We last looked at 3D-printed buildings about three years ago. At the time, the technology was still nascent. Although there was a lot of enthusiasm, many efforts were academic, with little commercial deployment, and practitioners were experimenting with a wide range of approaches.
I attended the SXSW ICON event and know a few folks over there and I think this is a fair characterization of their progress thusfar. I was surprised to learn that their printer (at least the gantry version) is capable of doing horizontal reinforcement. They seemed to gloss over this in the presentation but it seemed big to me. However, for vertical reinforcement they are still printing pockets for a laborer to slot the rebar into, and then filling the pocket.
On the construction costs side, I think they vastly understate the gains from specialization and streamlined processes that have come from the standardization of the conventional American house building process. Trades like electrical, plumbing, painting, HVAC, and framed fixtures like widows seem like they could get easily double in labor costs, considering most tradesmen and contractors will be unfamiliar with the new wall system and might struggle to work around it's constraints.
I was also a bit disappointed to see that their home designer Vitruvius can't match up floorplans to concept/tour photos, although it seemed to be something people were quite positive about despite this. I'm most excited for the design marketplace and their plan to offer commissions (no indication how significant this might be yet) to home designers for every house printed.
The CEO seemed to want to push people to live more remotely in these types of printed communities, which seemed a bit like a cope one might resort to so as to avoid acknowledging that none of this tech addresses land cost, which is the real root of the housing crisis.
Overall positive outlook, but deserves a few grains of salt.
Hey Connor! Is land cost really the driving factor of the housing crisis? I think only in a few high demand urban areas like New York and San Francisco. Where I live, in one of the nicer parts of Baltimore, the tax assessors and insurance adjusters say the value of my row home is 15% land value and 85% the replacement cost of the structure. America is full of underutilized urban and suburban land, and some of that is due to regulation but I think it would improve a lot (including regulatory improvements) if building high quality structures got cheaper.
It's part of it, but probably downstream from other factors and not the only lever to pull on to fix the problem. For instance, better transport (like autonomous vehicles, flying cars, high speed rail) effectively reduce distance, bringing more land into the "high value" commute bubble where people can still capture agglomeration benefits from these high land cost cities. Subsidizing demand but not supply and then over regulating supply would probably be the thing I would target for largest impact, but idk if it's the easiest to fix.
I have no idea why people were impressed as they were with Vitruvius. It's not great. The concept/tour photos are low quality and often don't make physical sense and, as you noted, they often (rarely?) don't match up with the plans.
I think they're either caught up in the hype or (more charitably) are applying the rate of improvement we've seen in LLMs and image-gen to their expectation of Vitruvius.
I'm a 3d printing enthusiast with a background in home construction. Finally, a subject which I might have something somewhat intelligent to say!
It's pretty amazing what kind of overhang angles you can get with traditional FDM 3d printing, but it's dependent on a lot of variables that require tweaking to get just right. The traditional rule of thumb is 45 degrees, but there's techniques that can get you 90 degrees! (https://www.youtube.com/watch?v=fjGeBYOPmHA just one example technique) I wouldn't be surprised if with the right setup that you could 3d print a 6/12 or possibly a 4/12 roof. It would likely have a pretty ugly underside at those low angles.
It'll be interesting to see what sort of possibilities there are for concrete with this. I'm wondering if the material (which cures somewhat slowly) and the volume (with a large print-bead, you have much lower surface area/volume ratios which might make it harder for a layer to stick without vertical supports) add some constraints that aren't there with conventional FDM and make this sort of thing difficult.
Yeah, it'd be neat to see what they've tried and work through a lot of variables on this.
I wouldn't be *terribly* surprised if there was a combination of additives, temperatures, air flow, and magic that encourages the concrete to cure enough to support a cantilevered next layer at a greater degree then you might expect. A lot of this will have to be tweaked per floorplan since if it's a larger cantilevered area each layer is going to have more time to cure before it gets another layer on top. With their boom-type system I'd investigate if you could print the minimal set of walls to print a roof layer, go off to print more walls somewhere else in the plan, go back print another roof layer, and repeat just to give more cure time per layer. The tradeoff would be longer print times, so you'd have to weigh that against just building the roof in a more traditional manner.
On some sites and floor plans you might be able to print the roof off to the side and take your time between layers and then fly it in via crane when it was complete.
Another thing that could be tweaked is the thickness and volume (as you noted) of the layer. Reduce the flow and reduce the layer height and each layer has more time to go off to a large enough extent.
