inchworm: This is the key workhorse robot of the excavated settlement and is based on the very same concept employed by the robots of the MUOL. It consists of a simply multi-jointed robot arm with two identical end-effector connectors that allow it to 'walk' inchworm-like between modular attachment points while picking up different modular tools from pallets. It can even be made self-mobile by the use of a special form of anchor module incorporating batteries and stabilizer 'toes'. Inchworms would be designed in several scales and reaches to suit different load handling capacities and levels of tool precision.

One of the interesting characteristics of the Inchworm robot -and a capability likely refined in their use in MUOF facilities- is that they would exist on the settlement's network backplane as largely independent devices with a very simple generic control interface architecture but would normally be used in groups or 'workstation sets'. These sets would be virtually assembled on-demand by the teleoperator who would summon a set of Inchworms from a service pool to a specific location in the settlement's spacial grid then assign them different positions, tools, and to different user interface controls within his operators station, the teleoperation software composing and coordinating these robots as a single system on demand. Some might be assigned to camera and sensor duty, others to carry work lights, others to use specialized tools, and others to perform the main work. It would be rather like having a community of robotic 'helping hands' akin to those featured in the Jim Henson fantasy film Labyrinth.

Carriers: Based on the chassis of the MultiRover or one of similar scale, Carriers are similar in design to the Transport Rover and would be used for the simple transport of goods within the excavated settlement, especially in areas where the socket grid has not yet been installed. Using plug-in attachments, the Carrier would be equipped with various pallets, containers, and cartridges according to what it is transporting and would even serve as a self-mobile anchor for smaller Inchworms to make them into a simple service robot.

Movers: These robots are essentially akin to the contemporary fork-lift but with a much more articulated lift mechanism that can manipulate items with both forks and a clamp that interfaces to the standard wall sockets. It would be used for the simple moving of equipment and goods and the loading and unloading of Carriers and the PTS but would also be used for the installation and removal of some standard modular equipment into the wall socket grid.

PTS: The Packet Transit System is an extension of the Personal Packet Transit technology developed for use on Aquarius and would consist of a conveyor system which shuttles goods on standardized modular pallets or in modular containers. Based on a supported conveyance technology, it would be used for the more direct automated integration between processing systems in the Industrial sector of the settlement and would integrate with the Ultra-Light Monorail for long distance external transit.

Conveyors: Not to be confused with the Carriers whose chassis they may share, the Conveyors are self-mobile segments of a conveyor belt used primarily for the transport of excavated material. They line up in long chains and link together on demand to form a conveyor which can be repositioned on demand as excavation work proceeds and used to divert portions of material to different areas for different handling. More specialized switch, bend, lift, sorter, and crusher units would be integral to this conveyor chain.

AWS: The Automated Warehouse System is another derivative of technology from Aquarius which would be based on combining the cargo handling systems of the Movers with the conveyor systems of the PTS as well as a box frame structure used as volumetric shelving. Located in storage facilities and directly integrated with the PTS, the AWS serves the purposed of automated storage and distribution for a large diversity of items. The production systems of the telerobotic settlement would tend to be less specialized than typical of industrial facilities today, likely relying on advanced forms of today's rapid prototyping technology and switching production on demand. This means a large rapidly changing diversity of goods calling for a more generic form of storage and handling, though probably with less stored volume for any given item -since the settlement is largely limited to production for its own internal uses. Using this AWS and the PTS it becomes possibly to treat the entire industrial production sector of the settlement like a single on-demand production facility of diverse goods.

Sweepers: These would be a variation of the Mover or Carrier chassis that is designed for the purpose of keeping the settlement's interior free of dust and small debris tracked in as goods and robots move between interior and exterior areas or the excavation areas in various parts of the settlement. In a vacuum or low pressure environment, this is not as simple an activity as it may seem with these robots needing specialized collection devices and perhaps stored gasses for the jet blowing of dust and debris. Stationary sweeping stations would be integral to most of the portals into the settlement and as most staging areas for excavation work so these robots would be serving a supplemental cleaning role.

Human Settlement Phase: Upon successfully completing the establishment of an industrial infrastructure that can meet human survival needs for an indefinite period for at least a small community the telerobotic settlement can finally begin hosting its first human inhabitants. With human habitation -even in small numbers- would come an order of magnitude increase in the pace of development by virtue of the elimination of latency in the control of the settlements robots and systems. Most work on the settlement would still be performed using the same systems because any activity done in a non-pressurized environment is inherently hazardous for human beings. But without the latency of long distance telecommunications robots can be controlled in real-time and with the benefit of greater spacial awareness for teleoperators resulting in a huge advance in the speed of activities.

