This. And after having managed some of these prefabricated housing projects, I found the big problem is in the details. The designers do not have the necessary experience to accurately make panels that fit together correctly. So you sink a ton of time in preconstruction, then fabrication also introduces errors in the final product and then the field teams spend even more time fixing it. Prefab is a great idea that is facing a steep learning curve.
Part of the problem with wood construction is that wood is a biological material. It’s never going to be perfect. It’s going to warp, twist, and bend. When carpenters build a house, they can work around this irregularity and fit pieces together one at a time as they go. But a factory robot needs to operate with every piece having the same predictable geometry. Even if you mill all of your lumber up perfectly square, it will warp a bit sitting in your warehouse as hygroscopic changes and retained stresses subtly change each piece’s geometry.
If you’re building with wood, you’re always going to need some manual hand-fitup to make all everything fit together nice. And if you have to do that anyway, you lose a lot of the efficiencies that come with manufactured homes. If your giant robot assembly line still has a bunch of carpenters at the end of it fitting everything together, you might as well skip the robots and just have the carpenters build the house.
There is still room for automation in home construction, but it’s most useful for the efficient production of individual components. Look how we mass-produce roof trusses in factories. That’s a great application. Or you can automate the production of things like cabinets, which then in turn will be installed and fitted by actual human carpenters.
This. And after having managed some of these prefabricated housing projects, I found the big problem is in the details. The designers do not have the necessary experience to accurately make panels that fit together correctly. So you sink a ton of time in preconstruction, then fabrication also introduces errors in the final product and then the field teams spend even more time fixing it. Prefab is a great idea that is facing a steep learning curve.
Part of the problem with wood construction is that wood is a biological material. It’s never going to be perfect. It’s going to warp, twist, and bend. When carpenters build a house, they can work around this irregularity and fit pieces together one at a time as they go. But a factory robot needs to operate with every piece having the same predictable geometry. Even if you mill all of your lumber up perfectly square, it will warp a bit sitting in your warehouse as hygroscopic changes and retained stresses subtly change each piece’s geometry.
If you’re building with wood, you’re always going to need some manual hand-fitup to make all everything fit together nice. And if you have to do that anyway, you lose a lot of the efficiencies that come with manufactured homes. If your giant robot assembly line still has a bunch of carpenters at the end of it fitting everything together, you might as well skip the robots and just have the carpenters build the house.
There is still room for automation in home construction, but it’s most useful for the efficient production of individual components. Look how we mass-produce roof trusses in factories. That’s a great application. Or you can automate the production of things like cabinets, which then in turn will be installed and fitted by actual human carpenters.