An interdisciplinary team in the Laboratory for Waste Management at PSI's Center for Nuclear Engineering and Sciences has developed a groundbreaking approach to cement production using machine learning. "This allows us to simulate and optimize cement formulations so that they emit significantly less CO2 while maintaining the same high level of mechanical performance," explains mathematician Romana Boiger, first author of the study. "Instead of testing thousands of variations in the lab, we can use our model to generate practical recipe suggestions within seconds - it's like having a digital cookbook for climate-friendly cement."
The PSI researchers trained their neural network using data generated from the open-source thermodynamic modeling software GEMS. "With the help of GEMS, we calculated - for various cement formulations - which minerals form during hardening and which geochemical processes take place," explains researcher Nikolaos Prasianakis. By combining these results with experimental data and mechanical models, the team derived reliable indicators for mechanical properties and material quality of the cement.
Among the cement formulations identified by the researchers, there are already promising candidates. "Some of these formulations have real potential," says John Provis, head of the Cement Systems Research Group at PSI, "not only in terms of CO2 reduction and quality, but also in terms of practical feasibility in production." The study primarily serves as a proof of concept - demonstrating that promising formulations can be identified through mathematical calculation. Before implementation, the recipes must first undergo laboratory testing.
In a parallel development, MIT researchers led by postdoc Soroush Mahjoubi have published an open-access paper in Nature's Communications Materials outlining a similar AI-based solution. The MIT team noted that materials like fly ash and slag have long been used to replace some cement in concrete mixes, but demand for these products is outpacing supply as industry looks to reduce climate impacts. "We realized that AI was the key to moving forward," notes Mahjoubi. "There is so much data out there on potential materials — hundreds of thousands of pages of scientific literature. Sorting through them would have taken many lifetimes of work, by which time more materials would have been discovered!"
Analyzing scientific literature and over 1 million rock samples, the MIT team used their framework to sort candidate materials into 19 types, ranging from biomass to mining byproducts to demolished construction materials. Mahjoubi and his team found that suitable materials were available globally — and, more impressively, many could be incorporated into concrete mixes just by grinding them.
These AI innovations are revolutionizing the cement industry, transforming manufacturing processes and becoming indispensable in the fight against climate change by enabling innovative and highly effective approaches to low-carbon cement production.