In preparation for its first residents, the settlement would begin the production and stockpiling of basic survival supplies such as potable water, foodstuffs, and the like and begin reorganizing the settlement's structure into pressurized and unpressurized zones, deploying either pneumatic hull modules or special pressure bulkhead modules. Industrial systems would begin tooling up for the production of goods for human needs and for the combined use of local and long distance telerobotic control. The settlement would also begin extensive clearing and expansion work in its soft landing zone, deploying new support equipment, new telemetry and navigation systems, safety/waystation shelters based on regolith bermed corrugated alloy shells with pneumatic inner hulls, and a new collection of large pressurized rover vehicles designed both for manned exploration support and the shuttling of passengers between the soft landing zone and the excavated settlement. These new vehicles would be designed for both manned and teleoperated control -since until humans arrive to drive them they must be teleoperated like everything else in the settlement.

It seems likely that the transportation technology for colonists will derive from the technology of the Asgard settlements themselves, using Transhab and EvoHab hull systems, MUOL style structural systems and modular drive components, and relying on a series of more specialized shuttle vehicles and the use of orbital stations as waypoints for the transfer of passengers. Thus we can envision the basic interplanetary vessel of the Asgard age as being simply a variation on the design of the Asgard orbital colonies themselves, composed of a core truss hosting thruster components and solar power systems and one or more transhab hulls -for the smaller scale vehicles- or a larger EvoHab hull with an option to host internal gravity deck structures. These would not be designed for landing on the surface and would instead use soft-lander shuttle vehicles either carried as cargo or stations at waypoint orbital stations. Given sufficient advance time for the cultivations of its industrial infrastructure, the teleoperated settlement has every possibility of being able to manufacture and launch all these vehicles on its own, though it is more likely that a combination of both Earth/Asgard sourced spacecraft and domestically produced spacecraft may be employed in a colonization program. As the number of prospective settlers increases along with a desire for routine two-way transit, these simple interplanetary spacecraft may evolve into a class of vehicles called Cyclic Shuttles. These would be transit vehicles scaled to the size of a modest Asgard settlement habitat placed in a perpetual orbit between key destinations in the solar system and relying on local transit vessels to service them when they are passing in close proximity to their destinations. They would not be fast, but they would afford the comforts of a full Asgard settlement for the duration of the trip. This, however, is more likely an eventuality for more distant settlement locations, such as Mars. The Earth's Moon is close enough and easy enough to transit to the surface to due to a lack of atmosphere that reusable soft lander vehicles may be able to shuttle whole between it and Asgard facilities.

The initial human habitat for the settlement may be based on the temporary use of the same grid vault sections of the excavated settlement employed by the teleoperated systems. Depending on the nature of the rock strata, using either modular prefab pneumatic hull units or modular bulkhead units and a reinforced sealant covering which seals any possible fissures or permeable inclusions in the rock -though some strata will hold adequate pressure without any additional treatment except at the interface to the bulkhead units which are likely to be mounted using a sealing method such as injected epoxy. Outfitting these spaces for human use would be done, again, using the plug-in socket grid and the same approaches to interior decor pioneered on Aquarius, though these initial habitats will likely be more strictly functional in their design. Later, these same spaces may be turned over to industrial and laboratory applications, the original industrial systems not having need of a pressurized environment for the most part but now needing to accommodate a situation where a growing number of systems will be managed both directly and telerobotically as well as increasingly repaired by human technicians. It will always be more efficient to leave systems in an unpressurized environment where it is not strictly necessary because of the atmospheric resources this saves and the reduction of fire hazards. But there will definitely be a growth in pressurized industrial facilities over time.

The ultimate permanent human habitat will require a more specialized architecture. Its construction will have begun in the telerobotic phase with key spaces likely pre-excavated. But full development may not begin until the first contingent of settlers arrives, partly because of the need to customize some features and the need for a more aesthetically oriented horticulture. Let's now consider the design of this permanent subterranean habitat.

The primary problem of living in space is the fact that one is spending virtually the entirety of one's life in an indoor environment -a fact very conveniently overlooked by most science fiction and most space advocates. Indeed, even the use of conventional windows of all but very small size will be a rarity because of their production cost, their poor radiation shielding, their fragility, their relatively complex installation, and higher potential failure rates for their mounting fixtures. And, of course, in the deep interior of an excavated structure window views to the surface -no matter what the window size- are not a possibility. It is therefore critical to the health of inhabitants who may live their whole lives in space that an architectural strategy be devised that can provide the necessary sense of naturalistic open space in spite of this situation.

