There is this factory for lack of a better word near me that makes houses, packages them on a truck in pieces, and will ship them around the US to a foundation. All is said and done it's _maybe_ 100k cheaper to go with them than to buy the land and find your own contractors (and when the cost is between 300k-750k either way it doesn't really matter).
The essay touches on why this is the case, but fundamentally the issue with homebuilding isn't that we haven't optimized how to build houses. It's that only certain small segments of the population have seen anything but crushing decreases in wages on top of rampant inflation. So of course, when the average income of a region is 35k and the average house is 650k, there are issues that optimizing can't solve.
We have effectively managed to demonize factory made houses by referring to the people that live in them as trailer trash. I honestly think 3-D printed concrete houses at some point will be how most things are built you’ll get tons of Americans talking about how do you can’t access the wires/plumbing and stuff and it basically just comes down to having enough windows or access points into the inner wall. Not to mention this is how everyone else in the world builds houses.
I wonder how much it’s due to the fact that homes are such an incredibly expensive endeavor compared to other things. I could see significant economies of scale in most of the materials and parts, but perhaps there has been a steady increase in quantity in those materials?
A question about prefab construction came up at a talk this year in Sydney with Lucy Turnbull (Former Sydney Mayor) and Alain Bertaud (planner and author order without design), Lucy mentioned someone tried this in Sydney and went under and they never heard from them again despite promising the world. Alain mentioned that tastes (think in terms of from finishes to floor plans) change often enough where prefabricating an entire house doesn't really make sense. Not to mention construction codes can change as well (I know in the US it can vary on a county level), they mentioned they saw more success with prefabricating components like windows or fireplaces or whatever.
Something like a factory requires an intensive upfront captial investment, if tastes change often enough the process would need to be amendable to adapt to changing tastes.
Combined with that, I think the fact there is no uniform standards for acceptable floor plans, compliant layouts and construction codes across the different jurisdictions really makes it hard for there to be economies of scale.
> note I don’t think construction codes are strictly a problem within the US, there’s apparently a manufactured housing code. However planning controls are a seperate thing and possibly still an issue.
An example from Sydney (which likely relates to other jurisdictions) Outsides construction code, in Sydney there is a quasi instrument called the apartment design guide which issues requirements on floor plans, floorspace, how far a bedroom wall can be from a window in a bedroom, ceiling heights a lot of things that act as constraints on the possible layouts of a home, and I have no doubt some form of this exists in other jurisdictions as well. I imagine when there is so much variation in different legislative constraints in different jurisdictions there isn't really economies of scales as there are actually several different non homogenous market segments with incompatible set of constraints, and where there's overlap it may not be a high demand end product.
I don't think this as much of a problem but I imagine there are cases where some unionised construction industries may refuse to use work on site using prefab components. I haven't really heard of such cases so I'm not convinced this is a real blocker.
Idk, but it seems like you could attempt to use this argument for absolutely anything that is manufactured.
Why do people buy manufactured cars instead of custom ones? Is it because they dont care or is it because a custom car would be 10x more expensive?
If they could actually manufacture a house that is 1/10th or less the cost but the tradeoff is its a little outdated or the layout isnt exactly what they want, i have a hard time beliving people wouldnt take that trade.
Why does a carpenter cut the end off a 10-foot board to get a required 9ft-2in, thereby wasting 8% of the input and incurring dumpster charges? Suppose the architect's design specified the cutlist, to be transmitted to the board "factory", which would cut boards to the required lengths, tagging them with RFID serial numbers indexed to the design, stacking them so the first ones to be used are on the top, and truck to the site without passing through Home Depot?
Lumber is transported from sawmills via rail cars to some depot where it will then get trucked to places like Home Depot.
There are some large-scale lumberyards that still have a railroad siding and will get lumber by the carload.
A delivery of lumber for a typical house is a full load (or more than one) via truck, and obviously a train isn't going to come to a construction jobsite. So the llumber at the construction site is already hand picked...
