A Printer That Can Print a 2,500 Square Foot House in 20 Hours • Hacker News Board

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#31

Post by jared314 » Mon, Oct 28, 2013, 7:02 PM UTC

ohwp wrote: post #19

oilytheotter wrote: post #12

Earlier quoted context omitted.

Did you see the video and pictures of the printer in the video? It wasn't full scale, but it wasn't far off either.

Yes I've seen the video's (a year ago). That's why I wonder why they didn't built a house already. They have the knowledge and tech. Some speculation: maybe it's very hard to print concrete out in the open with all kinds of weather messing up the print. Or companies just don't believe in the benefits so investors are hard to find.

I remember reading that D-Shape[1] had clogging problems with their printer. Perhaps CC has the same issue.

[1] http://www.d-shape.com/tecnologia.htm

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#32

Post by moocowduckquack » Mon, Oct 28, 2013, 7:03 PM UTC

JoeAltmaier wrote: post #15

oilytheotter wrote: post #14

Earlier quoted context omitted.

In the video, he proposes that the printer would have a second pick and place type arm that would be able to install plumbing and electrical systems. He claims that it could be used to do finishing work too, but there are no animations to suggest how that might work.

So, wave a wand and say "a robot will do it". That doesn't count as "A printer that can print a 2500 sq ft house". Even imagine the robot exists: how exactly does it just 'plug in' the plumbinb and electrical systems? Currently it takes a person to scramble all over the house, drilling and fitting and connecting. There is nothing being solved here. Just a blue-sky story about an imaginary printer.

how exactly does it just 'plug in' the plumbinb and electrical systems?

Lays them in as it builds presumably. Pick and place is easy with a big h-crane robot.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#33

Post by moocowduckquack » Mon, Oct 28, 2013, 7:13 PM UTC

JoeAltmaier wrote: post #9

Hm. A whole house, except the flooring, walls, wiring, plumbing, windows, roofing, lighting, heck anything but the bare frame and walls. You can already do that, prefab, in a few hours. This is custom, but so is prefab. The problem is the usual: a single material isn't enough to make anything but the simple constructs. Almost everything is complex these days. Materials Science hasn't been a science for a century for…

Prefab needs a factory to make the prefab sections.

Also, you can lay down multiple materials by making the head swappable. I would also say plumbing and electrics are nowhere near as complex as they used to be. Everything is available as clip-on fittings these days, which is perfect for automation.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#34

Post by toomuchtodo » Mon, Oct 28, 2013, 7:20 PM UTC

moocowduckquack wrote: post #33

JoeAltmaier wrote: post #9

Hm. A whole house, except the flooring, walls, wiring, plumbing, windows, roofing, lighting, heck anything but the bare frame and walls. You can already do that, prefab, in a few hours. This is custom, but so is prefab. The problem is the usual: a single material isn't enough to make anything but the simple constructs. Almost everything is complex these days. Materials Science hasn't been a science for a century for…

Prefab needs a factory to make the prefab sections. Also, you can lay down multiple materials by making the head swappable. I would also say plumbing and electrics are nowhere near as complex as they used to be. Everything is available as clip-on fittings these days, which is perfect for automation.

You don't even need to be a plumber to do plumbing; PVC the whole thing.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#35

Post by JoeAltmaier » Mon, Oct 28, 2013, 7:21 PM UTC

StandardFuture wrote: post #29

JoeAltmaier wrote: post #9

Hm. A whole house, except the flooring, walls, wiring, plumbing, windows, roofing, lighting, heck anything but the bare frame and walls. You can already do that, prefab, in a few hours. This is custom, but so is prefab. The problem is the usual: a single material isn't enough to make anything but the simple constructs. Almost everything is complex these days. Materials Science hasn't been a science for a century for…

You did not watch the accompanying YouTube video where these things are mentioned as being needed and could be accomplished using auxiliary automation techniques. EDIT: They also show visual demonstrations of techniques as to HOW this would be accomplished. http://www.youtube.com/watch?v=JdbJP8Gxqog

Sure I watched it. I was not convinced. Handwaving is not the same as research.

It may be true that fixtures could be set into place in an automated way. Its not true that they can be connected. And certainly no existing fixture is ready to be installed this way.

Housing construction depends upon massive efficiencies of scale. All the new fixtures for these imaginary houses would be designed differently (for robotic installation, not hand-installation). They will start out at a price many multiples of existing fixtures. They will go down ONLY if this process becomes very, very common.