This is an architectural area that has seen very little exploration in the past despite the very long history of underground dwellings, perhaps because contemporary designers have been so focused on the exterior appearance of structures and the merging of their interior and exterior environments. Concerns for surviving nuclear war and rare explorations of subterranean cities have fueled most experiments in completely enclosed underground architecture. This has left us with few definitive solutions for the problems of fully enclosed living environments while there are many practical applications for this such as the design of subterranean malls and terminal facilities for our increasingly sophisticated urban subway systems, the conversion of salt mines into health resort facilities, the conversion of mines into multi-purpose commercial facilities, and the development of comfortable subterranean homes in less than optimal property locations and regions now subject -thanks to Global Warming- to steadily increasing incidence of extreme weather. It's surprising that more space advocacy groups have not pursued subterranean dwelling research as the restriction to primarily interior design offers a lower cost in prototype habitat development than for any surface habitat design. With the external structure of such habitats essentially invisible, whether or not one has actual excavated structures to work with matters little to the research and design results while the use of actual excavations is still likely to be far cheaper than the fabrication of prototype shells for traditional surface habitats -based on what is essentially spacecraft hull technology- while also resulting in the creation of actual functional buildings with practical uses and real estate value.

When past futurists have considered the possibility of living permanently in space, they have tended to arrive at a notion which I've come to refer to as The Great Indoors. The premise here is that one overcomes the limitation of enclosed space living by simply making the space so vast as to be indistinguishable from an outdoor environment. Hence the common vision of space colonies as vast rotating orbital megastructures housing a recreated naturalistic landscape and ridiculously anachronistic free standing suburban homes. While such structures may be possible in the future -indeed, in sizes orders of magnitude larger than originally envisioned with the advent of diamondoid materials, the scale of their construction precludes the possibility of quick creation or low cost. Thus generations may be needed to realize them, while in the mean time they offer no practical solution for an intermediary form of habitat for the people who must spend their whole lives in space building them. Interestingly, while Marshal Savage originally considered these vast orbital colonies as over-kill and thus unlikely because of their protracted construction periods, he himself arrived at what was essentially the very same Great Indoors strategy when it came to surface settlements. Taking a page right out of classic science fiction, Savage's vision of Avalon was based on the creation of domed-over crater habitats hundreds of square miles in area and using the same pneumatic water-filled hull technology he envisioned for Asgard. Aside from the fact that this kind of hull technology was presented with no clear proposed fabrication method and faces serious problems with issues of tremendous static hydropressures even in a lowered gravity environment, the same problem exists for how people get by during a very protracted construction process. Even just filling them with an atmosphere could be a task taking some many years of time to complete.

Savage's solution to the transitional habitat needs was simply the same kind of crater dome structure on a much smaller scale and linked by an underground network of transit tunnels -a type of structure equally suitable, in theory, for early outposts and permanent settlements and readily increased in scale as the means to fabricate larger domes developed. A practical strategy, were it not for the speculation on this membrane hull technology. We now see this idea as probably not feasible because of the very limited life span of known transparent membrane materials, the limitations of water as a radiation shield and its much less than expected transparency, and the inability of such domes to resist micrometeororid puncture or -when so small- to provide an adequate buffer in atmosphere to allow evacuation and/or repair as might be the case with much larger domes. However, while the specific structural technology he chose may not be practical, the essential dwelling design he arrived at points toward a very practical solution to continuous enclosed space living with modest scale structures which is completely independent of any particular construction technology and potentially very practical in the context of both excavated and built-up habitat structures.

Thousands of years of urban development has demonstrated that it's not strictly necessary for a comfortable habitat to be based on vast open expanses of landscape. In fact, for most of human history people have tended toward small enclosed habitats well sheltered from a wilderness environment often regarded as dangerous. Today's notion of 'country living' as a luxury is a very recent invention and has more to do with the legacy of European upper-class land ownership and contemporary environmentalism's aesthetic veneration of nature than any practical need. Indeed, it's tragically ironic that environmentalists, because of their compulsion to live on the edge of wilderness they so venerate, are very often the vanguard of the ultimate suburban sprawl that threatens that wilderness. But it's also true that, prior to the advent of mechanized transport, the scale of cities was quite small relative to their high density and there was no suburbia as a transition between rural and urban habitats. Urban residents had quite ready access to rural environments within casual walking distance. In space, any excursions outside of a habitat will be limited by time issues related to radiation exposure and take on the nature of a rather grueling expedition. There will be no casual outdoor activity. So there is still a need for some sort of recreated outdoor environment, even if it must be of much more modest scale than The Great Indoors.