It's no big deal at all to cut off ends to make it fit exactly, but when framing most of the 2x4s, etc. are already the exact size you need. You just trim off the top and bottom plates at the edges of the rooms.
One thing that can’t be scaled is “prime location”
There will be locations that are more desirable than others, and even if you keep building houses where there’s space, the need to congregate in particular areas (such as for work) will result in particular locations being more desirable.
And, it’s hard to increase the density of an area once the housing supply is already built out.
So instead, that supply stays fixed, demand increases, and the price increases in turn.
This actually made me think then that an accelerator for scalability could be: public transit into population centers that ensure areas with abundant space (and cheaper housing supply) can still easily access the areas that would otherwise be hugely expensive to live near
I believe this was done near DC where the public transit buildout helped foster further housing development in those emerging areas. Not sure if other HCOL areas, e.g. CA Bay Area, have similar things going on for East Bay mobility / other cross-county transport
One problem in development is we keep trying to think big 'build a lot of homes at once'. That creates the suburbs which, long term, is very unhealthy for a city and its residents (and unfair to the core which ends up paying for their services). We need to push for smaller development, but more of it. It is a lr for cities. When you hear 'redevelopment' it generally means too big of a step is being attempted. It is too often a make or break, and too often that just means break so you get nothing bus held up development, and even when it does happen it is too much and you likely overshot in many ways and undershot in many more. Then, years down the line all those houses age out at the same time and their infra ages out at the same time leading to a sudden problem for the city. Smaller projects lead to a diverse and healthy city. You want to make homes cheaper? Publish, and maintain, pre-approved plans for homes and ADUs, but make sure the plans meet city density needs. Give incentives to clear out brownfill. Encourage development in ways that improve the health of a city and you will get healthier cities.
Homes last for 50+ years and are fixed objects that establish the visual look of our communities and outdoor space. They aren't disposable products. The way you get economies of scale is by repetitive builds and a highly optimized supply chain. You could get efficiencies if every home was built and looked the same but most people don't want to make that tradeoff
There are some things that could improve the situation. Post frame construction, Pre built trusses, macerating toilets that are more forgiving for sewer tie ins, localized instant hot so you don't have to run separate hot water lines, radiant heating so you don't have to run the duct work. It's all tradeoffs though and you aren't going to get a $500k house for $30k.
The other thing holding back progress are building codes and city laws. To be fair a lot of those codes exist for good reason but the inspection and permit system is suboptimal in most cases. You can buy a $30k small studio on Amazon right now that shows up on the back of a truck but good luck with your city allowing you to use it as a dwelling.
I think standardized codes would help a lot. Another major issue is labor mobility: Here in California, there's a big labor shortage for all blue collar trades, and even if they do live in the area, they have to spend a lot of time going to/from the job site.
I think this can be largely solved by technology, but with a change of regulations, code, and division of labor in the trades.
1) Put all the power conduit, plumbing and HVAC into standardized modules that can be cut to length with a circular saw, and attached with tools that cost a total of $500 with no skilled labor. It doesn't matter if this increases material costs by 50% for those components because they are cheap vs. labor. I'd rather waste a $50 piece of conduit than pay three different tradespeople $100+/hour to hand-build junctions where the wasted piece would end up being.
2) The next big cost is probably drywall finishing + doors. I don't have a great solution. I can imagine just 3d printing the whole interior once the conduits are placed.
3) Roofs can be cheap if rooflines are simple, since that allows stuff like metal roof trim to be fabbed at a factory. I don't think asphalt shingles are going to make much sense in many places 30 years from now, so probably just bite the bullet, and pick something wind and fireproof, then make it cheap.
4) Put solar panels somewhere other than the roof, or replace the roof material with them entirely.
5) Framing and insulation are already embarrassingly cheap vs the rest of the house. Probably not worth optimizing unless it saves finishing labor in the next step (e.g., 3d print a beautiful interior wall so you don't have to pay for someone to apply joint compound + paint).
