Join 3,572 readers in helping fund MetaFilter (Hide)


Print out your own home: WikiHouse
August 23, 2011 8:50 PM   Subscribe

Print out your own home with WikiHouse.

From the early days of computer aided design, the goal has been to not simply draw the building, but to build it in the computer and then drive the machinery to build it for real.

London architectural practice 00:/ has released WikiHouse as a plugin for Google Sketchup. It is an Open Community Construction Set that aims to make it possible for almost anyone, regardless of their formal skills, to freely download and build structures which are affordable and suited to their needs. Once your design is completed, the model is used by WikiHouse to create drawings, which are ready to be CNC milled out of 18mm locally sourced plywood. The pieces can be easily assembled with no power tools, with ribs spaced at 600mm and lateral stabilizers to ensure structural integrity.
posted by honey-barbara (57 comments total) 35 users marked this as a favorite

 
But I only have an A4 printer...
posted by greenhornet at 9:16 PM on August 23, 2011 [9 favorites]


Plywood framing? No.
posted by octothorpe at 9:19 PM on August 23, 2011


Here locally, an 18 mm plywood sheet is 1200 x 2440 mm and runs about 300 NOK. 48x98 mm (~2x4") construction lumber is about 15 NOK / m. So about 10 m of plywood construction (3 ply) from one sheet of plywood for 300, which is double that of regular framing with 2x4".
posted by Harald74 at 9:21 PM on August 23, 2011 [1 favorite]


You wouldn't download a house.
posted by lantius at 9:23 PM on August 23, 2011 [32 favorites]


Harald, maybe you'd save on labour to offset those costs? That said, I was also wondering about all the other costs of plumbing. wiring etc that obviously are add-ons. These are early days in 'digital home' creation, so I guess I applaud their notion of open source construction tools.
posted by honey-barbara at 9:25 PM on August 23, 2011


There's no source at the main site, it's just a solicitation.

It's not clear how this is better than standard construction: it isn't cheaper, the primary tool is more expensive than saws, hammers, nail guns, etc., and assembly time is shortened, but no different than a precut lumber package.

Grumble, grumble. I'm looking for an affordable CNC router service and they seem prohibitively expensive for a humble working lad. So I'll probably end up doing all the cutting with a bandsaw. By hand. In the snow. Uphill both ways.
posted by warbaby at 9:33 PM on August 23, 2011 [7 favorites]


Well, I had been wondering about the possibility of downloading items from the IKEA catalogue in the near future. So yeah, this is neat.
posted by vidur at 9:33 PM on August 23, 2011


Wikihouse is a terrible name.

Wikihouse brings images of a house where 4chan periodically drops by and redecorates.

(This is a very cool idea though).
posted by sien at 9:41 PM on August 23, 2011 [13 favorites]


Hope their design is more usable than their website.
posted by schwa at 9:45 PM on August 23, 2011


I can imagine that some places would want structural calculations. I can see that being a hurdle.
posted by Monday at 9:46 PM on August 23, 2011


An updated variation on the British style interlocking sheds and cabins.
posted by fairmettle at 9:46 PM on August 23, 2011 [1 favorite]


Wikihouse brings images of a house where 4chan periodically drops by and redecorates.

Would you prefer MetaHouse? Every room would feature a niche dedicated to the display of plated beans. The bathroom could be wallpapered in The Treaty of Westphalia. And when an obnoxious house-guest wears out his welcome, the house mod issues a bannin'.
posted by zachlipton at 9:47 PM on August 23, 2011 [7 favorites]


I'm thinking extruded concrete house made with a giant reprap like device might be a niftier idea, maybe something like this. Imagine living in a giant drip castle.
posted by smcameron at 9:50 PM on August 23, 2011 [2 favorites]


Would you prefer MetaHouse?

And get it painted professional white? No, thanks.
posted by vidur at 9:53 PM on August 23, 2011 [6 favorites]


That's not a house, that's a shed. Most folks with a saw and a hammer can whack together a shed out of 2x4s at a fraction of the cost, time and waste.