Luckily, the space environment presents us with a certain advantage for realizing this with excavated structures. While excavated spaces of may hundreds of square miles in area may not be possible, as noted previously, the lowered gravity conditions of most lunar and planetary locations in the solar system affords the possibility of domed and vaulted space of quite great spans compared to their terrestrial counterparts. In fact, in many locations there is a possibility of pre-existing lava tube structures possibly a mile or more across and hundreds of miles long! So we have the potential for a Not-So-Great Indoors that is still quite generous in potential area and fully adequate to the demands of continuous enclosed environment living. But how do we turn such spaces into a habitat that can emulate naturalistic outdoor spaces? What sort of environment will we create in such spaces?

Savage's small habitat concept points to a strategy that has been employed in pre-industrial urban environments for centuries; the walled compound. As human beings began gathering in large numbers in certain locations their strategy for dwelling design was organic. With the advent of agriculture and the use of permanent dwelling, many tribal villages employed the use of walled compounds as these walls provided shelter from wind, a barrier to dangerous animals as pests, and provided some protection in the event of tribal conflict or raids/thefts. The layout of these communities tended to treat the walled enclosure as an integral structural component of individual dwellings out of simply economy, since the labor involved in wall construction was large. This resulted in a roughly radial organization with a large open courtyard in the center used for communal activity. As small tribal community integrated into larger community this walled enclosure became the boundary of the extended family parcel which itself was arrayed around a new larger community center; the town square or court. Such courts often defined the boundary of variations on tribal community based on ethnic group, very extended family clans, or trade guilds. As cities grew even larger, new levels of hierarchy developed around even larger city 'square' centers from which branched streets and wide avenues hosting arrays of both older courts and individual family enclosures. Avenues served as primary transit vias through the city hosting carts and later early mechanized transport. Streets were traditionally narrow and of human scale. (as opposed to the contemporary American grid city strategy where avenues are two-way vias along the long side of rectangular blocks and streets are one-way vias on their short side) Remnants of this traditional organization can often be seen in Old World cities today.

Savage's small habitat concept suggests a similar strategy for the layout of a large excavated habitat; a series of domed chambers with functional spaces arrayed radially along their perimeter linking together in a hierarchy of progressively larger and more public dome spaces linked by a series of arched avenues. Given the availability of a large natural lava tube, this could serve either as a natural avenue and/or a primary community center space. The use of domed and arched spaces and radial organization here is predicated on the observation of ancient builders using domes for religious purposes. Domes in ancient architecture were quite simply a mechanism by which a new type of skyscape could be created to serve as the basis of representing, through its decoration, the cosmological order of a particular religion or belief system. These ancient builders realized that the dome shape, given sufficient height and a decor that can obscure its perimeter edge, creates the impression of a virtual horizon greater the physical boundaries of the domed building itself. Thus one enters a kind of 'interior outside' space within a dome. This observation was later rediscovered by Buckminster Fuller who noticed that his large geodesic domes would be readily accepted as a new skyscape by those inside them as long as the height of the dome was sufficient that the details of its frame structure could be obscured by distance, its framing further obscured by light transmission or reflected light (or in some cases by darkness, as is the case with planetarium domes) and that its edge perimeter was obscured by other interior structures. The smoothness of the dome and vault shape also obscures the actual area of the structure when there are no obvious visible structural elements from which the mind can extrapolate dimensions. Even when a geodesic frame is visible, its countlessly repeated pattern visually merges into a texture rather than a dimensional feature, leaving the specific dimensions of the shape hard to judge. Maintaining a radial order to the perimeter room structures further enhances this. They become, in effect, a kind of walled enclosure for the domed 'outside' space. The virtual horizon and ceiling established by the dome are perceptually infinite. Only these perimeter features give us a limit to apparent space.