That leaves the foundation + architecture / engineering work as the hard part. Most of the design work for that stuff could be automated. Let the homeowner and builder boss an LLM around, and then run the LLM output through code compliance + simulation gates. The latter is really important because most local code is hazard or climate dependent, and having good deterministic vetting of designs would let the construction process apply to multiple climates.
Prefab could make sense, but, in practice, those people don't pick up the phone. One major issue is delivering the house to things like hills, or at the end of windy / suburban roads. (The prefab sections want to be 30-50ft long, but your residential road doesn't want to support trailers over 20-30' or so).
Most homes built don't use conduit for electric anymore. A few jurisdictions like Chicago require it. In any case, doing it new construction is quite easy and usually the job given to the newest apprentice electrician.
Plumbing (including HVAC lines and gas lines) already come in standard sizes and can be cut with a saw; connecting them is trivially easy. (HVAC lines might be slightly harder but everything else is easy.)
Hanging drywall and mudding it is not particularly expensive and is necessary for good fire resistance. There are cheaper techniques like pre-printing wallpaper on drywall, but most people would rather pay more to have a conventionally painted wall.
Metal roofs are already fabbed at the factory. You simply order what you want and install it. Obviously a simpler roofline is less work to install.
Solar panels can't be the roof because few want to engineer a solar panel which also is capable of being a watertight, weather-resistant roof tile that can hhandle things like snow load, driving rainstorms, and has the ability to do simple repairs. There have been attempts at this, but then it costs more than just a conventional roof plus conventional solar panels.
"3D print" a wall doesn't make sense - where is the 3D printer going to come from and where it will be set up? 3D printing stuff requires having a base, rigidity, and so on; it's just not well suited for putting up walls which is a pretty trivial task to do.
Hope this didn't sound dismissive, but the problems in homebuilding that could be solved by manufacturing techniques and automation already have been solved by the people who are in the business of making singlewides, doublewides, and modular homes; it saves a little money but doesn't save that much, and the resulting product is often quite a bit less desirable and less durable than a conventionally built house.
I work in the construction industry, you did not ask the LLM the right questions.
1) is already the way residential construction works. Conduit isn’t used in homes aside from Chicago and a handful of other jurisdictions but all conduit can be cut with a sub $500 tool called a portaband. Plumbing pipe can be cut with a portaband. There are multiple power tools that can be used to cut sheet metal. Fittings for the electrical, plumbing, and HVAC trades are commoditized and standardized.
2) Drywall and drywall finishing is the cheapest labor on a building site aside from the cleaners and maybe the insulators. The framers, finish carpenters, equipment operators, flooring workers, and MEP trades all make more than the drywallers, mudders/tapers, and painters.
3) Raised seam metal roofs are great and last longer than asphalt shingles, but it’s more expensive up front.
4) Solar roofs have far too many terminations/connections to be reliable, the upfront cost may possibly never be paid back. That could change, but it’s a tricky problem.
> That leaves the foundation + architecture / engineering work as the hard part. Most of the design work for that stuff could be automated. Let the homeowner and builder boss an LLM around, and then run the LLM output through code compliance + simulation gates. The latter is really important because most local code is hazard or climate dependent, and having good deterministic vetting of designs would let the construction process apply to multiple climates.
The architect and engineer design the construction documents to adhere to the codes in the jurisdiction the structure is being built in, no LLM needed. Most jurisdictions just adopt one of the standardized codes for general code like the IBC and then various trade codes like the NEC.
Things like roof load calculations for areas with snow make designing a house that works in both California and Minnesota not worth it unless you want to overengineer your roof for California.
There’s an entire massive industry focused on shaving costs to increase margins that have been working on these problems for a long time.
Key statement: "The fact that the ratio between costs of constructing a home and the costs of the various materials is already so low is fundamentally what makes achieving substantial economies of scale difficult."
That's striking. Building houses looks labor-intensive, but, if that's correct, labor cost isn't that large a fraction of the final cost.
