Meta uses artificial intelligence to produce greener concrete

The most popular building material used every year, between 10 and 30 billion tons of concrete, cement, the building block of concrete, accounts for 8% of annual global carbon dioxide emissions. New formulas developed with artificial intelligence provide a greener way to produce concrete.

A greener method for concrete production

University of Illinois, Meta AI and researchers at concrete supplier Ozinga, with the artificial intelligence-supported formulas they developed, managed to reduce the carbon footprint by 40% while maintaining the durability of concrete.

Artificial intelligence can design formulas for higher strength and low carbon concrete. Artificial intelligence trained on more than 1,000 concrete formulas along with structural properties, including 7-day and 28-day compressive strength data, generated many potential formulas.

Researchers performed tests by selecting five of the formulas created. In the test applications, a concrete formulation that meets the long-term strength requirements and has a much lower carbon effect compared to the regional comparison was obtained.

Although it offers many advantages, the concrete produced with the formulation obtained dries later than the existing methods and this causes the construction process to be prolonged. Researchers will continue to develop by examining how the concrete produced in the following stages will affect different areas of use and weather conditions.