If I worked at Icon (I'm currently looking for work as a technical product manger/software engineer/vp of engineering sort of person!) I'd also be investigating different support structures. You probably don't want a concrete column in the middle of your house but it'd be interesting if you could maybe use some sort of robotically-stackable column system that could be easily broke down by human labor. Or heck, maybe even a standardized support system that can be thrown into place by humans until our robot overlords are advanced enough to do it all.
There is definitely a place for printed homes, just like other prefabricated solutions. The hype, however, is unjustified.
I have similar concerns that apply to both prefabricated and printed structures. Because of the large amount of resources required, buildings of all types should be as long-lasting as possible. What is often forgotten in this space is that in order to achieve longevity, buildings must be adaptable to changing usage patterns. The greatest advantage that traditional construction has is that you can cost-effectively add to or update traditional structures several times over their useful lifespan without compromising structural integrity. But when it comes to prefabricated structures or printed ones, the design criteria are exceedingly rigid for that initial use case. And if one goes in to make changes, that will require a pretty in-depth engineering analysis of what is feasible, much more involved than a traditional model. And likely to be completely impractical at the end of the day.
Anyone who has gutted an old wire lathe and mortar tile bath knows how profoundly important flexibility is.
Some folks, especially a lot of amateur historic preservationists, bitch about ”crap materials” like cement board and drywall.
They can be made near as durable, as soundproof, as lasting, but when it comes time to change the use or layout of a building they’re a totally different ballgame from the old stuff.
And as we’re about to see with a lot of larger, newish class B downtown office blocks, shit that can’t be changed easily or cheaply will die a lot sooner of a death than the durability of their structural system would dictate.
I think you’re right about the newer stuff ending up demolished. Seems wasteful. And going through the engineering analysis that would enable me to safely repurpose the buildings is likely not cost effective
The challenge isn't the engineering analysis. I'm not even sure there'd be much analysis required for most offices to convert to residential use; office live loads are 50 PSF, there are punching checks all over the place, filing rooms are 250 PSF, all of these are very conservative for residential loading. No, it's the ability to utilize the floorplate efficiently...
Old masonry and early steel-framed office buildings were built with small enough floorplates, usually, that we can convert them into units which will meet both customary and regulatory requirements to make attractive condominiums and apartments, lots of corner units, no rooms without natural light.
The "big but not specialized and a bit older" stuff that makes up most class B markets is not nearly as amenable to conversion; what do you do with the interior space far from the windows, and how do you completely reconfigure large, concentrated plumbing stacks and MEP systems that were designed to circulate air within 10-20,000 sq. ft. open floorplans?
I'm curious how you would relocate something like an electrical or plumbing run in one of these structures. I can't find any specifics about how things that typically run through the walls are installed in these houses.
It seems like maybe the walls are hollow and thus you just need an entry and exit point to fish wire. It's still a beast to cut large cavities in mortar as opposed to drywall or wood paneling. That wouldn't work for non-flexible plumbing or HVAC though. Some models appear to have cross-bracing inside the walls as well which means moving runs inside walls isn't feasible without just tearing them completely open (again, a much more difficult task for a mortar wall). At least for electrical, maybe you just attach and run conduit along the outside of the walls like in commercial settings for the sake of flexibility and people have to adapt to the aesthetic effects.
I grew up in a 400 year old masonry house. I live in a 100 year old one today. Have recently finished a high profile redevelopment of another 100 year old masonry building, bringing it into the 21st century. It’s always a challenging balance of function and aesthetics with safety. But if you know how the building is put together, you can work with it because of familiarity of how materials behave.
Traditional masonry is very different than stick built.
It seemed strange that we are using 3D printing to make relatively simple modernist buildings. You'd imagine the new technology would enable the return of ornamentation heavy construction that was more commonplace before the 20th century. That would be especially appealing to Hindu Americans living in California.
The problem, at least in many rich countries, is not the price of building housing, but restrictive zoning and expensive land. It is definitely great to hear that there is still lots of innovation in construction, but it won't solve any housing crises (not that they claim it would).
This is the truth. Houses aren’t expensive in America because construction is expensive houses are expensive because land is artificially inflated in price due to NIMBYism
Brian's post that breaks down housing costs (immediately prior to this one) is a summation of why Icon will ultimately fail. Builders "solved" the efficiency problem of structural systems for small scale buildings with stick framing by the 1850's. Cast in place concrete is also pretty impressive. Icon's technology might have some spillover applications someday, but I doubt manufacturers of dimensional lumber are losing sleep over this construction method.
I'll be interested to see how these concrete structures hold up to the notorious clay soil in Texas. Are they going to do the typical concrete thing and crack all over the place once settling and shifting occurs due to the underlying soil? Is the roof rated to last the lifetime of the house as well? I imagine repairing the roof on one of these buildings is a massive expense relative to traditional roofs.