Lighting is also critical to the illusion of an outdoor space and it would be advantageous to employ natural sunlighting as much as possible, both for health reasons and for energy efficiency. But how can this sunlight be provided without windows? Though the use of large arrays of surface mounted heliostats which funnel concentrated sunlight to a series of optical shafts into the habitat for further distribution by fiber optic cable and a variety of emitter devices. This technology is commercially available today and would have already been well advanced during the Asgard phase as a means to provide natural lighting in heavily shielded EvoHab structures -especially those used for farming. Similar systems will have already been deployed during the telerobotic phase for similar farming purposes. The domed and arched ceiling spaces of the habitat would be used as the primary emitters for this collected light, using either geodesic arrays of emitter panels or perimeter edge emitters with the domes and vaults serving as a reflective diffuser. This approach has the great advantage over windows in that it filters most IR and UV spectrum (that light energy being recoverable by in-line photovoltaics) and can be easily made to compensate for the differences in ambient sunlight intensity in different locations in the solar system by simply varying the area of light collection accordingly. Reflective light diffusers also have the potential to serve double-duty as projection surfaces for information and entertainment purposes, and can be employed as 'virtual windows' where spherical exterior views or computer generated skyscapes can be displayed.

Large area virtual windows may also be a common feature for other interior decor -another concept likely pioneered in the Asgard phase- and though expensive and limited in performance today, should by the time of Avalon ultimately be capable of great economy and vast display areas though the use of flexible polymer materials and with potential holographic display characteristics making them far more realistic in the views displayed. Again, multi-duty from these for such uses as telecommunications, virtual window-walls for teleconferencing, data and entertainment displays, and CAVE (computer augmented virtual environment) displays for teleoperation and telepresence application makes them all the more practical.

Artificial lighting is, of course, a necessity but here too we can expect to see, by the time of Avalon, advances that will likely make LED based illumination the dominant technology with broader spectrums and large area laminate panel fabrication akin to current electroluminescent devices but with far larger surface areas possible -perhaps continuously produced panels many meters wide and infinitely long and with all sorts of possible integrated electronics such as motion sensors and dimmers, touch switches, and color tuning. Thus the largest of chambers and the longest of tunnels could employ continuous ceiling surface illumination and an endless variety of artifacts could have illumination built-in. So one can anticipate this underground habitat to be rather indistinguishable from at least an urban outdoor habitat.

Thus we arrive at our basic architectural model in the form of a domed central space with a garden landscape in the center and functional space arrayed bout its perimeter in the form of simple rectilinear vaults using window-wall openings which can be fitted with glass windows (for sound barrier purposes) screens, or various decorative facades and a kind of short roof overhang whose primary function is to obscure the perimeter edge of the dome ceiling and host a variety of lighting and environmental equipment. This basic form would extend from the small individual dwelling with little more than a garden atrium at its center to various scales of communal space using terraced perimeter structures of many levels and a central space landscaped with everything from forests and small mountain-like rock features to lakes simulated ocean shorelines. In fact, the community could chose to try and recreate specific biomes in different chambers for aesthetic purposes. Each such chamber, from small to large, would be treated as a single pressurized unit, its access portals equipped with doubled quick-close bulkhead hatches that allow them to be sealed in the event of a pressure emergency or fire. (the dual-hatch arrangement would allow them to function as airlocks)

These largest of domed chambers are likely to require a special excavation technique because of their great height. Excavated at lowered levels compared to other smaller chambers, their creation would be based on the tunneling of an upward and inward spiral corridor which roughly defines the perimeter of the dome. The dome wall is then carved out from the top-down, the tunnels becoming a series of terraces incrementally removed and the 'floor' of the dome is carved out. As the dome is carved out incrementally, its space could be made habitable, though its habitable structures would have to be removed on a regular basis as the dome is expanded, its floor progressively lowering and widening. In this way the settlement could incrementally grow domed chambers over time to suit settlement expansion, though they must pre-plan a maximum are so that perimeter structures are accounted for. Obviously, a vast quantity of material would be removed in the process of this construction and need processing and transport to the outside for disposal.

The overall layout of the habitat would be hierarchical. The simple grid vault spaces created during the telerobotic phase would be closest to the exposed edges of the rock strata the settlement is using. Deeper in would be the radial habitats arrayed roughly in a star branched network off key avenues which may be established by pre-existing lava tunnels. Smaller scale dwellings would be arrayed in clusters about community centers in the form of larger chambers, mimicking the arrangement of pre-industrial cities. These clusters, in turn, could be arrayed around larger central chambers and/or along avenue tunnels which may be made of natural lava tunnels. This presents three basic dwelling options; private chambers, communal chambers, and avenues.