For a large part of residential construction if everything is new it only takes a few days to assemble for each step, the rest is just the logistics of choosing and buying the materials and getting delivery. Like even a large house can be entirely framed in a few days, a week at most, with two or three if all the materials are available and the frame design is finalized. With dimensional lumber construction there is a pretty simple way to stack and nail the floors then walls then ceilings then next floor if there is, then rafters. Work goes as fast as you can toss lumber and shoot nails. Everything so with simple blueprints of where the walls are, the height of the walls, and window and door placements. For a framer its like constructing with legos. Then a roof can be put on in half a day, siding 2-3 days, electrical and plumbing a day or two, drywall a day or two to hang and a day or two to mud and sand, painting can be done in a day or two. Getting decent trim might take longer, but it is also a more skilled and detail orientated work. With framing you can ignore 1/16th gaps and everything will settle and flex into a solid foundation. With pieces of trim a 1/16th gap could be a visible problem.
You could potentially go from nothing to an entire nearly finished house in 2 weeks worth of labor, however because the framers, electrocutions, plumbers, siding, flooring, drywall guys are often all separate contractors and the shipping for all those materials are all on different schedules it takes longer. Or if a few guys are doing it all, they will lose a bit of speed by not being specialists for each task, but can make up that time having a more centrally coordinated plan of attack.
Even then the permits and design planning and choosing the windows and siding and flooring and all that is often still the biggest speed bottleneck.
Lots of interest comments on the efficiency of building homes.
The bigger issue is the cost of land. The differential for land to build is often 10:1.
So even if the prefab shave 10-20% the price for a custom build, it's still not making much of a difference for the normal buyer.
unless a develop or government was opening up mutiple parcels of land well below the costs to build a house, prefab is not really going to be worth it.
There's also the other path of making homes more durable, less maitainance intensive. This would also increase the housing stock and probably reduce total costs over the utilization time by a household.
More durable materials and construction techniques would also reduce the insurance costs which are basically overhead in the economy.
Building an Affordable House goes into some real, actionable things that could be done - and often aren't.
A big part of the problem is the same with cars; nobody makes used cars, and nobody builds used houses. The buyers are the ones with money and they drive the demand.
(Some of the obvious wins have taken over quite quickly; almost no builders frame roof trusses anymore and instead bring them in from a factory on trucks and crane them into position - three men can do in a day what would have taken an entire team a week.)
Maybe we are solving the wrong problem? Should it be, where at the economies of scale in "building places for people to live in". I'd be interested to hear from others about relative costs for high rise, multi-family dwellings, double storey dwellings and so on relative to single-storey single-family dwellings.
There's no good technical reason why you shouldn't be able to pick up flat pack housing at your local Ikea and fix something very serviceable with just an Allen key in an afternoon or so. But it would burst the bubble that housing is expensive and devalue the property of people that indebted themselves to own a house. That's why Ikea is not in the housing business.
Housing is not a technical problem. Our medieval ancestors build housing using just twigs and mud. It's not that complicated to build something vastly better with modern materials. Modern conveniences like heating, electricity, sewers, water, etc. add a bit of complexity of course. But there's no logical reason why you should spend north of half a million on that. If you have a few spare thousands, you can own a pretty nice recreational vehicle that come with most of what you'd need. But good luck finding a spot in most densely populated areas where you would be allowed to live in one.
We keep finding extremely petty reasons not to do pragmatic things to fix housing and the cost of living crisis. Simply stopping the process of policing this sector would in short order lead to most cities gaining uncontrolled slums, camp sites, and what not. The irony of policy failure is that this is in fact happening in lots of places.
I believe all these engineering/technology/economic discussions are missing the human element, humans do human kinda things.
The #1 feature of housing in US is to keep the undesirables out. The best way to implement it is costly housing via zoning, deed restrictions and HOAs. And with costly housing, emerge good schools. Which creates a vicious/virtuous cycle. People who want to live in safe places (and/or good schools) create more demand in these specific locations even though they may not have the original motive (of keeping the undesirables out). This is the power of defaults.
Now, add to this that most of the wealth in US is housing, it creates a perverse incentive to stop any more supply, which they can accomplish at the city/county level.
