Detailed discussion by Eric Hunting:
“The ideal situation for P2P architecture is where you can produce structures of small to large scale using intuitively simple modular systems with components on a human scale that are easy for the solitary individual to manipulate and which encode aspects of safety and structural engineering into their interface standards in the same way that the sub-components in a personal computer encode lower levels of engineering into them so that assembly and design higher up the food chain doesn’t have to think much about them. This is the province of true plug-in architecture systems. We’re not quite there yet technologically but there are some things pointing in the direction.
I’ve been talking a lot in recent years about a system I’ve dubbed Utilihab, which is a generic name I’ve given to a series of building systems based on aluminum T-slot framing as used in industrial automation and which go by such names as Tomahouse/Tomatech, Jeriko House, iT House, Kit Haus, etc. I’ve personally worked with Tomahouse and Jeriko House.
You can see these here;
The designs take their cues from classic Modernists but the structural system derives from Japanese/Polynesian pavilion architecture such as traditional Japanese homes and more recent derivatives like the modular resort cottage homes of Tony Gwilliam’s Bali-T houses, based on traditional Indonesian ironwood post&beam framing with a synthesis of Asian and Modernist-Minimalist design.
The concept of aluminum profile based modular housing may actually have its origins with profile based housing developed in the late 1940s by Jacque Fresco -who some may recall as the itinerant inventor and futurist founder of the Venus Project. This was very radical for the aluminum industry of the time, with extruders few and timid about pushing their equipment to the limits -something American extruders seem to have never gotten over to this day. Alas, he was also never a particularly smart businessman and came from an era long prior to the notion of open source…
The current forms of this building system tend to be based on 240mm-250mm square aluminum primary frame profiles with two or four T-slot channels per face and secondary and floor/ceiling grid framing of 120x240mm and 120mm square -usually based on newer and larger standardized profiles (60mm series) and sometime with custom profiles, though there’s no particular advantage to that over the standard industrial profiles except for profiles with integral thermal breaks. (used when designers want the framing to be visible externally as an architectural detail) Some companies still have compulsively old-fashioned thinking and so try to employ custom profiles as a gimmick to lock-in market share -which is sort of silly given that, currently, there is no established ‘market’ to divvy up! Trusses are also possible for much larger unit spans and particularly convenient using ready-made ‘web plates’ that slip and lock between a pair of profiles to link them into a truss beam. Buildings are composed of simple post&beam structures with a typical 4 meter square unit module supporting 1 meter floor/ceiling grids (actually, a meter plus interstitial frame thickness) and then finished in various forms of non-load-bearing pre-finished retrofit paneling, concealment strips, and planking. Light panels, usually about a meter wide, use pop-in or friction mount attachment while exterior walls use bolt or other mechanical attachment and have interstitial volume for insulation, where not using some for of SIP. Some fixtures are designed to attach to the framing slots in the same way accessory components are added to industrial automation structures and with the hollow channels within the framing sometimes serving extra duty as utilities conduits. Typical structures are one or two storeys high, but the systems have been rated for ten storey structures -usually with the addition of things like cross-brace tension cables and gusset plates.
Though some developers have patented special profiles and connector designs only they use, the T-slot framing technology itself is largely public domain. In fact, none of the manufacturers of T-slot parts for industrial automation seem to have any clear picture of it’s origins, which may go back to some time very early in the 20th century -most-likely the 1920s when things like ball socket space frames appeared, though T-slot attachment schemes with machine tools and optical benches go back well into the 19th century. Companies sort of sprang-up simultaneously around the globe and started making this for industrial automation use in the late 1970s, with all of them thinking/claiming they had invented it then suddenly finding everyone else’s prior art. It was almost like Sheldrake’s Morphic Field effect.
This ‘frame and panel’ system -along with similar space frame variants like Universal Node System/Min-A-Max- is one of a couple basic forms of small component plug-in architecture currently being developed; the other key one being ‘planar backplane’ systems as explored in Shigeru Ban’s Furniture House series. Here the structure of homes consist of largely independent floor and ceiling plane elements composed of a modular planar grid. These are then supported by strong modular furniture units serving triple-duty as load-bearing structural support, partitions, and active furnishing elements. These planar backplane elements become very similar in character to the motherboards of backplanes of computers, determining the attachment grid for the other components and integrating most infrastructure and climate control systems. This is a potentially more advanced system because of the potential to integrate a great deal of technology within the structural elements while being intuitively simple for the user, though more limited in overall structural shapes. Users can basically design on-the-fly by the simple arrangement of these modular furniture elements, keeping within simple limits of span, cantilever, and vertical load communication according to the planar grid. This set of structural rules is easily automated by having the backplane of the house made digitally active, a ‘structural integrity network’ basically live-modeling the house by identifying the parts plugged into it and graphically communicating to users where their changes approach safety limits and showing them when parts are broken, worn, or stressed. Frame and panel systems like Utilihab are likely to evolve toward or be superseded by planar backplane systems in the future, based on these advantages. However, that form of technology is much less developed and, to be fully demountable and freely adaptive (which Shigeru Ban’s designs aren’t as yet), requires much more sophisticated component interfacing and integral structural intelligence.