While the individual dwelling would offer the most privacy and personal square footage, it would also generally offer the smallest open space area. A typical design may feature a central atrium of 20-50 feet in diameter. It would be possible for an individual to invest in the construction of a larger more distant private chamber if they're willing to deal with the overhead of its landscape maintenance and more independent life support systems. This offers the possibility of a kind of homesteading, though by approval of the main settlement and using structures whose location and size might later allow them converted into communal chambers as the main settlement expands. (it would still be too hazardous for people to live in disconnected habitats and all structures must be planned relative to the growth plan)

The larger communal chambers would offer dwellings in the form of terraced residential units along their perimeter, offering a much larger but shared central open space. They would also offer a greater potential variety in the aesthetic style they could employ, some choosing to make their perimeter structures seem like highly visible buildings and others merging them into the synthetic landscape to give precedence to the naturalistic aesthetic. Owing to the much larger scale of space, the illusion of an outdoor habitat within the dome would be more effective and the shared labor in maintaining its landscape could reduce the amount of individual work. There is also room for a larger variety of 'outdoor' recreational activities.

The avenues would tend to present a more urban environment and concentrate most commercial and light industrial activity so its likely residences will be similar to contemporary townhouses. The exception, though, are the larger lava tunnels which could be as large across as the larger communal domes, presenting what is essentially a valley environment with a terraced side landscape. These would use much the same form of housing. Avenues would also concentrate the use of mechanized transit, which would consist of a hybrid PPT/PRT system located behind terraced dwellings rather like the configuration likely employed with Aquarius. Each small dwelling and communal center would feature at least three primary tunnel exits, one of which could be dedicated to PRT use.

The basic construction and finishing method employed with these human habitats would be generally the same as that employed by the earlier telerobotic habitat with the same reliance on a grid socket system. However, the more complex topology would require the use of both a rectilinear and a geodesic grid with some more specialized components to translate between their different geometries. This would be done to allow for the use of more convenient rectilinear finishing components in the more functional area of perimeter chambers while more specialized components would be employed to accommodate the domed ceilings. (which must still be robot traversable for sake of maintenance -especially when they are large) While pneumatic hull modules were sufficient for pressurized environments in the simple grid vault spaces, these large domes and vaults with their highly variable topology will require a more surface integrated method of sealing -assuming the rock strata itself is not sufficient. This would favor techniques based on spray applied carbon or alloy fiber reinforced ceramics and epoxies. The socket mounts would provide an off-set between the wall surface and the socket points which smooths out variations in surface topology while also allowing all surfaces of the structures to function as utilities routing -including the area of domes and vaults. This is necessary because the need to bore numerous small tunnels through rock -and create more potential pressure failure points in the process- to accommodate a more conventional utilities routing makes surface mount routing a much more practical alternative. So in general the actual rock surface would be quite invisible in most inhabited areas, covered by a series of both functional and decorative panels concealing these utilities routes but keeping them easily accessible.

The choice of materials and pattern of fixture, furniture, and human artifact design in the interior design of these human habitats would strongly reflect the nature of the industrial production and the need for materials economy in these new communities. While the telerobotic settlement will have achieved a sophisticated level of industry prior to human settlement, the limited scale of its industrial systems and the somewhat more limited spectrum of resources in likely surface locations in space will favor materials that need the simplest processing to produce and are well suited to recycling. Interestingly, this would actually result in a much more naturalistic environment than is commonly envisioned for space habitats because it would mean a reliance on a large variety of simple organic materials where agricultural production and the re-use of agricultural wastes offers a simpler alternative to large scale industrial production. For instance, we might see a heavy use of materials like bamboo, wheatboard, sisal, hemp, and kanaf fabrics, and similar products because the types of plants these come from are very easy to produce in a high density hydroponic agriculture systems and easy for machine processing. Plastics like polyethylene will be popular because it can be grown directly in the tissues of some genetically modified plants. Terra cotta and other clay and ceramic materials will be popular because their raw materials are easily processed in small volumes and with little energy. High tech ceramics will supplant many alloys, particularly steel alloys which have very high energy overhead in production. And, of course, the natural regolith and rock will find its way into a variety of materials. So we can easily imagine an environment here far less hard and industrial than has been commonly envisioned by space advocates and science fiction to date.