Note: The above is US specific. There are other things at play in other countries. I'm not sure what drives costly housing in Canada and Australia.
> This means that the ratio of the costs of the output to the costs of the physical inputs (drywall, lumber, concrete, etc.) is around 2.
> You could find a way to use fewer and/or cheaper raw materials.
You can delete almost all of those components, almost all of that assembly work, much of the mass of the structure, much of the supply chain for the raw materials, almost all of the inventory costs, and you can make the financial carrying cost negative. You build it in a factory at room temperature mostly out of nearby rocks, and ship it flat-packed to the destination, where it mostly self-assembles over the course of a few minutes. The result is beautiful, requires little energy to heat or cool, and is impervious to insects, salt, caustics, weather, fire, and resistant to bombs.
Uh, the demonstration is forthcoming, bear with me.
Housebuilding is motivated as much by need for housing as the prospect of increasing market value, "building equity" and so on. This prospect depends most strongly on location, which means the land price dominates the cost of construction.
There isn't a lot of room for scaled improvements because labor speed is almost never the bottleneck in residential construction. It is the logistics of all the supplies, of all the subcontractors, the permitting, which all take considerable time and wait. And if people are building their own custom home, getting them to actually choose what windows, floors, doors, siding, roofing materials, or even just finalize the layout for framing can take a lot of time. The windows could all be installed in a few hours, but getting somebody to choose what windows they want could take days or longer, days for delivery, days for the contractors schedule to align while they are juggling a bunch of different jobs because rarely do they get to just sit on one jobsite for multiple days having everything they need ready to go.
I think people overestimate the labor requirement/cost for building new construction because so much of their experience is dealing with repairs and maintenance and adapting old construction. While it might take a day or two to fix one already existent bathroom because of tearing old stuff out, not making a massive mess, fishing tools and materials through little access holes, you can plumb an entire new framed house with multiple bathrooms that doesn't have drywall in the way or people living in it in the same time or less. In the time it takes to fix your old existent leaking tub, in new construction a plumber could install 3 tubs, 5 toilets, 7 sinks, and hot water heater with all new pipe. And it will be a much easier task with far less struggle and both physical and mental anguish. Construction guys absolutely love new construction because it is literally 10x easier and without having to hack together old standards into new standards.
Material costs are at minimum atleast 50% of the cost of new construction, some contractors will just automatically bid everything at double the base material cost and make decent money off that. And potentially even higher total cost will be just material if everything is decided beforehand and all the materials are ready to go on site, because half of your labor cost is paying for all these contractors and subcontractors to travel around to different sites at different times. If they can put in a solid 20-30 hours on a single project without leaving the job site, they are going to be happy and be making bank and able to put in low bids. If they gotta return to a single job site 20 separate times from repeatably going to stores, waiting for supply deliveries, waiting for decisions, they are going to want some compensation for it and it gets rolled into the labor cost.
35 comments
[ 2.8 ms ] story [ 51.8 ms ] threadThere is this factory for lack of a better word near me that makes houses, packages them on a truck in pieces, and will ship them around the US to a foundation. All is said and done it's _maybe_ 100k cheaper to go with them than to buy the land and find your own contractors (and when the cost is between 300k-750k either way it doesn't really matter).
The essay touches on why this is the case, but fundamentally the issue with homebuilding isn't that we haven't optimized how to build houses. It's that only certain small segments of the population have seen anything but crushing decreases in wages on top of rampant inflation. So of course, when the average income of a region is 35k and the average house is 650k, there are issues that optimizing can't solve.
This is my takeaway: to reduce home-construction costs, we need to apply economies of scale further to the inputs.
What is the idiot index for lumber, drywall, et cetera?
Something like a factory requires an intensive upfront captial investment, if tastes change often enough the process would need to be amendable to adapt to changing tastes.
Combined with that, I think the fact there is no uniform standards for acceptable floor plans, compliant layouts and construction codes across the different jurisdictions really makes it hard for there to be economies of scale.
> note I don’t think construction codes are strictly a problem within the US, there’s apparently a manufactured housing code. However planning controls are a seperate thing and possibly still an issue.
