Regular readers of Construction Physics know that I formerly worked for the construction startup Katerra, which raised several billion dollars in venture capital in the hopes of revolutionizing the construction industry, and then went bankrupt. In fact, it was my experiences at Katerra that inspired me to start this newsletter in the first place. Katerra’s collapse became a cautionary tale about overenthusiastic investment, too-rapid scaling, and the hubris of outsiders assuming they could improve industries they didn’t really understand.
If they're not immediately cost-competitive, they're going to have to do one of two things:
1. Get the end consumer to buy into (both metaphorically and literally) the benefits they (supposedly) offer on lifecycle costs.
2. Target the luxury market.
To realize point 1, they are going to *have* to find some way of financializing those cost savings into a form of NPV that can be offered to consumers up front. Running their own insurer and bundling 30 years of (lower-cost) insurance into the sale price so it can be amortized across the mortgage would be extremely risky if their assessments of repair costs and resilience don't pan out, but it's the only thing I can think of off the top of my head.
A similar mechanism could be used to bundle "home warranty" maintenance plans into the sale price at below their market cost. There are ways to financialize basically any long-term cost advantage into net present value, after all.
Point 2 seems quixotic to me, personally, but that's because I'm a cynic. Fashions change faster in the luxury market than down-market and people expect to undertake more frequent renovations for stylistic reasons. I can already envision exactly how much of a copper-plated b**** a "bathroom pod will be to renovate.
Luckily, the presentism the average buyer displays will work in their favor here; if it's fashionable enough today, few buyers will think "what happens when I want to gut the bathroom?" That can be further enhanced if they look to historic, high-value materials which have basically always been in style, like traditional waterproof plaster finishes for bathrooms or marble flooring in the kitchen.
Overall, I doubt they and their investors have sufficient faith in their long-term lifecycle cost savings to embrace the first strategy, and the second is psychologically more feasible. But looking at the floorplans they've built I'm skeptical that's what they're doing.
Precast is, as you noted, a largely mature field; the automation tools embraced in central Europe cost out at rough parity against more labor-intensive models used by US precasters, there's no huge learning curve/scale benefit remaining to find and recover, scale is seriously limited by geography due to shipping costs and challenges... a cynic might suspect Katerra's team understands this but the VC-funded salary is shouting louder than their business sense.
Isn’t this approach still limited by transportation cost/distance? And presumably the need for cranes on site to assemble the panels.
Brian, I have always wondered why Lean/JIT principles have seemingly not worked out for site-built homes. It certainly appears there is still a lot of operational waste in the typical process for residential construction.
Wouldn't fast built manufactured homes be a better business in emerging markets where countries are urbanising at a rapid clip? You can gain greater economies of scale and often build out an entire neighbourhood in one shot. Or does the lower labour costs in developing countries make the strategy completely uneconomical.
They didn't bet $200 mil of their own money, it's the investor's money. The new idea is the same old idea, promises of rapid scale-up that can't possibly materialize.
Precast has all the problems of transporting heavy loads on large distances and/or requiring bespoke manufacturing facilities close to the place of installation, without really addressing the main pain points of construction, the high complexity of a modern home that requires substantial on site work. Just look at that metal drywall framing, that took longer to install then building the entire concrete shell. Never mind the delicate tango of electric, plumbing, AC, all while perforating concrete walls.
The claim of speedups is thus largely false, because they can only speed up a limited part of the construction, and even if it was true, it wouldn't be a substantial selling point on the residential market, where paying the mortgage an extra month or two is acceptable compared to things like perceived durability, long term investment value etc. This is why the residential market is very conservative and why high productivity methods have seen a much stronger impact in the commercial market, where time is money.
It will be interesting to watch Katerra 2.0 (ONX Homes) try again. Thanks for the thoughtful, balanced assessment of their odds, Brian. I think they have a reasonable chance at making precast concrete work in the south, I like the bathroom pods concept, and the interior framing using LGS/FRAMECAD seems an appropriate match for the structural concrete panel construction. How competitive (and lucky?) their approaches to land development and B2C sales will be is unclear, and I'd like to better understand how they will insulate and seal the buildings (devil's in the details). But, the technology that seems most problematic to me is their "AI-driven HVAC system (whatever that means)". I don't see anything exciting about their in-house HVAC system claims, which reek of "smoke and mirrors". We'll see which of these unknowns pans out, and how they react to those that don't. I hope you plan to keep us updated, Brian!