Similarly, the design of many furnishings, fixtures, and artifacts will reflect a design ethic of multifunctionality and design for reuse. On earth we have tended to cultivate a design culture rooted in the wasteful agenda of consumerism which overspecializes our artifacts making them less adaptable and therefore more easily obsolesced. This simply isn't practical in the space environment and in the situation of early human settlement. So the dominant design strategy will be embodied in the principle of 'min-a-max'; maximum diversity of function from the minimum diversity of components. So instead of a vast assortment of electronics devices filling out the spectrum of entertainment, communications, utilities, and computing uses, a single computing platform with a standardized set of simple components would comprise all these functions. And even on the inside of these devices, instead of a diversity of specialized components, a higher reliance on more generic components which can be freely adapted to many functions such as Field Programmable Gate Arrays. (some readers may recall my past proposal for the FMF to develop its own computing platform based on a more advanced form of FPGA called the Homogeneous Processor Array that performs the functions of any logic structure and RAM in the same device) Furniture and fixtures would employ a modular component strategy where the wall socket grid, space frame and truss systems, and perhaps a derivative of contemporary T-slot framing all integrate to support an environment of assembled-on-demand and perpetually repurposed furnishings. We can envision a style of furniture akin to the Living Structures of designer Ken Isaacs and similar work by designer Andrea Zittel. A great deal of the preliminary design work leading to this design ethic will probably have already been done in the Aquarius and Asgard stages as these types of communities will face the same kind of agenda of materials efficiency and employ much the same strategies.

Recreation and exercise will be critical to the health of human settlers -especially in a lowered gravity environment- and there are particular challenges to this in an environment imposing continuous indoor living. The larger domed chambers will certainly offer more room for recreational activity but their design is still going to be based on the illusion of space rather than actual area which means favoring activities which economize on the need for space. The presence of gravity, even if lowered, will allow for the creation of swimming pools and various water based forms recreation, though this will depend on the efficiency of water recycling and the availability of water in large volumes. Martial arts, yoga, and tai-chi will likely be popular and practical forms of exercise and recreation thanks to their reduced space needs. Dance would be popular for similar reasons, the lowered gravity allowing for spectacular gymnastic and acrobatic moves. Other activities would likewise be enabled by the lowered gravity. Rock climbing may be popular, being relatively compact and enhanced in ease by the lowered gravity. Perhaps this may be supplemented by 'trapeze webs'; jungle-gyms of vast scale based on large webs of cables and net and fabric platforms made easy to explore by the lowered gravity. Marshal Savage envisioned human flight as a recreational activity in lowered gravity, though this notion was contingent on his very vast crater domes. However, this could still be a possibility in the lava tunnel spaces or could be alternatively based on the creation of 'freefall' chambers using fans. (lower powered and over much larger areas compared to the 'freefall' rides on Earth) Small vehicle racing is another possibility, with human powered vehicles the safest option. The use of digital media, computer games, and virtual environments will be very popular thanks to their great space and material economy and their ease of communication -by radio- from Earth and other locations but to enhance their physical activity the community may choose to repurpose the systems used for teleoperation to create CAVE systems for entertainment purposes, allowing for more physical interaction with virtual environments. As noted in the piece on Asgard, the importance of sexual interaction in health and recreation should also not be overlooked. In the zero-g environment this required some special consideration because the very mechanics of intercourse are effected by the microgravity environment and privacy is at a premium in these habitats. With the surface settlements of Avalon, these issues will be less of a concern. There would be more venues for privacy and meeting places with a more naturalistic environment. But this will most definitely still be a very close-knit community and a culture which is not sufficiently progressive in its attitudes toward sexuality will most definitely generate social and psychological problems as a result.

Surface Habitat Phase: This is the last phase we will discuss in this piece and it basically concerns the expansion of the initial settlement and the creation of new settlements into regions where excavated structures are not allowed by the particular geology. The evolution from individual settlement to the creation of a true colony of many settlements.

The chief limitation of the excavated habitat is its reliance on nature to provide a suitable structure. Initially, with little of a moon's or planet's surface exploited for habitation, odds are high for there to be some locations which, on the mean, are both logistically advantageous and offer suitable strata for simple excavation. But as the colonial community reaches farther and farther away from initial settlements for resources and seeks to deploy large transportation systems -such as mass drivers and orbital tether systems- to make export more economically viable it will become necessary to create locate habitats independently of what nature does or doesn't provide. This will compel the development of methods to create habitats in any location. In general, if one goes deep enough in the strata of a moon or planetary body, one is bound to find strata suited to excavation. But the deeper one goes the more complex and expensive the excavation process becomes and the more one may be faced with problems of natural heat if there is any kind of latent geothermal heat or tectonic activity. (though it seems most places we've been interested in settling in the solar system have no appreciable tectonic activity) At a certain point the cost of the excavated habitat will reach parity with the cost of the built-up structure -especially when already established settlements become capable of producing suitable building materials in large quantities. Here then is where the use of surface settlements starts to become practical.