If these buildings prove sturdier than wood-framed structures over time across a variety of environments, then long-term operational and maintenance costs are an additional advantage (notwithstanding Patrick's point about adaptability to structural changes). It sounds like they are already better insulators for climate control than wooden structures, which means lower energy costs.
And just some other questions I had when looking at these houses:
1. Is everything that is anchored to the walls done with masonry anchors? Are any adjustments or hardware needed to deal with the ridged wall surfaces?
2. How is wifi signal in a structure where all of the walls are concrete? Concrete tends to block signals pretty badly, meaning you'll either need to go back to lots of wired connections or have wifi repeaters and/or access points all over the place.
3. How are exterior walls waterproofed?
4. What are the limits on the size, particularly height, for buildings constructed this way? I know there were plans to 3D print a skyscraper in Dubai, but that project looks to be DOA. If this process becomes common but is only economical for single-family homes, I could see a future where apartment, condo, and office interior finishing practices diverge quite a bit from single-family home construction since the former is still done using traditional stud wall building methods. Thus economies of scale for materials and labor gets "cut in half" to some extent.
Some yes, but automobiles were pretty much developed for getting around and replacing horses. Apollo mission developed all kinds of new technologies for it's mission. 3D house technologies are still looking for a practical home after two decades.
That's far too picture-perfect of a historiography. People spent basically all of the 19th century screwing around with internal combustion engines and trying to find the combination of reliability and market fit that would make them a useful tool at scale. Just like they'd spent most of the 18th screwing around with steam engines.
For every piece of technology that was purpose-built, often by a government to win a war, there's at least one that stumbled about drunkenly in the private sector for decades until someone finally bolted on that last bit to make it work well or mated it to the right market.
Very interesting. I’m 40 yrs in the trade in NY and I can vouch that here $/sf for traditional homes is about $400/sf. Also RS means costs are natl average: one needs to factor by location. Good luck with the 3d kit house.
Interesting. I think labor will continue to get more expensive, so they don't need to come done that much in price. As long as the order of magnitude in cost is correct, they have a very good chance of being successful. When prices are close bet on automation.
I attended the SXSW ICON event and know a few folks over there and I think this is a fair characterization of their progress thusfar. I was surprised to learn that their printer (at least the gantry version) is capable of doing horizontal reinforcement. They seemed to gloss over this in the presentation but it seemed big to me. However, for vertical reinforcement they are still printing pockets for a laborer to slot the rebar into, and then filling the pocket.
On the construction costs side, I think they vastly understate the gains from specialization and streamlined processes that have come from the standardization of the conventional American house building process. Trades like electrical, plumbing, painting, HVAC, and framed fixtures like widows seem like they could get easily double in labor costs, considering most tradesmen and contractors will be unfamiliar with the new wall system and might struggle to work around it's constraints.
I was also a bit disappointed to see that their home designer Vitruvius can't match up floorplans to concept/tour photos, although it seemed to be something people were quite positive about despite this. I'm most excited for the design marketplace and their plan to offer commissions (no indication how significant this might be yet) to home designers for every house printed.
The CEO seemed to want to push people to live more remotely in these types of printed communities, which seemed a bit like a cope one might resort to so as to avoid acknowledging that none of this tech addresses land cost, which is the real root of the housing crisis.
Overall positive outlook, but deserves a few grains of salt.
Hey Connor! Is land cost really the driving factor of the housing crisis? I think only in a few high demand urban areas like New York and San Francisco. Where I live, in one of the nicer parts of Baltimore, the tax assessors and insurance adjusters say the value of my row home is 15% land value and 85% the replacement cost of the structure. America is full of underutilized urban and suburban land, and some of that is due to regulation but I think it would improve a lot (including regulatory improvements) if building high quality structures got cheaper.
It's part of it, but probably downstream from other factors and not the only lever to pull on to fix the problem. For instance, better transport (like autonomous vehicles, flying cars, high speed rail) effectively reduce distance, bringing more land into the "high value" commute bubble where people can still capture agglomeration benefits from these high land cost cities. Subsidizing demand but not supply and then over regulating supply would probably be the thing I would target for largest impact, but idk if it's the easiest to fix.
I have no idea why people were impressed as they were with Vitruvius. It's not great. The concept/tour photos are low quality and often don't make physical sense and, as you noted, they often (rarely?) don't match up with the plans.
I think they're either caught up in the hype or (more charitably) are applying the rate of improvement we've seen in LLMs and image-gen to their expectation of Vitruvius.
I'm a 3d printing enthusiast with a background in home construction. Finally, a subject which I might have something somewhat intelligent to say!