But,once we can do the same thing with recycled plastic composite molded into 4'x8' sheets, that's something to get excited about. A plastic hut would be very durable, and commercial composite building materials are ridiculously overpriced.
posted by CynicalKnight at 9:54 PM on August 23, 2011 [2 favorites]


I thought of faxing but the paper is a problem at the Elmers stage.
posted by clavdivs at 9:58 PM on August 23, 2011 [1 favorite]


Ooh! Are interlocking sheds like that available in the US?
posted by ocherdraco at 9:59 PM on August 23, 2011


CynicalKnight, I was thinking the same thing. I've been experimenting with 3D printing and I'd love to see recycled plastics re-formed into durable housing.
posted by honey-barbara at 9:59 PM on August 23, 2011 [2 favorites]


This is how we build molds in the boat building industry. And also how many plywood kit boats are constructed.

I guess it's nice to see applied to home construction, but yeah plywood is expensive.
posted by Mei's lost sandal at 10:11 PM on August 23, 2011 [1 favorite]


If you're looking for an affordable CNC service, look on www.cnczone.com - that's where the self-build hobbyists and small business people hang out.

This particular design is about as unornamental as possible, however CNC technology is capable of extremely ornamental designs. The technology has two distinct user groups: engineers, whose designs, like this one, are dull as dishwater, geometric, sparse, and tend to be AutoCAD-based; and artists, whose designs are fantastic (the link shows a laser-built object but CNC would work as well) however are often impractical and extremely slow to actually make.

Honey-barbara, your 3D-printed plastic house will take weeks to print even at 5mm resolution. The better (IMO) approach is to build the "bones" of the house as a stark box with attachment points, and add on the minimum of printed stuff for decoration, packing with resin etc where required.

I would say that Vectric Aspire is the best program in the industry on the artistic side of router technology, and OneCNC is the leader in the engineering side. You can of course use an art tool to do engineering, and vice versa.

If you want to make a CNC router that's good enough to build a house with, budget $5,000 if you know what you're doing or have on-demand access to someone who does. Budget twice that if you just want to buy one that will work for that purpose. Budget at least $20K for the commercial use of one for long enough to build parts for a whole house out of it. That's still a lot cheaper than conventional building methods. This very rough estimate does not include materials.

For this job, you don't actually need ultra-high precision and you could probably get away with making a lot of the router assembly out of wood or even MDF. It doesn't really matter a lot if a house's bits are a couple of mm out, you can deal with that sort of gap with sealant and paint. However, if you anticipate using your router for smaller, more noticable things like furniture, you're best to use at least aluminum, preferably steel, for your router. Weight is a good thing as it keeps it stable. Bigger bed sizes are great; you can always do a little job with a big router, in fact you can do a large number of little jobs at once with a big router. High clearance (Z-axis) is good; if possible build it in such a way that the table surface can be removed, which allows you to machine things beneath the table, such as the tops of columns. A particularly clever design system is to put a rotary system on the outer side of the long axis, and set up your spindle such that it can be re-mounted on the gantry above this rotary system. Thus you replace the short axis motion with the spindle, and are able to machine columns etc. This decision must be made early in the router's build life. :D
posted by aeschenkarnos at 11:18 PM on August 23, 2011 [10 favorites]


Harald, maybe you'd save on labour to offset those costs?

I don't know, but it seems to me on first glance that bolting it all together isn't any quicker than using a nail gun.
posted by Harald74 at 11:32 PM on August 23, 2011


"Computer, 1 house, earl grey, hot"
posted by arcticseal at 11:45 PM on August 23, 2011 [5 favorites]


It depends how you set it up, Harald74. Sticking things together with a nailgun is fine if it's a really simple design that you can see in your head, however all user-assembled objects are basically puzzles, where the objective is to make the puzzle as simple as possible to solve. Where the design is fairly complex to start with, figuring out how to put it together can be quite time-consuming and where it's a big and heavy object, undoing an error can be a major pain.