Though these technologies aren’t ready for immediate large projects like relief efforts (the companies I noted all are still stuck making prefab luxury homes -even though most have relief/low-cost housing aspirations of one kind or another), they’re ready for discrete housing and experimental use. (I plan to use these for my own home in the near future) Their chief functional limitation is roofing technology. We still haven’t devised a practical means of fully demountable small component modular roofing that’s freely extensible/variable in two directions. There’s a lot of room here for innovation and it’s accessible given that you can base these on the off-the-shelf components of the industrial automation industry. But it is pretty sophisticated and needs comprehensive fabrication facilities and engineering to push the envelope.
What’s really significant about these technologies is that, in order to develop them effectively, one has to employ a completely different industrial model than has been common to the Industrial Age. They need ‘industrial ecologies’ like that of the computer industry, where your products represent open ‘platforms’ based on p2p defined standards supported by ecologies of competitive product developers in a food chain of sub-component integration. This is the world-changing part. This is what made the computer industry different from all other previous forms of industrial development and is why this most complicated of all artifacts ever devised has enjoyed such an unprecedented pace of evolution, so very quickly transforming from rare 100 million dollar behemoths to small, ubiquitous, and cheap enough that some homeless people can afford them and simple enough that a child can assemble them from parts made around the world and it will work perfectly the first time it’s turned on.
There is much that can be done on the level of simpler and even lighter technology, though the application to permanent housing gets trickier the less physically robust structures are. From a nomadic architecture standpoint, it’s hard to improve on the accumulated experience embodied in traditional technologies of nomadic cultures. There’s a sort of singular refined perfection in the yurt, tipi, lavvu, bedouin tent, Romany caravan, etc. that only -and rarely- gets improved upon with new materials and fastener technology. (tension structure materials like ETFE and teflon impregnated fiberglass cloth like Sheerfill, new cable materials, and fasteners like Grip-Clips) However, traditional nomadic architecture evolved in a context of large open spaces and extremely light lifestyles of minimal personal possessions (though traditional Tuvan furniture -albeit designed to be portable- is, thanks to Chinese influence, often far from what might be called ‘light’) So these structures don’t integrate well into an urban environment. Proponents of the original Urban Nomad movement, such as Ken Isaacs, sought first to focus on the re-appropriation and adaptive reuse of found urban space -primarily indoors. This was the premise behind Isaacs’ ‘Living Structures’; the simple wood framed multi-function ‘furnitecture’ that produced the Box Beam and now Grid Beam building systems. Relying on another larger enclosure structure to provide the more basic environmental shelter, one can employ much more freely adaptive structures within them and use very simple modular component systems.
I recently proposed a p2p architecture experiment based on this called Vivarium whose premise was to repurpose a generic commercial/industrial urban space as a community-evolved recreation facility based on the notion recreational architecture -building structures as group play- using Grid Beam and Living Structures. One would simply repurpose a functionally generic space, like a warehouse, into a recreational space based on the participants’ individual and collective notions of fun, pleasure, and comfort as realized in structures they alternately individually or collectively build and combine, through negotiation, into the space. So in this context the Grid Beam system is being used like an adult Tinker Toy system with which to spontaneously build furnishings and structures. This seemed a fun and ‘low stakes’ setting in which to explore p2p community design concepts.
Many kinds of small light ‘furnitecture’ are possible and suited to this sort of sheltered generic space environment. The Japanese Capsule Hotel unit is a good model here that can be explored in many variations. I think it was Archigram that, in the 60s, experimented with ‘pod living’ based on the idea of using generic open dwelling space to host functional rooms in the form of enclosed appliance-like furniture unit pods that could be freely moved about. (an idea reinvented today with the rooms of Shigeru Ban’s Naked House, which I’ll link to shortly) Such pods are possible with many different materials, such as T-slot, light space frames, and rigid composite shell units and fabric covered structural foam that have the option to be used outdoors. I once considered the idea of making a slightly larger form of Japanese Capsule Hotel unit into a rigid composite shell microcabin, complete with solar power, communications, and other gear built-in, that could be pulled on bicycle wheels and provide a highly insulated durable alternative to tents at a lower cost than trailers. Though bulkier than tents when moved, fabric covered structural foam cabin pods would also be suited to this application. N55 explored a similar idea called Snail Shell System based on repurposing a rotomolded HDPE tank. Andrea Zittel explored this concept in the form of a stationary variant of the traditional ‘teardrop’ trailer called the Wagon Station.
