WikiHouse – Open source buildings and interiors for self-build

https://www.opendesk.cc/

I think we'll be seeing a lot of things like this in the next few years.

The WikiHouse project began as just an experiment into digital, distributed housebuilding, but has turned into a serious open source project aiming to change the way building information works. We’re doing it the hard (and slow) way – so currently busy working through certification, regulation, liability, commercialisation etc. It’s already in use in various pilot projects by a bunch of teams around the world already, but as you can imagine, change happens slowly in this space.

In a nutshell, the advantage of digital fabrication is similar to any fabricated components (think IKEA). It brings speed, value, precision, performance-predictability, safety and mistake-proofing (which lowers skill and risk thresholds for small players) to construction projects, plus it’s editable as code. The opportunity we see is to push this from big centralised factories to distributed networks. The lead technology, WikiHouse WREN, is just intended as a starter for ten really. We started designing for machines and materials that are pretty widely available (Plywood, CNC), and then invited others to fork and edit, which is what’s happening. It is strong, buildable, airtight, performs well on cost etc, but yes, at the moment, it is limited in terms of span, height... Please do fork and improve it, or get into developing other open building technologies.

As a couple of comments mentioned, the key issues are around the limitations of sharing static design files, because, yes, price, conditions, regulations etc vary depending on where you are, and yes, there is a legal risk of any specific design shared being interpreted as professionally certified, and therefore exposing professional contributors to risk. So, over the last few months we’ve been working on some browser-based generative software for designing and delivering homes digitally without relying on external proprietary software like Rhino and Sketchup, with support for distributed supply chains and with checking protocols built-in. We’re hoping we’ll have some stuff to release before too long for people to get their teeth into, but as you can imagine, we have to moderate our excitement with a degree of sensible caution and thorough testing. But as a rule, we’ll be open-sourcing everything we can, as soon as we can.

We don't really get tornado's or earthquakes however. we do get winter gales regularly.

There is a huge need for cheaper, and more efficient (in terms of energy needed to make the materials) homes here. Due to aesthetic requirements (and as a traditional way of making a house that will survive the damp in the UK), most homes are still built with fired-clay brick skins forming a cavity wall, although the inner masonry skin is often a prefab timber frame these days. The roofs are usually clay tiles (or concrete copies if local planners allow) or natural slate. The foundations are a solid trench of 1 metre deep concrete (more if the ground is soft). Ground floors are usually thick concrete onto thick hardcore bases with insulation incorporated either in the concrete or under the floor covering. The whole thing is massive, heavy, concrete and clay focused and expensive. Of course land prices are high in the UK so the resultant houses are really expensive no matter what. There must be a more efficient way of building.

And by that I mean, you have spent a good deal of thought on a pretty marketing page, but your spiel doesn't really demonstrate knowledge of the wooddust and nails that make up homebuilding. You're not being careful with which codes you satisfy, you're not talking physical tolerances (earthquake? hurricane? heavy rain? wind?), you're not discussing the existing prefab systems and what makes yours better.

There is, of course, potential in prefab. But I would suggest more sweat of talking to carpenters and other tradespeople. Hiring a top carpenter for 6 months to throw darts at the thing would be useful. willyt provides some good feedback, but I think its revealing that you don't treat that kind of information, in that kind of depth, in your current information copy. You can get there though!

best of luck.

https://www.kickstarter.com/projects/622508883/open-building...

http://opensourceecology.org/wiki/Main_Page

Especially if one can source local CnC/machine shops to send material to be pre-cut then delivered on site.

I do see 3D printing eventually being a more efficient way to build spec houses.

The thing that I noticed is the material waste is pretty high. Getting things fit to 4'x8' sheets of plywood in the US has lots of waste in the sheets. When you stick build a house, most of the materials are used, with a little careful thought things work out. With the long arches, only a few fit in the sheets.

I wish the basic issues could get fixed.

Structure. What is the advantage of all the machining that is being done here? It looks like there is not enough structure? I'm not convinced that the clip together frame screwed together on site from 18mm ply will form a convincing portal frame? That stuff costs about £40/sheet what happens to all the wastage from the machining? Once these are assembled and filled with insulation they will be pretty heavy to lift into position without a crane which is a safety/legal issue, so if we need a crane, what is the advantage over other cassette systems like dan-wood, hof house, cross laminated timber or even just prefabricated stick frame panels? A ply/foam sandwich could make for a pretty rigid structure (think surfboard), but the insulation is loose filled here; opportunity missed! I note that Arup are listed as partners in the project, they should be able to advise on this.

Design. Its a massive amount of effort expended on developing a new construction method to create a 40m2 one bedroom single storey house like many others before. Why is it better to build like this? What is the maximum structural span of this machined plywood slot together construction? What advantages does it give me if I want to have walls at angles other than 90 degrees or rooms wider that 4m? How high can I go? It seems like you have chosen eco friendly materials but single storey construction is fundamentally not environmentally friendly because the density is too low to support public transport and efficient public services provision.