If you design your parts to slot together, such that slots of particular widths and shapes only fit the converse slots in the correct places with a fairly tight fit, you can make it considerably easier for yourself or whoever is doing the assembly to do so in minimal time. You can also engrave numbers etc to aid in matching part-to-part.
posted by aeschenkarnos at 11:47 PM on August 23, 2011 [2 favorites]


I'm an OK framer but I absolutely love putting together stuff created via CNC router. A whole house that arrived as a kit with super-tight tolerances? Oh, yeah.
posted by maxwelton at 1:28 AM on August 24, 2011


That's a good point, aeschenkarnos. BTW, I was reminded of a short story I hazily recollect where all the construction materials for emergency shelters had a small electronic localizer thingies embedded, and could tell verbally the volunteers or displaced persons putting the shelters together where any particular element fit. Can't remember where I read it, though.
posted by Harald74 at 2:02 AM on August 24, 2011


Sounds like Cory Doctorow's "Makers". I quite liked that one.
posted by Omnomnom at 3:14 AM on August 24, 2011 [1 favorite]


Plastics, both new and recycled, are being used more and more in construction products. The problems with using plastic for structural members are many though. Most plastic is not structurally very strong in one direction or the other, members would have to be big and molded into some kind of structural shape - like an i-beam. Plastic also expands and contracts much more than wood or metal.

It'll come, but it's not right around the corner. (Wood is the ultimate recyclable material anyway, FWIW.)
posted by Benny Andajetz at 3:41 AM on August 24, 2011


If you're interested in this you might also be interested in the RuralZED. This is a prefabricated housing design that is very energy efficient, and sells for UK£89,000 (not inc. labour). The producer is the same architect responsible for at least one of the greenest buildings in the UK.
posted by biffa at 3:51 AM on August 24, 2011 [6 favorites]


Or just get yourself a Huf Haus. They'll build it to your specifications in a factory then bring it to your site and put it up in a couple of days.
posted by The Ultimate Olympian at 3:56 AM on August 24, 2011


gah. no way this is cheaper than standard US-style stick framing. this seems to be a fad among architects, like the shipping container houses: small luxury houses with a "socially responsible" or affordable veneer...
posted by ennui.bz at 4:32 AM on August 24, 2011


You'd want to be particularly careful putting the roof together if you live somewhere wet, or you might find that your WikiHouse WikiLeaks.
posted by ZsigE at 4:50 AM on August 24, 2011 [6 favorites]


"your 3D-printed plastic house will take weeks to print even at 5mm resolution. "

Whereas your normally built house takes MONTHS to build. How is weeks to print not an improvement?
posted by lollusc at 5:01 AM on August 24, 2011


How is weeks to print not an improvement?

That only gets you the frame; you still need a foundation, plumbing, wiring, heating, etc. Even a typical platform construction McMansion can be framed out by a crew in a few weeks. It's all the other stuff that takes time.
posted by octothorpe at 5:28 AM on August 24, 2011 [1 favorite]


As 3D printing gets faster, it will become an improvement; also the ability to print conductive wires, voids for water plumbing, etc will be a total game changer. The technology has huge potential. At the moment though, what we want, in terms of a 2D printing analogy, is like 4200dpi photo quality poster print, and what we have is dot matrix ASCII art.
posted by aeschenkarnos at 6:09 AM on August 24, 2011 [1 favorite]


This sort of thing is the 21st century equivalent of Archigram designs. Clever, nifty, thought-provoking & irrelevant.
posted by aramaic at 6:16 AM on August 24, 2011 [1 favorite]


For us Americans, 18mm is 3/4", or your average floor-decking sheet of plywood. I'd be very worried about using 3/4 plywood -- even in double layers -- to replace 2x4s or 4x4s. Then they drill a lot of holes through the boards, further weakening them. Plywood doesn't have a lot of strength over long free runs as it is. Plywood wants to delaminate as it flexes, and long thin layers are going to fall apart after enough wind. The "frames" they show, the large flattened pentagons, do not look like they'd be able to support much at all. The "cathedral" ceiling they have is a structural fault; there should be an 'attic floor' in there, to help against the outward pressure of the roof weight. It's essentially two light, filmsy flying buttresses with weight added in the wrong places. A flying buttress holds weight 'in', not 'up'. Then add in running electricity, insulation, installing doors and windows, drains for toilets and sinks...this is more like a project for a masters in architecture or civic planning, and not really a practical application of prefabricated homes.