These look like a panelized structure but site assembled & finished single family homes. Not sure the cost advantage of these, as most of the labor-intensive work is the site finishing. Maybe they have over-learned the lessons from Katerra and are now being too conservative. And concrete, really? Their web site sure flirts with green-washing.
I watched a YouTube video shot in their Florida factory, and remain underwhelmed. There are no real economies of scale when the production run is less than one house a day. A section of a pre-cast concrete panel shows a smidgin of insulation between a lot of concrete - not sure how "cool" that would be in Florida or Texas. A primary constraint with any factory system (pre-cast, modular, kit, tilt-up, whatever) is that the largest module you can ship on a public road is basically a 40-foot container.
A possible exception could be the brand-new (pending voter approval) "California Forever" project in Solano County, California - they have over 50,000 acres including an industrial area, so they could in theory build a factory to assemble building blocks on, say a 25' wide by 50' long floorprint for row townhouses, and just haul them across their own property and stack them in place.
Until a more carbon-friendly concrete can be developed and brought to market at scale, I don’t see precast concrete or 3D printed homes supplanting wood frame homes in the US or Canada.
Further, unless specialty trade contractor consolidation occurs at some point in the market, the punishing overhead and upfront capital costs of factory construction will likely never compete with the millions of “Chucks in a truck”.
There will likely be industry progress, but not disruption, using engineered wood products, like MPP light frame construction, that doesn’t require factories beyond the manufacturing of the initial raw material. Factories that inevitably shutter during the cyclical downturns of the industry.
Lastly, until zoning and permitting can be streamlined, the housing problem will never be solved by supposed technology enabled construction conglomerates. This is likely why Onx is focused in Florida, where the bureaucratic impediments are likely lower. Developing raw land can take decades in places like California; so good luck with that business model, at least in the short term.
If they're not immediately cost-competitive, they're going to have to do one of two things:
1. Get the end consumer to buy into (both metaphorically and literally) the benefits they (supposedly) offer on lifecycle costs.
2. Target the luxury market.
To realize point 1, they are going to *have* to find some way of financializing those cost savings into a form of NPV that can be offered to consumers up front. Running their own insurer and bundling 30 years of (lower-cost) insurance into the sale price so it can be amortized across the mortgage would be extremely risky if their assessments of repair costs and resilience don't pan out, but it's the only thing I can think of off the top of my head.
A similar mechanism could be used to bundle "home warranty" maintenance plans into the sale price at below their market cost. There are ways to financialize basically any long-term cost advantage into net present value, after all.
Point 2 seems quixotic to me, personally, but that's because I'm a cynic. Fashions change faster in the luxury market than down-market and people expect to undertake more frequent renovations for stylistic reasons. I can already envision exactly how much of a copper-plated b**** a "bathroom pod will be to renovate.
Luckily, the presentism the average buyer displays will work in their favor here; if it's fashionable enough today, few buyers will think "what happens when I want to gut the bathroom?" That can be further enhanced if they look to historic, high-value materials which have basically always been in style, like traditional waterproof plaster finishes for bathrooms or marble flooring in the kitchen.
Overall, I doubt they and their investors have sufficient faith in their long-term lifecycle cost savings to embrace the first strategy, and the second is psychologically more feasible. But looking at the floorplans they've built I'm skeptical that's what they're doing.
Precast is, as you noted, a largely mature field; the automation tools embraced in central Europe cost out at rough parity against more labor-intensive models used by US precasters, there's no huge learning curve/scale benefit remaining to find and recover, scale is seriously limited by geography due to shipping costs and challenges... a cynic might suspect Katerra's team understands this but the VC-funded salary is shouting louder than their business sense.
Isn’t this approach still limited by transportation cost/distance? And presumably the need for cranes on site to assemble the panels.
Brian, I have always wondered why Lean/JIT principles have seemingly not worked out for site-built homes. It certainly appears there is still a lot of operational waste in the typical process for residential construction.
Wouldn't fast built manufactured homes be a better business in emerging markets where countries are urbanising at a rapid clip? You can gain greater economies of scale and often build out an entire neighbourhood in one shot. Or does the lower labour costs in developing countries make the strategy completely uneconomical.