Energy. They are proposing a Mechanical Heat Recovery unit and very little heating which means they need airtight construction, but the airtightness membrane is very close to the inside of the house and will be easily damaged when the electrician and plumber fit the services. Airtightness is critical. A more robust design would rely on a taped board as the airtightness layer, preferably somewhere in the construction where it won't be damaged by someone putting a picture up. The integrity of the vapour check layer is also very important as any significant interstitial condensation will degrade the plywood structure quite quickly. Loose fill insulation is notorious for slumping and leaving cold spots that go mouldy.

Cost. Double it. E.g. £1000 labour allowed for the foundations installed. That pays for 4-5 days of a 1 skilled labourer's time, it's not enough. There are no plant hire costs in there either. Also only £1500 allowed for the electrician!

Legal. Who gets sued if this design goes wrong? Who pays for the insurance? You have a disclaimer, but if any of the designers contributing to this are architects, engineers or other construction professionals the construction case law in the UK suggests that a professional can't blanket disclaim all advice given in their professional capacity as potentially incorrect. So any professional contributing to this could still be held responsible for design failures. It's rather terrifying what has been blamed on architects giving free advice off hand. I don't know if the creators of this are architects, but bear in mind that Architects in the UK are required by law to have adequate insurance to cover the full cost of any mistakes they may have made for undefined period after the building is completed I don't remember the exact figures but you are generally advised to carry insurance at least a decade into your retirement. Hence no professional will contribute to this website for free unless you find a way of guaranteeing that they are insured as otherwise it will have an unknown ongoing cost to them into the future.

The reason can be condensed in what pnathan already said in his comment: "too much designer, not enough carpenter".

This is nothing new, this project is at least 3-4 years old and I clearly remember donating a small amount to "the cause".

Yet the number of full designs are pretty much the same as when I donated.

What this project is lacking, imho, is a full-blown project, as in "from plywood to finished/furnished/habitable house".

I ultimately came up with a couple of ideas in this regard; one seemed potentially doable, and the other seemed absolutely doable. Both rely on the idea of the main building material being plywood or OSB - like 1/2 - 3/4 inch thick 4x8 sheets (or the metric equivalent).

The first concept was something that a "do-it-yourselfer" could own or rent. Essentially it was to be a plotter you would attach (in some manner) to the sheet of plywood, and it would "plot" the outlines of the cuts and such on the wood (at the time, I envisioned an ink-jet like approach - today a small LED laser engraver might work better), with annotations and other such things to guide the builder in fabrication and assembly. These annotations could also refer to areas in a reference assembly diagram or software that could take the user step-by-step in cutting out the pieces (using various tools - a jigsaw and circular saw mainly) and assembling them (perhaps with 3D views, audio, graphics, etc). The plywood pieces would also be numbered or otherwise noted so that each is used in the proper order. The whole thing could be sold as a simple robot, perhaps attached with cords or something to the wood, in a case with software - user supplies a laptop (or perhaps it comes with it?).

This machine actually does exist - in a way nowadays. I know of a CNC router that attaches directly to a piece of plywood, which could cut the parts directly; however they wouldn't be numbered (but perhaps they could be engraved?). There's also another tool that is something like a "user guided" CNC router; the user holds the tool, and guides it based on an on-board reference (I'm not sure, but I think it can also help to auto-correct problems). Ultimately, both of these machines come really close to the first concept I had. It's nice to know someone had the same ideas and took it further - but that's to be expected, I guess.

The second concept I had was something that I could only see as either an owned piece of equipment, or as a rental/lease system. Essentially, this would be a container (ideally a standard shipping container). One end, you'd load an "elevator" of plywood sheets (like a printer of sorts), the pieces would go inside the "magic box", and would be spit out the other end (again, stacked) as mostly pre-cut pieces of plywood containing the parts of the building.

Think of these pieces as being similar in scope as those "pop-out" wood puzzles of dinosaurs and such you can buy as toys, just larger scale. The cuts made by the machine wouldn't be "complete", but would have to be completed by the user using a jigsaw or handsaw. This would also allow shipping pre-cut pieces (which could be made in a factory) to a building site. Perhaps these could be "ordered online". But the original concept was that the container box would be dropped off at the on-site building location, with a load of "blank" pieces of plywood/OSB.

Inside the box would be a simple system of conveyors and other plywood material handling devices that would move the pieces from the input stack, to a cutting machine, then to the output stack. The cutting machine could be either a CNC router (like a 4x8 shopbot or similar) or some kind of large (2-300 watt?) laser cutter. The latter might be preferable, because it could also engrave assembly annotations, etc on the parts.

The downside of this machine (actually both ideas have similar downsides) would be the fact that you'd need to monitor and do tool bit changes as they wear - or in the case of a laser cutter/engraver - monitor and prevent fires from starting. Which would seem to make this kind of a system more suited to a "factory floor" installation rather than an on-site system rented by a diy'er or a contracted builder. Then you'd just go online, select a floorplan/style, make certain mods as available, and order. Completed sheets of the parts would be delivered to your building site (perhaps with all needed fasteners and such).

Lastly - I'd be willing to bet that such a system (and the "WikiHouse" designs) - at least in the United States - will suffer from the same issues that plague people who want to use other alternative construction techniques (aka - rammed earth, straw bales, earthship, etc): That of the hassle of getting permits approved for the construction, then obtaining homeowners insurance for the resulting structure. In some areas, both can be so onerous as to make such construction impractical, more expensive, or impossible...