Now, calling these 'surface' settlements is something of a misnomer because, in practice, they may still be predominately subterranean and in basic interior organization they would be completely identical to the earlier excavated habitats. The difference is that they would be based on the construction of thick built-up shells of reinforced stabilized regolith -regolete- which are wholly or partially buried in the granular surface regolith material. Their construction would be based on the surface strip-excavation of foundation spaces which would be at least as deep as the 'height' of the main functional habitable spaces. In these depressions, the built-up construction of the habitat shell structures would be done using methods akin to contemporary slip-form concrete construction, though based on a regolete material likely reinforced by homogenous admixtures like fiber reinforcement, with carbon or nanofiber a likely choice. Mound forming methods, where a carefully sculpted mound of loose material is piled up to serve as the form for loose poured shell structures, is one like method -especially near initial habitats whose own excavations will have produced vast amounts of waste material for this purpose. But the use of climbing form systems is also likely, exploiting the very same socket grid system that would ultimately be used for outfitting the shell interiors. Once complete, these habitat shells would be completely or partially buried allowing vehicle transit and surface equipment installation around them, particularly solar power systems, radiators, and heliostat collectors. These reinforced regolith shells would be very thick -perhaps many meters- and so would generally limit the habitat in terms of window access in much the same way as the excavated structures even where they are exposed on the surface. Their exterior surface would host much equipment, especially solar and radiator panels. However, they have the option of being made directly light transmissive by incorporating fiber optic light guides in their shell matrix with matched pairs of heliostat collectors on the outside and emitters on the ceiling inside. It may eventually be possible for these kinds of optical systems to be image transmitting, allowing for a virtually transparent structure much as Savage envisioned even if very different in composition.

Inside, these habitats would be no different from their excavated counterparts and the very same strategies of interior design would apply. However, their design has more flexibility because these built structures would not be limited in layout by the vagaries of natural rock strata and, unlike excavated structures, can be demolished and removed in the event of obsolescence -though with shell structures so thick this would not be a quick and easy task. The use of nanofiber reinforcement in regolete may also allow for much larger span structures than the natural rock chambers, bringing us a little bit closer to those vast crater domes of Savage's imagination. Though they would most likely still be far from that in scale, shells with spans of a few miles may actually be possible.

With the ability to make such structures in any surface location, so too would come a new option for 'homesteading'. With the advent of both telerobotics based on largely self-replicating robotics and a similarly automated industrial capability relying on local materials, the cost of building additional habitats starts to shrink to the cost of its energy over time, which would itself be renewable. (in other words, if one wanted to build a house on Earth today its cost is determined by the market value of its materials and labor. If one wanted to build a habitat on Mars, it's labor cost is nil because it's being done by robots whose own replacement parts overhead is handled by other robots and its materials cost is nil because their extraction is done by robots and so the cost comes down to the energy used -which is renewable and expandable in spot demand as a function of robotic product so it too becomes nil) This effectively reduces the cost of settlement to little more than the cost of the settler's transport and the cost of local temporary accommodations during the time he spends conducting the construction of permanent dwellings of his own or the cost of hiring existing settlers -and/or their robots- to do this work in advance for him -which would basically mean a debt based on some kind of industrial/agricultural production from the homestead or the trade of some exclusively Earth or Asgard produced product sent as cargo with the settler whose value is indexed to its transport cost. This means that, while becoming a space settler would still represent a personal expense -because of transport costs- so high that it represents a once-in-a-lifetime trip, it would be more accessible to more people the larger the colony grows. Indeed, settlements may be constantly expanded with spare production in a deliberate attempt to attract more settlers with offers of free accommodations just so so they can help flesh-out the social ecology.

As I noted at the start, I foresee this pattern of development and approach to habitat repeating itself throughout the natural bodies of our solar system with variations in the structural technology according to the differences in local environment. For instance, the moons of Jupiter would present special challenges due to radiation and unusual forms of strata. Manned spacecraft may require powerful magnetic or plasma shield technologies to protect them and their service shuttle vehicles may need to be deployed from deeply excavated launch and landing facilities in the form of large deep wells. All habitats in such heavy radiation environments may be limited entirely to deeply excavated structures, and on Europa this means deep within a strata of ice which the habitats must insulate themselves from to prevent their own latent heat from melting their shelter. We might also see the development of a kind of inverted 'surface' shelter there where habitat structures relying on their own buoyancy are planted in the underside of its vast ice sheets. But no matter the location and no matter the technological variations, we can expect a living environment very similar in terms of development pattern, form, strategy for dealing with continuous enclosed environment living, and in many ways general appearance. Only with the advent of the Diamond Age is this picture like to change, with the advent of a new class of self-replicating materials and structures capable of a very radical approach to habitat which we will be discussing in the next piece.

Eric Hunting