An example from Sydney (which likely relates to other jurisdictions) Outsides construction code, in Sydney there is a quasi instrument called the apartment design guide which issues requirements on floor plans, floorspace, how far a bedroom wall can be from a window in a bedroom, ceiling heights a lot of things that act as constraints on the possible layouts of a home, and I have no doubt some form of this exists in other jurisdictions as well. I imagine when there is so much variation in different legislative constraints in different jurisdictions there isn't really economies of scales as there are actually several different non homogenous market segments with incompatible set of constraints, and where there's overlap it may not be a high demand end product.
I don't think this as much of a problem but I imagine there are cases where some unionised construction industries may refuse to use work on site using prefab components. I haven't really heard of such cases so I'm not convinced this is a real blocker.
Why do people buy manufactured cars instead of custom ones? Is it because they dont care or is it because a custom car would be 10x more expensive?
If they could actually manufacture a house that is 1/10th or less the cost but the tradeoff is its a little outdated or the layout isnt exactly what they want, i have a hard time beliving people wouldnt take that trade.
https://www.construction-physics.com/p/the-elusive-cost-savi...
There are some large-scale lumberyards that still have a railroad siding and will get lumber by the carload.
A delivery of lumber for a typical house is a full load (or more than one) via truck, and obviously a train isn't going to come to a construction jobsite. So the llumber at the construction site is already hand picked...
It's no big deal at all to cut off ends to make it fit exactly, but when framing most of the 2x4s, etc. are already the exact size you need. You just trim off the top and bottom plates at the edges of the rooms.
There will be locations that are more desirable than others, and even if you keep building houses where there’s space, the need to congregate in particular areas (such as for work) will result in particular locations being more desirable.
And, it’s hard to increase the density of an area once the housing supply is already built out.
So instead, that supply stays fixed, demand increases, and the price increases in turn.
This actually made me think then that an accelerator for scalability could be: public transit into population centers that ensure areas with abundant space (and cheaper housing supply) can still easily access the areas that would otherwise be hugely expensive to live near
I believe this was done near DC where the public transit buildout helped foster further housing development in those emerging areas. Not sure if other HCOL areas, e.g. CA Bay Area, have similar things going on for East Bay mobility / other cross-county transport
There are some things that could improve the situation. Post frame construction, Pre built trusses, macerating toilets that are more forgiving for sewer tie ins, localized instant hot so you don't have to run separate hot water lines, radiant heating so you don't have to run the duct work. It's all tradeoffs though and you aren't going to get a $500k house for $30k.
The other thing holding back progress are building codes and city laws. To be fair a lot of those codes exist for good reason but the inspection and permit system is suboptimal in most cases. You can buy a $30k small studio on Amazon right now that shows up on the back of a truck but good luck with your city allowing you to use it as a dwelling.
I think this can be largely solved by technology, but with a change of regulations, code, and division of labor in the trades.
1) Put all the power conduit, plumbing and HVAC into standardized modules that can be cut to length with a circular saw, and attached with tools that cost a total of $500 with no skilled labor. It doesn't matter if this increases material costs by 50% for those components because they are cheap vs. labor. I'd rather waste a $50 piece of conduit than pay three different tradespeople $100+/hour to hand-build junctions where the wasted piece would end up being.
2) The next big cost is probably drywall finishing + doors. I don't have a great solution. I can imagine just 3d printing the whole interior once the conduits are placed.
3) Roofs can be cheap if rooflines are simple, since that allows stuff like metal roof trim to be fabbed at a factory. I don't think asphalt shingles are going to make much sense in many places 30 years from now, so probably just bite the bullet, and pick something wind and fireproof, then make it cheap.
4) Put solar panels somewhere other than the roof, or replace the roof material with them entirely.
5) Framing and insulation are already embarrassingly cheap vs the rest of the house. Probably not worth optimizing unless it saves finishing labor in the next step (e.g., 3d print a beautiful interior wall so you don't have to pay for someone to apply joint compound + paint).