They didn't bet $200 mil of their own money, it's the investor's money. The new idea is the same old idea, promises of rapid scale-up that can't possibly materialize.
Precast has all the problems of transporting heavy loads on large distances and/or requiring bespoke manufacturing facilities close to the place of installation, without really addressing the main pain points of construction, the high complexity of a modern home that requires substantial on site work. Just look at that metal drywall framing, that took longer to install then building the entire concrete shell. Never mind the delicate tango of electric, plumbing, AC, all while perforating concrete walls.
The claim of speedups is thus largely false, because they can only speed up a limited part of the construction, and even if it was true, it wouldn't be a substantial selling point on the residential market, where paying the mortgage an extra month or two is acceptable compared to things like perceived durability, long term investment value etc. This is why the residential market is very conservative and why high productivity methods have seen a much stronger impact in the commercial market, where time is money.
It will be interesting to watch Katerra 2.0 (ONX Homes) try again. Thanks for the thoughtful, balanced assessment of their odds, Brian. I think they have a reasonable chance at making precast concrete work in the south, I like the bathroom pods concept, and the interior framing using LGS/FRAMECAD seems an appropriate match for the structural concrete panel construction. How competitive (and lucky?) their approaches to land development and B2C sales will be is unclear, and I'd like to better understand how they will insulate and seal the buildings (devil's in the details). But, the technology that seems most problematic to me is their "AI-driven HVAC system (whatever that means)". I don't see anything exciting about their in-house HVAC system claims, which reek of "smoke and mirrors". We'll see which of these unknowns pans out, and how they react to those that don't. I hope you plan to keep us updated, Brian!
These look like a panelized structure but site assembled & finished single family homes. Not sure the cost advantage of these, as most of the labor-intensive work is the site finishing. Maybe they have over-learned the lessons from Katerra and are now being too conservative. And concrete, really? Their web site sure flirts with green-washing.
I watched a YouTube video shot in their Florida factory, and remain underwhelmed. There are no real economies of scale when the production run is less than one house a day. A section of a pre-cast concrete panel shows a smidgin of insulation between a lot of concrete - not sure how "cool" that would be in Florida or Texas. A primary constraint with any factory system (pre-cast, modular, kit, tilt-up, whatever) is that the largest module you can ship on a public road is basically a 40-foot container.
A possible exception could be the brand-new (pending voter approval) "California Forever" project in Solano County, California - they have over 50,000 acres including an industrial area, so they could in theory build a factory to assemble building blocks on, say a 25' wide by 50' long floorprint for row townhouses, and just haul them across their own property and stack them in place.
Do you self driving electric trucks will make factory built homes more viable since you will have significantly lower transportation costs?
It'd be great to have their custom HVAC system if they go out of business and you want warranty and repairs...
Until a more carbon-friendly concrete can be developed and brought to market at scale, I don’t see precast concrete or 3D printed homes supplanting wood frame homes in the US or Canada.
Further, unless specialty trade contractor consolidation occurs at some point in the market, the punishing overhead and upfront capital costs of factory construction will likely never compete with the millions of “Chucks in a truck”.
There will likely be industry progress, but not disruption, using engineered wood products, like MPP light frame construction, that doesn’t require factories beyond the manufacturing of the initial raw material. Factories that inevitably shutter during the cyclical downturns of the industry.
Lastly, until zoning and permitting can be streamlined, the housing problem will never be solved by supposed technology enabled construction conglomerates. This is likely why Onx is focused in Florida, where the bureaucratic impediments are likely lower. Developing raw land can take decades in places like California; so good luck with that business model, at least in the short term.
So, ONX Homes is something like Soviet House-building Plant https://ru-m-wikipedia-org.translate.goog/wiki/%D0%94%D0%BE%D0%BC%D0%BE%D1%81%D1%82%D1%80%D0%BE%D0%B8%D1%82%D0%B5%D0%BB%D1%8C%D0%BD%D1%8B%D0%B9_%D0%BA%D0%BE%D0%BC%D0%B1%D0%B8%D0%BD%D0%B0%D1%82?_x_tr_sl=ru&_x_tr_tl=en&_x_tr_hl=uk&_x_tr_pto=wapp