It's pretty amazing what kind of overhang angles you can get with traditional FDM 3d printing, but it's dependent on a lot of variables that require tweaking to get just right. The traditional rule of thumb is 45 degrees, but there's techniques that can get you 90 degrees! (https://www.youtube.com/watch?v=fjGeBYOPmHA just one example technique) I wouldn't be surprised if with the right setup that you could 3d print a 6/12 or possibly a 4/12 roof. It would likely have a pretty ugly underside at those low angles.
It'll be interesting to see what sort of possibilities there are for concrete with this. I'm wondering if the material (which cures somewhat slowly) and the volume (with a large print-bead, you have much lower surface area/volume ratios which might make it harder for a layer to stick without vertical supports) add some constraints that aren't there with conventional FDM and make this sort of thing difficult.
Yeah, it'd be neat to see what they've tried and work through a lot of variables on this.
I wouldn't be *terribly* surprised if there was a combination of additives, temperatures, air flow, and magic that encourages the concrete to cure enough to support a cantilevered next layer at a greater degree then you might expect. A lot of this will have to be tweaked per floorplan since if it's a larger cantilevered area each layer is going to have more time to cure before it gets another layer on top. With their boom-type system I'd investigate if you could print the minimal set of walls to print a roof layer, go off to print more walls somewhere else in the plan, go back print another roof layer, and repeat just to give more cure time per layer. The tradeoff would be longer print times, so you'd have to weigh that against just building the roof in a more traditional manner.
On some sites and floor plans you might be able to print the roof off to the side and take your time between layers and then fly it in via crane when it was complete.
Another thing that could be tweaked is the thickness and volume (as you noted) of the layer. Reduce the flow and reduce the layer height and each layer has more time to go off to a large enough extent.
If I worked at Icon (I'm currently looking for work as a technical product manger/software engineer/vp of engineering sort of person!) I'd also be investigating different support structures. You probably don't want a concrete column in the middle of your house but it'd be interesting if you could maybe use some sort of robotically-stackable column system that could be easily broke down by human labor. Or heck, maybe even a standardized support system that can be thrown into place by humans until our robot overlords are advanced enough to do it all.
Good analysis
There is definitely a place for printed homes, just like other prefabricated solutions. The hype, however, is unjustified.
I have similar concerns that apply to both prefabricated and printed structures. Because of the large amount of resources required, buildings of all types should be as long-lasting as possible. What is often forgotten in this space is that in order to achieve longevity, buildings must be adaptable to changing usage patterns. The greatest advantage that traditional construction has is that you can cost-effectively add to or update traditional structures several times over their useful lifespan without compromising structural integrity. But when it comes to prefabricated structures or printed ones, the design criteria are exceedingly rigid for that initial use case. And if one goes in to make changes, that will require a pretty in-depth engineering analysis of what is feasible, much more involved than a traditional model. And likely to be completely impractical at the end of the day.
Anyone who has gutted an old wire lathe and mortar tile bath knows how profoundly important flexibility is.
Some folks, especially a lot of amateur historic preservationists, bitch about ”crap materials” like cement board and drywall.
They can be made near as durable, as soundproof, as lasting, but when it comes time to change the use or layout of a building they’re a totally different ballgame from the old stuff.
And as we’re about to see with a lot of larger, newish class B downtown office blocks, shit that can’t be changed easily or cheaply will die a lot sooner of a death than the durability of their structural system would dictate.
I think you’re right about the newer stuff ending up demolished. Seems wasteful. And going through the engineering analysis that would enable me to safely repurpose the buildings is likely not cost effective
The challenge isn't the engineering analysis. I'm not even sure there'd be much analysis required for most offices to convert to residential use; office live loads are 50 PSF, there are punching checks all over the place, filing rooms are 250 PSF, all of these are very conservative for residential loading. No, it's the ability to utilize the floorplate efficiently...
Old masonry and early steel-framed office buildings were built with small enough floorplates, usually, that we can convert them into units which will meet both customary and regulatory requirements to make attractive condominiums and apartments, lots of corner units, no rooms without natural light.
The "big but not specialized and a bit older" stuff that makes up most class B markets is not nearly as amenable to conversion; what do you do with the interior space far from the windows, and how do you completely reconfigure large, concentrated plumbing stacks and MEP systems that were designed to circulate air within 10-20,000 sq. ft. open floorplans?
Enable one, not me
I'm curious how you would relocate something like an electrical or plumbing run in one of these structures. I can't find any specifics about how things that typically run through the walls are installed in these houses.