I used to do delivery to construction sites -- and this sort of prefrabrication is already done, and in a slightly better way. You can get an entire house shipped as floor and wall panels, ready for assembly; they're called panelized homes, but they use studs and other modern construction processes to hold the structure up.
posted by AzraelBrown at 6:21 AM on August 24, 2011 [5 favorites]


Still waiting for my Xanadu House.
posted by The Deej at 6:37 AM on August 24, 2011


I want my house FASTER! I can't take BEING HOMELESS for even a few more WEEKS.
posted by jasonsmall at 7:00 AM on August 24, 2011


I'd be very worried about using 3/4 plywood -- even in double layers -- to replace 2x4s or 4x4s.

A doubled 3/4 sheet doesn't get close in compressive strength to an equivalent width 2x lumber, never mind a 4 by. Theoretically, if you put the layers into strong compression, it could -- but you'd need bolts/rivets/whatever every couple of inches, fiberglass wrap, or something of that nature to prevent the delimitation and make sure the layers flex as little as possible, and flex together.

At the very least, I'd be using OSB, not plywood -- but OSB has a strength axis, and you'd have to be aware of that when designing the panel cuts *and* loading the router.

I'm even less impressed by the speed to build that one truss -- a small crew can frame out a whole shed in two-four hours, starting with stock dimensional lumber, these guys made one truss unit. Never mind the time for the CNC router to cut the panels -- which is going to be part of your build time, and time is money, even if you're doing it yourself. The time you're building is the time you're not earning money. That may be a good tradeoff, but it *is* a tradeoff. If you lose three months income because you're building this, that's a real cost.

For fast, you need big, powerful and stable -- three things that portable gear is really bad at. This is why the various prefab systems have the big powerful stable machines at a central factory, and ship prefab elements to the site.

I just don't the the idea of putting prototyping systems into use as production machines. Prototypes are designed to be flexible, which makes them inefficient. They're brilliant -- I'm glad I live in a world with them. But if I need to make a couple of million parts, I need to be making my parts so they can be made by simpler, faster machines -- or the cost is going to be outrageously high.

Think of it this way. Materials cost what they cost, be they wood, steel or plastic. After that, the cost to a make a part is directly correlated to the time it takes to turn that chunk of raw material into that part. The more time it spends on the machine, the more expensive it becomes.
posted by eriko at 7:04 AM on August 24, 2011 [2 favorites]


I think it's worth exploring the idea of building houses like this, even if it's wildly impractical now. I think the ultimate goal is always to have a completely self-building house where you just kind of throw the materials on the lot and come back a few days later to a finished house.

But baby steps.
posted by empath at 7:12 AM on August 24, 2011 [2 favorites]


A more impressive use would be a jig to align lumber into frames and frames into structure, so only two people could frame a house quickly. Or they could apply this to corrugated plastic, using a few interlocking layers.

Plywood is expensive, doesn't do well with load bearing, and the process is wasteful.
posted by Slap*Happy at 7:17 AM on August 24, 2011


The idea of using a CNC machine for home construction is Ionesco-level absurd. Framing is easy, people. Measure and cut. Repeat. And you don't need millimeter level accuracy because guess what. Houses settle. They move. They are more like giant articulated linkages than rigid objects.

People have been framing houses the way they do today since before power tools. The hard part about building houses is not the wood framing, it's the masonry, plumbing, electrical, HVAC, and carpentry (mouldings, flooring, etc.).

The easiest way to speed housing construction is to get more people. This is why the Amish can put up a 2-story barn in a day--without power tools.
posted by Pastabagel at 7:29 AM on August 24, 2011 [6 favorites]


Great. A new way for Americans who do manual labor to have their industry re-organize their jobs out of existence.
posted by clockzero at 7:31 AM on August 24, 2011 [1 favorite]


So you just open your box, assemble the plywood framing, and presto, you have a house!! Nothing left to do but some quick plumbing, electrical, heating, masonry, appliances, windows, doors, flooring, painting, roofing, tiling, weatherizing, insulation, lighting, trim, and paving/landscaping!
posted by brain_drain at 7:34 AM on August 24, 2011 [8 favorites]