Wanting to move beyond the limits of the found urban space, Isaacs also explored the microcabin/microhouse concept based on stressed skin plywood structures. An interesting aspect of this in the p2p context is that he devised the use of these with external multi-level frame structures based on modular pipe fittings like Kee Klamp that would externally support complexes of these microhouses up to several storeys high. The intention was to use these to host small constantly evolving villages. You can visualize these as open scaffold-like structures sprawling volumetrically which the microhouses could rest in, each module fitting within the unit cubic grid space. These would include decking, walkways, screens, and simple corrugated metal roofing in different areas, all of it freely demountable. Problem was that Isaacs never found a good way to waterproof his microhouse designs and the plywood of the time was pretty crude, giving these microhouses a short life span. This was similar to an early plug-in architecture concept based on large space frame structures that complexes of pod-like room modules would be suspended within. Sometimes proposed for megastructure architecture, one of the most interesting forms of this was based on webs of tension cables suspended between canyon walls into which whole cities might be retrofit as complexes of suspended pods and decks. More recently, this same concept re-emerged with the Shimizu Try 2004 megacity concept where such pods were taken to the scale of whole skyscrapers suspended within a space frame pyramid. Rather over-the-top, but you can easily imagine how this works at a more human scale.
This concept of adaptive structures sheltered by larger structures extends to more permanent -or should we say ‘continuous’- habitats based on purposely built ‘skybreaks’. The term ‘skybreak’ originates with students of Buckminster Fuller who proposed the concept as the ideal approach to the use of the geodesic dome for housing. Typical dome homes are based on using a dome in much the same way one uses the basic frame of a house, which is then partitioned into rooms. This has never worked particularly well with the dome shape. Fuller’s student’s realized that a more appropriate role for the dome was as a large area shelter against the extremes of the basic elements -a largely independent barrier or shield against the rain, wind, and greater temperature extremes. Thus one would shelter one’s whole ‘lot’ space with a dome that could be open on the perimeter in warm weather and inside which one would cultivate a garden environment and build light freely-changed structures to actually inhabit and provide the functions of different rooms. Essentially, it’s like using a greenhouse to shelter another lighter house. Decoupled from the function of weatherproofing, these light internal structures would use materials not normally practical and could get away with modular systems that were easy to owner-assemble but not yet sophisticated enough to be weatherproofed. The catch with the idea was that it wasn’t until close to Fuller’s own death that silicone sealed planar glazing systems and ETFE based ‘pillow panel’ systems were devised to make the transparent skybreak a practical concept. (Fuller’s earlier attempts with translucent plastic sometimes failed dramatically)
This isn’t actually that radical or new a concept. It originates with the pavilion architecture of Asia and Polynesia where traditional housing never used load-bearing walls but instead relied on clear-span post and beam structures that created functionally generic space made functional by mobile furnishings and light, sometimes free-standing, screens and partitions. This later inspired the Modernists, the concept made iconic by Philip Johnson’s New Canaan CT Glass House then reapplied in thousands of variations. I personally consider the pavilion the most practical form of housing in the contemporary cultural context, particularly for non-toxic home applications, though it doesn’t suit the conventional western suburban environment where you have no use of walled enclosures as in Asia (because western suburbs are about the public display of social status and affluence, not about living well…) and where you lack the space to articulate landscape for sake of privacy.
Today we have a huge variety of structural types -many prefab- that can be used to make skybreaks of most any scale or shape. Tension roofs and pneumatic structures in particular are promising for large scales, though there hasn’t been much experimentation with this. The potential of the concept well demonstrated, in small scale, by such things as Shigeru Ban’s Naked House where a translucent-walled clear span enclosure shelters a series of traditional Japanese rooms contained within boxes on casters that can be freely moved about. A similar concept I’ve often considered trying would use a translucent dome or hypoid/conic tension roof transported inside one of several 20′ containers on casters or short legs which would be used as simple rooms in a ‘compound’ home once the skybreak was deployed over them.
One of the most intriguing skybreak concepts I’ve come up with employs the strategy for large urban microgravity habitats on orbit. Called EvoHab, the concept derives from the Transhab -a pneumatic-hulled space station module once intended for the ISS but now the focus of the Bigelow space tourism projects. The Transhab employs a tough but flexible foam-filled skin as a pressure hull, supported by a rigid core truss around which functional elements are radially organized. Small scale, these things don’t seem very different from a typical NASA-style space station module, organized into circular decks. But I’ve anticipated this technology would produce progressively larger habitat structures, eventually trading the pre-made flexible hulls with composite hulls built on-orbit by combining a supporting frame with external shield panels and internal pressure hull panels sealed with plastic materials. In this way the same basic structure using this same core-truss radial organization would continue to grow in scale until you could enclose an entire small city within large spherical and cylindrical hull shapes. These hulls would be made ‘light transmitting’ by using externally mounted holographic membrane heliostats and internal light emitter panels linked by thin optical fibers or light pipes. These would eventually become image-corrected, making the hull virtually transparent from the inside. This vast spherical skybreak would then house an ‘urban tree habitat’ based on using the large core truss as primary attachment for radially mounted equipment and dwellings made from modular building systems. This deployable orbital housing would, in smaller forms, be akin to Japanese Capsule Hotel units but could evolve into large multi-chambered pods and clusters made primarily of semi-rigid structural foam, fabrics, and light alloy frames. The largest dwellings and work spaces could be based on stacked planes outfit by furnishings that attach between them -the same concept as the planar backplane plug-in architecture but here adapted to microgravity.
There are a lot of potential building concepts to explore as the basis of p2p architecture, depending on the scale, location, and at-hand fabrication capability. We could probably devise something for just about any situation from the at-hand technology -even if some of it remains a bit primitive.”