It seems like maybe the walls are hollow and thus you just need an entry and exit point to fish wire. It's still a beast to cut large cavities in mortar as opposed to drywall or wood paneling. That wouldn't work for non-flexible plumbing or HVAC though. Some models appear to have cross-bracing inside the walls as well which means moving runs inside walls isn't feasible without just tearing them completely open (again, a much more difficult task for a mortar wall). At least for electrical, maybe you just attach and run conduit along the outside of the walls like in commercial settings for the sake of flexibility and people have to adapt to the aesthetic effects.
I grew up in a 400 year old masonry house. I live in a 100 year old one today. Have recently finished a high profile redevelopment of another 100 year old masonry building, bringing it into the 21st century. It’s always a challenging balance of function and aesthetics with safety. But if you know how the building is put together, you can work with it because of familiarity of how materials behave.
Traditional masonry is very different than stick built.
It seemed strange that we are using 3D printing to make relatively simple modernist buildings. You'd imagine the new technology would enable the return of ornamentation heavy construction that was more commonplace before the 20th century. That would be especially appealing to Hindu Americans living in California.
Judging by the tourists in Italy, baroque ornamentation in architecture is popular among a lot of demographics.
That's the market 3D printed homes should be aiming for instead of just trying to replicate modern architecture but cheaper.
The problem, at least in many rich countries, is not the price of building housing, but restrictive zoning and expensive land. It is definitely great to hear that there is still lots of innovation in construction, but it won't solve any housing crises (not that they claim it would).
This is the truth. Houses aren’t expensive in America because construction is expensive houses are expensive because land is artificially inflated in price due to NIMBYism
Brian's post that breaks down housing costs (immediately prior to this one) is a summation of why Icon will ultimately fail. Builders "solved" the efficiency problem of structural systems for small scale buildings with stick framing by the 1850's. Cast in place concrete is also pretty impressive. Icon's technology might have some spillover applications someday, but I doubt manufacturers of dimensional lumber are losing sleep over this construction method.
I'll be interested to see how these concrete structures hold up to the notorious clay soil in Texas. Are they going to do the typical concrete thing and crack all over the place once settling and shifting occurs due to the underlying soil? Is the roof rated to last the lifetime of the house as well? I imagine repairing the roof on one of these buildings is a massive expense relative to traditional roofs.
If these buildings prove sturdier than wood-framed structures over time across a variety of environments, then long-term operational and maintenance costs are an additional advantage (notwithstanding Patrick's point about adaptability to structural changes). It sounds like they are already better insulators for climate control than wooden structures, which means lower energy costs.
And just some other questions I had when looking at these houses:
1. Is everything that is anchored to the walls done with masonry anchors? Are any adjustments or hardware needed to deal with the ridged wall surfaces?
2. How is wifi signal in a structure where all of the walls are concrete? Concrete tends to block signals pretty badly, meaning you'll either need to go back to lots of wired connections or have wifi repeaters and/or access points all over the place.
3. How are exterior walls waterproofed?
4. What are the limits on the size, particularly height, for buildings constructed this way? I know there were plans to 3D print a skyscraper in Dubai, but that project looks to be DOA. If this process becomes common but is only economical for single-family homes, I could see a future where apartment, condo, and office interior finishing practices diverge quite a bit from single-family home construction since the former is still done using traditional stud wall building methods. Thus economies of scale for materials and labor gets "cut in half" to some extent.
Still seems like technology trying to find a reason for it's existence rather than technology that was developed from the ground up for a purpose.
That’s… the entire history of all technology?
Some yes, but automobiles were pretty much developed for getting around and replacing horses. Apollo mission developed all kinds of new technologies for it's mission. 3D house technologies are still looking for a practical home after two decades.
That's far too picture-perfect of a historiography. People spent basically all of the 19th century screwing around with internal combustion engines and trying to find the combination of reliability and market fit that would make them a useful tool at scale. Just like they'd spent most of the 18th screwing around with steam engines.
For every piece of technology that was purpose-built, often by a government to win a war, there's at least one that stumbled about drunkenly in the private sector for decades until someone finally bolted on that last bit to make it work well or mated it to the right market.
Very interesting. I’m 40 yrs in the trade in NY and I can vouch that here $/sf for traditional homes is about $400/sf. Also RS means costs are natl average: one needs to factor by location. Good luck with the 3d kit house.
Interesting. I think labor will continue to get more expensive, so they don't need to come done that much in price. As long as the order of magnitude in cost is correct, they have a very good chance of being successful. When prices are close bet on automation.
Nice post. Check out Danish company COBOD for an alternative take on 3D-printing homes.
Oh, I see that the “3-year old” post linked in the article was very aware of their printing tech :-)