Excerpted from a discussion by Eric Hunting:

“I’ve long been interested in construction automation, both for use in space and for domestic uses. Unable to employ much sweat equity in building myself, I’ve long been interested in the idea of building the machines that might build for me. One concept that I’ve wanted to explore was self-assembling space frame systems based on a scheme devised by designer Scott Howe;

http://www.plugin-creations.com/us/ash/research/projects/proj33/project.htm

http://www.plugin-creations.com/us/ash/design/projects/proj35/project.htm

Howe has spent a long time exploring construction automation and self-constructing structures, but usually with machines and components of very large scale. In this Trigon system, developed with NASA, he got down to a much more modest scale that seems accessible to possible hobbyist experimentation while still producing structures of significant size. In this system a panel-type space frame consists of simple modular robots with powered hinge and linking mechanisms on their edges allowing them to climb over each other by flipping themselves and alternately attaching and detaching to assemble themselves into rigid structures. Connected units would form a distributed power and data bus. This is a potentially easy machine to make as a combination of two flat plates (wood, aluminum, composite) with active components sandwiched between them. It could be sized such that a relatively modest number of units would make up structures of significant scale. Many kinds of functional elements could be integrated into the center space of the frame panels; quick-attachment mounts, lighting, fans, speakers, displays, solar panels, sensors, WiFi nodes, and so on.

Howe also proposed a Cubolding system based on similar but cube-shaped units;

http://www.plugin-creations.com/us/ash/research/projects/proj32/project.htm

Here each cube is outfit with automatic corner connectors and internal actuators that push cube units forward or turn them over corners. Sets or trains of cubes traverse the surface of structures in inch-worm fashion. This strategy could be employed with either mechanisms integral to every cube or using a robot that can move inside and between the hollow cubes. This makes for structures with more rigid connections but the articulation needed is more elaborate.

These kinds of systems would be very entertaining to watch in demonstration and would make for a very valuable venue of robotics research. They could make just the setup of an exhibition a very interesting event. But, by themselves, they probably can’t serve as resilient housing and probably aren’t going to be strong enough to support the loads of multi-floor buildings. They would need additional materials that, right now, would end up being put on by hand. It wouldn’t be terribly difficult. If you’re producing simple self-standing enclosure forms like domes, pre-made outer tent skins and plug-in foam and cloth interior panels. Unless you’re dealing with a structure that has to be dismantled, moved, and redeployed very frequently it’s hard to justify the cost of so much active hardware in every component, even if this system leverages it more than other cellular robots. You need to be getting a fairly large amount of shelter area relative to the unit hardware investment. Still, even if they have a long way to go, they would be fun to explore and would have a lot of visual impact. And if we are talking about a festival sort of exhibition, maybe we are taking things apart and moving them frequently enough to justify this approach.

But if we’re talking about nomadic/temporary structures that are easy to fabricate and set examples for other people to follow and use, self-assembling and robotic systems may be overkill. We can save human labor by much simpler means. Technically, there is no reason building comfortable, attractive, and resilient housing isn’t already as easy as assembling office partitions other than the fear of the impact on design from standardization and because it suits the interests of banks for housing to be as expensive as possible. (no banker is going to get rich on housing people can build for themselves in hours or days–which is why it was necessary for the mobile home to become ‘damned’ architecture) It would seem to make more sense to leverage technology and design on high performance structures with the elimination of skill in assembly to leverage modest human labor–the same logic of PC parts. If we’re talking about something on the scale of a small village, we’re not likely building things so big and complex that it really compels automation–as fun as that certainly may be to explore as experiment.

I think we could readily meet the needs of a pop-up village project with some combination of three types of open building systems;

Aa post and beam framing system using structural members light enough for one person to carry but capable of supporting structures of up to three stories and spans of around 6 meters. (which is what Utilihab is designed as)

A pavilion system based on a modular planar/deck truss plug-in ‘backplane’ serving as both floor and roof deck. (what Utilihab is expected to evolve into by second or third generation)

And a high-performance ball-socket-node space frame system supporting both relatively short and long strut lengths for truss and enclosure uses. (alternately, a plate dome system could serve, since primary uses would be for enclosures/skybreaks)

All these would ideally feature tool-less or few-tool assembly and employ pre-finished plug-in/snap-on/bolt-on panel systems. These things would not only meet our project needs, they could revolutionize housing in general–no other high-tech required.

With such easy assembly of basic structure, a nomadic architecture would differ from more permanent architecture largely by choice of materials. Yurts and walled tents can define our basic approach, though these systems afford us much larger and more diverse shapes than the traditional forms. Current architectural fabrics and membranes (ETFE, PTFE) offer vastly greater performance, resilience, and duty life than typical tent materials and can function well even for permanent structures. (I’d avoid anything with PVC, though…) We now have flex-cell PVs that can be integrated with these materials, allowing tension roofs to function as quickly deployable solar power systems. ( http://www.ifaipublications.com/iaa/repository/8/9534/large_2914_369_1280.jpg ) Felts used in living wall systems can also integrate with these materials to make temporary living structures. And we now even have paint-on EL lighting. ( http://www.lumilor.com/ ) Lighter alternatives to alloys for structure could also be used in this mobile context, particularly FRP, wood composites, and the like.

The nomadic context also offers potential for exploring another, more specialized, structural technology; soft architecture. This is something I often discuss in the context of ‘furnitecture’ and as the basis of deployable dwellings in large microgravity space habitat environments. (think somewhat larger versions of Japanese capsule hotel units made out of fabric and foam that plug into open space frame core structures of a large space habitat–space frames which can also double as host for hydroponics, hence the notion of ‘urban tree’ habitats) Basically, it’s soft-sculpture technique applied to the creation of modest semi-rigid shelters using insertable light framing, modular structural foam, or inflatable elements. An easy visualization is to imagine those hut-shaped pet beds made of foam and fabric increased to the size of human-scale furniture or, larger still, a traditional yurt. Such structures offer much more resilient, highly sound and thermal insulated, and easier to setup shelter than traditional tents with many possibilities for novel integral features, integral furniture, and the like. The compromise is that they are potentially bulkier than tents depending on the nature of insert elements. New compressible memory foams offer interesting possibilities for this.

Another technology more specific to the nomadic context is the use of pneumatic frame structures offering some advantages over conventional space frames. These use high-pressure inflatable elements as alternatives to the usual struts of a space frame and can be fashioned to suit ‘gridshell’ systems or, with a larger parts count, more modular conventional space frame geometries. ( http://www.planex-gmbh.de/air-supported-membrane.html ) Unlike more typical inflatable shelters, these systems do not need continuous power for pressurization and with some modern elastomerics pressurized struts can support such high internal pressures they can approximate the strength of wood or aluminum. Pressurized corrugated membranes have long been speculated as an eventual alternative to glass windows. (since materials like ETFE are more transparent than glass) The concept has been applied at small scale to support systems for conventional tents ( http://heimplanet.com/ ) and some off-the-shelf versions of this technology have appeared offering very multi-functional pavilion forms. ( http://www.bec.es/imagenes/estructuras-hinchables/estructura-neumatica-hinchable-pileri/estructura-neumatica-hinchable-pileri-F03.jpg )

Technologies like these seem to offer more near-term possibilities, even if they might not seem as cutting edge as the idea of programmable matter. And, as long as we rely on largely modular design, there are any number of ways we can ‘tag’ large collective structures for spime tracking and the creation of semantic webs around them. That’s basically what a ‘spime’ is, isn’t it? Semantic webs for things and their designs that link to those physical things by the software used to create them (output) and by various means of digital tagging, use and activity logging, or integral networked information gathering.”