Damn you Pastabagel
posted by brain_drain at 7:34 AM on August 24, 2011


Back in the early 1900s there were companies like Aladdin Homes that shipped "Build in a Day" kits that included pre-cut framing lumber, stairs, interior trim, shingles, etc. They offered tons of different designs. You can order copies of their catalogs from Dover or download one here.
posted by bonobothegreat at 7:39 AM on August 24, 2011


For cost and simplicity, I like prefab housing. Have factories spit out cubes you could ship anywhere, lift into place with a crane, and bolt together to make buildings that would pass inspection. Make the plumbing and wiring as plug-and-play as possible, but leave conduits for threading more stuff in if needed.

Toilet and shower unit, bedroom unit, kitchen unit, recreation unit, hallway and stairway units, etc. Start with a kitchen-bathroom-bedroom triple, then save up to buy extra units later. Make them all the same size and with the same plug arrangements so you can get modules from different manufacturers. When you decide you want to sell, you could sell the whole thing in place or sell off bits to different people to cart away. And make toy kits like doll houses that would let people experiment with designs before ordering real units.
posted by pracowity at 7:52 AM on August 24, 2011 [1 favorite]


Yeah pracowity, there's a 'democratising of shelter' angle that I'm seeing. I know it's a long way off in actuality, but I think getting people more involved in/thinking in detail about the design and DIY construction of their home is a good thing.

The Ultimate Olympian: I LOVE Huf Haus too. Those guys NAIL [sorry] every detail. So inspirational!
posted by honey-barbara at 8:05 AM on August 24, 2011


I have a friend who has one of these Shopbots, which he used to build this

Those flame-ribs, which are each about 2 stories tall, were made of sections cut from plywood on the shopbot (as are all the decorative panels—the one with the repeating circular motif is one I designed); they're three layers thick, with the sections staggered to avoid weak spots at the breaks; the layers were glued and screwed together.

While I think it was a wonderful application of CNC router technology, and while I think Wikihouse is a neat idea in principle, it doesn't strike me as a really viable application of the technology. I think that structural insulated panels are a much more useful Building Method of the Future.
posted by adamrice at 8:09 AM on August 24, 2011


At the very least, I'd be using OSB, not plywood -- but OSB has a strength axis, and you'd have to be aware of that when designing the panel cuts *and* loading the router.

OSB is "oriented strand board". All the wood strands are glued together in one direction and it definitely has different capabilities depending on which axis the strands point.

Plywood has no "grain". Each ply is perpendicular to the next, making it extremely strong and rigid in either direction. I wouldn't have any reservations using plywood for framing. I'm not convinced it would be effective cost- or timewisd, though.
posted by Benny Andajetz at 8:12 AM on August 24, 2011


One other thought. The strength of the framing is a small part of the strength of the structure. Even a standard stick-built wall is pretty flimsy by itself, and very prone to racking. Attaching the walls to each other adds strength. The sheathing is actually what serves the purpose of actually tying everything together and giving it the most strength.

And buttresses don't hold weight "in", they transfer loads outward and down, giving the structure a bigger footprint and more points to distribute load.
posted by Benny Andajetz at 8:27 AM on August 24, 2011


I printed my WikiHouse from barley rice. CNC walls and googly eyes. Tables of paper wood, bricks out of shite. And everything empty as shadows at night.
posted by It's Raining Florence Henderson at 9:41 AM on August 24, 2011


Structural strength? Bah, burn it down every night and build a new one in the morning!
posted by blue_beetle at 9:59 AM on August 24, 2011 [1 favorite]


The big drawback I can see besides the cost is if you need to get the structure inspected you'll need to get an engineer to stamp it and that' going to be expensive. Now if the cutting companies were willing to stamp the designs the way truss companies do this system might have a chance.

And if it was me I wouldn't be using wing nuts. Battery operated impact drivers are cheap and you could drive hundreds of nuts on a single charge.

warbaby writes "It's not clear how this is better than standard construction: it isn't cheaper, the primary tool is more expensive than saws, hammers, nail guns, etc., and assembly time is shortened, but no different than a precut lumber package."

