Concrete is the most widely used building material on Earth, and making its key ingredient, cement, accounts for roughly 8 percent of human-caused carbon emissions, according to MIT. One of the simplest ways to cut that footprint is to use less concrete. MIT engineers have now shown a surprisingly low-tech way to do it: pour the concrete into molds sewn from fabric.
In a paper in the Journal of Building Engineering, announced by MIT News on November 6, 2026, a team led by civil engineer Josephine Carstensen built a 3-meter (about 10-foot) test beam that used about one-third as much concrete and half as much steel reinforcement as a conventional beam designed to carry the same weight.
The problem with clever shapes
Engineers can already use a computer method called topology optimization to work out where material is really needed in a beam and where it can be removed. The results are organic, often intricate shapes. The trouble, MIT explains, is that these shapes are hard or expensive to cast in ordinary rigid molds. The few that have been built mostly relied on specialized equipment such as concrete 3D printers or machine-carved molds, which are scarce in regions where much of the world’s new construction is expected.
Fabric formwork, where wet concrete is poured into a textile mold, has been explored for years. Its weakness is that the fabric sags and deforms under the weight of the concrete, so the final shape has been difficult to predict.
Turning flexibility into a feature
The MIT team, which includes co-authors from Boston University, decided to design for that deformation instead of fighting it. “The key realization was that this flexibility is not just a limitation to manage; with the right computational pattern design, it can become a way to produce optimized structural forms,” Carstensen says in the MIT release.
The workflow combines two design methods. Graphic statics, which produces structural layouts almost instantly, was used to shape the steel bar so it carries an equal share of the load along its length and stays simple enough for builders to make. Topology optimization, slower but free to place material anywhere, decided where concrete could be removed. An automated process then converted the 3D design into flat fabric pieces that can be cut and sewn, accounting for how the cloth will stretch once filled. A simple timber frame held the sewn mold in place.
What the tests showed
The team built three molds for the same design using 36, 56 and 76 fabric pieces. After casting, 3D scans compared each beam with the target shape. More seams gave finer local control but did not necessarily improve the overall shape, because small cutting and sewing errors can add up. The 56-piece mold struck the best balance.
The researchers also noted practical details. Woven cotton lets water seep through, so the mix had to be moderately wet and sand-rich with only small stones, so it could flow into narrow, angled legs. Workers packed it by hand and kneaded the outside of the fabric, which removed the need for a vibrator that a rigid mold of this shape would require. All three beams came out free of wrinkles and creases, although the bottom sagged somewhat and the angled legs folded slightly inward.
In a load test, the team stacked 36 water-filled buckets on one beam, up to about 30 percent of its design load to mimic everyday service. The measured deflection at the center was within about 2 percent of the computer model’s prediction.
Where the economics work
MIT’s cost comparison adds a useful reality check. In the Global South, where materials cost more relative to labor, the fabric versions were cheaper, with savings of about 10 to 20 percent. In the Global North, higher labor costs for sewing flipped the result, making fabric beams about 16 to 55 percent more expensive. The approach looks most promising for custom components made one at a time or in small batches, such as exposed beams, long-span roof framing, pedestrian structures and landscape elements. For repetitive rectangular beams in high-wage markets, conventional reusable molds remain the cheaper choice.
What it means
The world’s building stock is expected to nearly double by 2050, with about 85 percent of that growth in developing countries, MIT notes. If designers there can get strong structures from less cement using a sewing machine, suitable fabric and a timber frame, the climate benefit does not have to wait for robotic or 3D-printing infrastructure.
There are limits. The beam is a proof of concept comparable to a short building span, and the researchers say a larger beam needs a new design because its own weight and the pressure on the mold change. The next step is to test multiple beams to failure, a key safety milestone before real buildings could use the technique.
Bottom line
MIT’s fabric-formed beam shows that smarter geometry, not exotic machinery, can cut concrete and steel use sharply. It is a research result, not a building product yet, but it points to a more accessible path toward lower-carbon construction.