New research from MIT provides a proof of concept for an innovative approach to structural design, arguing that if the technology is widely adopted, it could have a major impact on climate change.
The production of construction materials is one of the largest single contributors to carbon emissions, accounting for 7-9% of overall emissions. While construction itself emits greenhouse gases, just providing a project with raw materials is a polluting process in its own right.
Much of that pollution comes from concrete production, but a lot of it also comes from steel production. Bridges and skyscrapers might not read as inherently polluting, but the sheer volume of building materials still means they represent a real environmental footprint.
This new paper offers a solution called topology optimization. The team designed a computer program (which, incidentally, is not an AI model) that not only figures out how to reduce the volume of material used, but does so while keeping the final output buildable in the real world.
Historically, topology optimization had the problem of producing designs that were impractical to construct. This new approach lets users limit how many components meet at each point and how small the fine parts can be.
The original bridge (top left) is near Buffalo, New York. On the top right is a wood-only design, bottom left is steel-only, and bottom-right is wood and steel combined.
Credit: Schemmer et al.
It’s also capable of designing with multiple materials, using each for its specific weight and other properties, to create even more efficient designs. Though it might be a bit of work, if the designs are indeed practical to construct, they could save contractors a large fraction of overall build costs by reducing material use.
That alone should motivate the use of the technique and, in turn, help reduce the carbon footprint of the construction industry.
The researchers estimate that about 3.7 gigatons of CO2 are released by manufacturing construction materials each year. Meanwhile, their approach enabled the design of structures that used less than 10% as much material as conventional ones. That means the idea could theoretically reduce carbon output on the order of gigatons. The researchers were able to get the entire model running on a conventional MacBook Pro, too, making it both accessible and portable.
Often, it can feel like new environmentally friendly innovations chip away at a problem at best, rather than taking a big bite. Here, we have a genuinely large source of emissions to address, and a means of addressing it that offers industry savings rather than an expense.
Though it might not be as flashy as a new solar panel or even a fusion reactor, it’s this type of innovation that will be needed to arrest climate change and begin to undo the damage we’ve caused over the past 50 years of technological progress.