That leaves the foundation + architecture / engineering work as the hard part. Most of the design work for that stuff could be automated. Let the homeowner and builder boss an LLM around, and then run the LLM output through code compliance + simulation gates. The latter is really important because most local code is hazard or climate dependent, and having good deterministic vetting of designs would let the construction process apply to multiple climates.
Prefab could make sense, but, in practice, those people don't pick up the phone. One major issue is delivering the house to things like hills, or at the end of windy / suburban roads. (The prefab sections want to be 30-50ft long, but your residential road doesn't want to support trailers over 20-30' or so).
Plumbing (including HVAC lines and gas lines) already come in standard sizes and can be cut with a saw; connecting them is trivially easy. (HVAC lines might be slightly harder but everything else is easy.)
Hanging drywall and mudding it is not particularly expensive and is necessary for good fire resistance. There are cheaper techniques like pre-printing wallpaper on drywall, but most people would rather pay more to have a conventionally painted wall.
Metal roofs are already fabbed at the factory. You simply order what you want and install it. Obviously a simpler roofline is less work to install.
Solar panels can't be the roof because few want to engineer a solar panel which also is capable of being a watertight, weather-resistant roof tile that can hhandle things like snow load, driving rainstorms, and has the ability to do simple repairs. There have been attempts at this, but then it costs more than just a conventional roof plus conventional solar panels.
"3D print" a wall doesn't make sense - where is the 3D printer going to come from and where it will be set up? 3D printing stuff requires having a base, rigidity, and so on; it's just not well suited for putting up walls which is a pretty trivial task to do.
Hope this didn't sound dismissive, but the problems in homebuilding that could be solved by manufacturing techniques and automation already have been solved by the people who are in the business of making singlewides, doublewides, and modular homes; it saves a little money but doesn't save that much, and the resulting product is often quite a bit less desirable and less durable than a conventionally built house.
1) is already the way residential construction works. Conduit isn’t used in homes aside from Chicago and a handful of other jurisdictions but all conduit can be cut with a sub $500 tool called a portaband. Plumbing pipe can be cut with a portaband. There are multiple power tools that can be used to cut sheet metal. Fittings for the electrical, plumbing, and HVAC trades are commoditized and standardized.
2) Drywall and drywall finishing is the cheapest labor on a building site aside from the cleaners and maybe the insulators. The framers, finish carpenters, equipment operators, flooring workers, and MEP trades all make more than the drywallers, mudders/tapers, and painters.
3) Raised seam metal roofs are great and last longer than asphalt shingles, but it’s more expensive up front.
4) Solar roofs have far too many terminations/connections to be reliable, the upfront cost may possibly never be paid back. That could change, but it’s a tricky problem.
> That leaves the foundation + architecture / engineering work as the hard part. Most of the design work for that stuff could be automated. Let the homeowner and builder boss an LLM around, and then run the LLM output through code compliance + simulation gates. The latter is really important because most local code is hazard or climate dependent, and having good deterministic vetting of designs would let the construction process apply to multiple climates.
The architect and engineer design the construction documents to adhere to the codes in the jurisdiction the structure is being built in, no LLM needed. Most jurisdictions just adopt one of the standardized codes for general code like the IBC and then various trade codes like the NEC.
Things like roof load calculations for areas with snow make designing a house that works in both California and Minnesota not worth it unless you want to overengineer your roof for California.
There’s an entire massive industry focused on shaving costs to increase margins that have been working on these problems for a long time.
That's striking. Building houses looks labor-intensive, but, if that's correct, labor cost isn't that large a fraction of the final cost.
You could potentially go from nothing to an entire nearly finished house in 2 weeks worth of labor, however because the framers, electrocutions, plumbers, siding, flooring, drywall guys are often all separate contractors and the shipping for all those materials are all on different schedules it takes longer. Or if a few guys are doing it all, they will lose a bit of speed by not being specialists for each task, but can make up that time having a more centrally coordinated plan of attack.