So, maybe this could be used to make novelty houses for the rich. That's about it. The video makes grand claims about cheap housing for the masses. That is preposterous. Folks in India are not suffering from expensive housing because construction is a dangerous occupation. Its land and materials that are expensive. Labor is the cheapest part of their equation I think.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#36

Post by moocowduckquack » Mon, Oct 28, 2013, 7:21 PM UTC

ChuckMcM wrote: post #18

Sigh, if this guy had an actual 3D printer he would realize those doors and windows are a no-go :-) More seriously, there are unsolved issues in printing unsupported structures, the current best solution is a second 'support' material that can be dissolved/blown away once the part is completed but that rules out voids which are part of what you hope to make with this sort of thing. Also as others point out this is ju…

if this guy had an actual 3D printer he would realize those doors and windows are a no-go

Is easily solvable though by having the robot drop a plank across the gap before printing the lintel, so I don't think that is a showstopper.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#37

Post by moocowduckquack » Mon, Oct 28, 2013, 7:24 PM UTC

ohwp wrote: post #10

The 3d visuals of this printer have been on the internet for more than a year but there are still no pictures and movies of a working version. Concrete printing is not new so I wonder what is keeping them from building the printer over a year later.

Here's CC's 2007 video of the scale prototype robot working - http://www.youtube.com/watch?v=JOlRI_Hm1Po

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#38

Post by angersock » Mon, Oct 28, 2013, 7:25 PM UTC

6d0debc071 wrote: post #30

angersock wrote: post #3

So, I realize that the return on investment for VCs in something like this is probably pretty low--the AEC industry is terrible (though real estate is alright). That said, can we at least be honest that having that much capital locked up in stupid consumer and enterprise gambles is actually a pretty obviously sub-optimal use of that money for the common good?

So, you, as venture-capital-extraordinaire, spend all your money on helping the poor. Maybe you do better in the short term. But what do you do once that money's gone? There's no long-term in that picture and charity for developing nations is not a short-term endeavour. Meanwhile, trivial enterprise and consumer products may not be going to uplift a country any time soon, but they can be recursed upon, and the techno…

CAM machining advancements were mostly pushed forward by car companies and government spending, right? Almost all of the computing technology we use today began as a .gov or .mil project, or came from some private lab with funding from the .gov or .mil.

I think that that rather neatly undermines your point.

It wasn't the VCs spending lots of money. The big desktop revolution (outside of Apple) was initially driven by IBM, Commodore, Tandy, and Atari...all relatively established businesses partnering with younger firms for software or whatnot.

Sure, we had Sun Microsystems and Silicon Graphics, but at the end of the day it seems like they're all gone and blown away while the more lasting impacts remain attributed to academia (.edu and .gov again), defense (.mil), or consumer software not really driven by venture capital.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#39

Post by scotty79 » Mon, Oct 28, 2013, 7:28 PM UTC

ChuckMcM wrote: post #18

Sigh, if this guy had an actual 3D printer he would realize those doors and windows are a no-go :-) More seriously, there are unsolved issues in printing unsupported structures, the current best solution is a second 'support' material that can be dissolved/blown away once the part is completed but that rules out voids which are part of what you hope to make with this sort of thing. Also as others point out this is ju…

> Sigh, if this guy had an actual 3D printer he would realize those doors and windows are a no-go :-)

If you notice guy obviously competent and expeirienced in the field presenting something stupid it's more likely than not that you are missing piece of knowlege that is required for understaning why it actually isn't stupid.

Just watch the movie. It shows how to deal with doors and windows.

Re: A Printer That Can Print a 2,500 Square Foot House in 20 Hours

#40

Post by sologoub » Mon, Oct 28, 2013, 7:28 PM UTC

ChuckMcM wrote: post #18

Sigh, if this guy had an actual 3D printer he would realize those doors and windows are a no-go :-) More seriously, there are unsolved issues in printing unsupported structures, the current best solution is a second 'support' material that can be dissolved/blown away once the part is completed but that rules out voids which are part of what you hope to make with this sort of thing. Also as others point out this is ju…

Just read the info on the "Why Blu" page as the photos look intriguing, but as soon as the page mentioned "pre-fab", it became a non-start for me in the SoCal market.

The problem is that pre-fabricated or pre-manufactured homes are a very specific category of homes and require specific disclosures when selling and financing. It's basically a derogatory term used to describe cheaply built structures that are a step above mobile homes.

Home being one of the most valuable assets a typical family holds. In SoCal and many other places, land makes up the bulk of the value. Place a structure on that land that then detracts from the value doesn't seem very prudent. Not to mention difficult to finance, as banks do not look favorably on that class of building.

So, unfortunately, innovation has to deal with current market and regulatory frameworks. 3D printing as proposed (though my not be valuable), would probably fall under the "custom house" label, which is the most valuable classification. Even if it's just a tract home, it's still a lot better than being considered pre-fab/pre-manufactured.