Believe it or not framing is a skilled trade even if your only tool is a hammer. In theory anyways you could make this essentially idiot proof by only allowing the pieces to fit together one way.

ennui.bz writes "no way this is cheaper than standard US-style stick framing."

One immediate advantage that I can see is this allows for super insulated wall systems for essentially the same assembly cost. Super insulated stick framing is significantly to wildly more expensive labour wise depending on the method you choose. Plus you'd have less low insulating wood and more insulation than in the equivalent stick framed wall. Plus it would allow people in remote areas to assemble their house themselves. And the framing waste is at the cutting plant where it is easily re purposed.

eriko writes "A doubled 3/4 sheet doesn't get close in compressive strength to an equivalent width 2x lumber, never mind a 4 by. Theoretically, if you put the layers into strong compression, it could -- but you'd need bolts/rivets/whatever every couple of inches, fiberglass wrap, or something of that nature to prevent the delimitation and make sure the layers flex as little as possible, and flex together. "

The replacement studs wouldn't be all that strong compared to a 2X4 but attached to 3/4" siding you'd end up with an impressive stressed skin especially if the sheathing was dadoed to interlock with the studs. Those studs would keep the sheathing from buckling and a sheet of 3/4" plywood on edge has more than 9 2X4s worth of cross sectional area. See for example TJI joists where 1/2" OSB kept from buckling by 2X3s at each edge are easily twice as stiff as the same depth of 2X material.

Pastabagel writes "The idea of using a CNC machine for home construction is Ionesco-level absurd."

It's already a widely deployed technique. I'd guess at least half the log homes constructed nowadays are in whole or in part cut out by CNC equipment.

Pastabagel writes "People have been framing houses the way they do today since before power tools. The hard part about building houses is not the wood framing, it's the masonry, plumbing, electrical, HVAC, and carpentry (mouldings, flooring, etc.)."

The framing is actually hard; at least to get right. With the wide spread adoption of pex supply and ABS drain piping I'd posit that getting the framing right is harder than getting the plumbing right for anyone who is familiar with both. Framing just has so much redundancy baked in that you can get it wrong with out noticing.
posted by Mitheral at 10:55 AM on August 24, 2011 [1 favorite]


It's probably worth noting, too, that thinner framing and engineered joists like TJIs scare the shit out of firefighters. A TJI is crazy strong, but if the web burns through they collapse fast.
posted by Benny Andajetz at 11:18 AM on August 24, 2011


You beat me to it, Benny. Various modern construction techniques such as roof trusses and engineered lumber become commercially successful because they trim actual material mass requirements to a minimum—but what appears to the builder as excess is actually the margin of structural stability that make it possible to perform interior firefighting and search/rescue.

The energy in the average live fuel load has gone up as the modern home includes a lot of plastic. Why not make the structural members out of plastic too? We could be treated to even faster structural failure and even hotter temperatures than we have today.

Stepping back 10 paces, the issue of what do we give up in safety to gain something in fabrication seems linked with the desire expressed upthread to replace inputs of skilled labor with fancy technology (and some different skilled labor inputs) and call that progress. We're so sure there's a magic process that will let us pass go and collect $200. This kind of thinking is a disease, an irrational aversion to recognizing limits. So what if it takes a lot of time and money to build a house? What's wrong with that? Why the built-in assumption that we must always be making housing cheaper and faster?

I'm not anti-innovation, but i'm not intrinsically pro-innovation either. This particular innovation doesn't look like it's faring too well as we add in the various levels of implementation-related constraints. And we haven't even started on a comparison of the total EROEI for this versus traditional stick building.
posted by maniabug at 12:36 PM on August 24, 2011 [2 favorites]


Here's Endless Forms - Cornell University on 3D printing. [There's also coverage of these developments at M.I.T.'s Technology Review] It's early days, but so exciting!
posted by honey-barbara at 10:34 PM on August 26, 2011


« Older [Harold] Brodkey produced fiction that was epic to...   |   Tabletop: Analog Game Design -... Newer »


This thread has been archived and is closed to new comments