Even then the permits and design planning and choosing the windows and siding and flooring and all that is often still the biggest speed bottleneck.
The bigger issue is the cost of land. The differential for land to build is often 10:1.
So even if the prefab shave 10-20% the price for a custom build, it's still not making much of a difference for the normal buyer.
unless a develop or government was opening up mutiple parcels of land well below the costs to build a house, prefab is not really going to be worth it.
More durable materials and construction techniques would also reduce the insurance costs which are basically overhead in the economy.
A big part of the problem is the same with cars; nobody makes used cars, and nobody builds used houses. The buyers are the ones with money and they drive the demand.
https://www.amazon.com/Building-Affordable-House-Fernando-Pa...
(Some of the obvious wins have taken over quite quickly; almost no builders frame roof trusses anymore and instead bring them in from a factory on trucks and crane them into position - three men can do in a day what would have taken an entire team a week.)
Housing is not a technical problem. Our medieval ancestors build housing using just twigs and mud. It's not that complicated to build something vastly better with modern materials. Modern conveniences like heating, electricity, sewers, water, etc. add a bit of complexity of course. But there's no logical reason why you should spend north of half a million on that. If you have a few spare thousands, you can own a pretty nice recreational vehicle that come with most of what you'd need. But good luck finding a spot in most densely populated areas where you would be allowed to live in one.
We keep finding extremely petty reasons not to do pragmatic things to fix housing and the cost of living crisis. Simply stopping the process of policing this sector would in short order lead to most cities gaining uncontrolled slums, camp sites, and what not. The irony of policy failure is that this is in fact happening in lots of places.
The #1 feature of housing in US is to keep the undesirables out. The best way to implement it is costly housing via zoning, deed restrictions and HOAs. And with costly housing, emerge good schools. Which creates a vicious/virtuous cycle. People who want to live in safe places (and/or good schools) create more demand in these specific locations even though they may not have the original motive (of keeping the undesirables out). This is the power of defaults.
Now, add to this that most of the wealth in US is housing, it creates a perverse incentive to stop any more supply, which they can accomplish at the city/county level.
Note: The above is US specific. There are other things at play in other countries. I'm not sure what drives costly housing in Canada and Australia.
> You could find a way to use fewer and/or cheaper raw materials.
You can delete almost all of those components, almost all of that assembly work, much of the mass of the structure, much of the supply chain for the raw materials, almost all of the inventory costs, and you can make the financial carrying cost negative. You build it in a factory at room temperature mostly out of nearby rocks, and ship it flat-packed to the destination, where it mostly self-assembles over the course of a few minutes. The result is beautiful, requires little energy to heat or cool, and is impervious to insects, salt, caustics, weather, fire, and resistant to bombs.
Uh, the demonstration is forthcoming, bear with me.
I think people overestimate the labor requirement/cost for building new construction because so much of their experience is dealing with repairs and maintenance and adapting old construction. While it might take a day or two to fix one already existent bathroom because of tearing old stuff out, not making a massive mess, fishing tools and materials through little access holes, you can plumb an entire new framed house with multiple bathrooms that doesn't have drywall in the way or people living in it in the same time or less. In the time it takes to fix your old existent leaking tub, in new construction a plumber could install 3 tubs, 5 toilets, 7 sinks, and hot water heater with all new pipe. And it will be a much easier task with far less struggle and both physical and mental anguish. Construction guys absolutely love new construction because it is literally 10x easier and without having to hack together old standards into new standards.
Material costs are at minimum atleast 50% of the cost of new construction, some contractors will just automatically bid everything at double the base material cost and make decent money off that. And potentially even higher total cost will be just material if everything is decided beforehand and all the materials are ready to go on site, because half of your labor cost is paying for all these contractors and subcontractors to travel around to different sites at different times. If they can put in a solid 20-30 hours on a single project without leaving the job site, they are going to be happy and be making bank and able to put in low bids. If they gotta return to a single job site 20 separate times from repeatably going to stores, waiting for supply deliveries, waiting for decisions, they are going to want some compensation for it and it gets rolled into the labor cost.