Portuguese 
Spanish 
3 min of reading 
0 comments 
Researchers at the Paul Scherrer Institute Create New Low-Carbon Cement Recipes Using Artificial Intelligence and Advanced Genetic Algorithms
Eco-friendly cement may be about to gain traction in the construction industry. Researchers at the Paul Scherrer Institute in Switzerland have developed a way to create new eco-friendly cement recipes in a matter of seconds using artificial intelligence.
Instead of lengthy laboratory tests, scientists used artificial intelligence to create new cement recipes with lower emissions.
The Problem with Conventional Cement
Cement acts as a binding agent in concrete. When mixed with water and gravel, it forms minerals that make the material hard and durable. The problem lies in its manufacturing. To produce cement, limestone must be heated to 1,400 °C. This generates a large amount of CO₂.
-
Helicopter Drops 180 Tons of Sand and Gravel on Swedish River to Revive Ecosystem Damaged by Decades of Exploitation
-
Brazilian Company Launches Retro Smartphone with 48 MP Camera, T9 Keyboard, and Privacy Features for Social Media-Free Use
-
Self-Taught Chinese Farmer Builds 5-Ton Submarine from Scrap, Launches It in Anhui River
-
Invisible Induction Technology Enables Wireless Power for Blenders, Coffee Makers, and Air Fryers, Reducing Countertop Cables
Surprisingly, most of these emissions do not come from fuel combustion but from the limestone itself. When heated, it releases the carbon dioxide that was chemically trapped in its composition.
Today, some of this cement is replaced by byproducts such as iron slag or coal ash. But the global demand for cement is so high that other solutions need to be found.
AI Finds the “Right Recipe”
Roman Boiger, the mathematician leading the PSI team, states that artificial intelligence serves as a “digital cookbook for eco-friendly cement.”
The created system can simulate different cement mixtures and predict, in seconds, which ones emit less CO₂ while still maintaining the quality of the material.
The scientists trained a neural network with data produced by a software from the institute itself, GEMS.
This program simulates geochemical reactions and the formation of minerals during the cement curing process. As a result, the system learned to predict the behavior of various formulations.
According to researcher Nikolaos Prasianakis, the neural network can calculate the properties of a cement mixture in milliseconds—which is about a thousand times faster than conventional methods.
Genetic Algorithms Assist in Selection
In addition to the neural network, the team also used genetic algorithms. This type of technique is inspired by natural evolution. The idea is to automate the testing of which material combinations produce efficient, low-carbon cement that is viable for large-scale production.
John Provis, co-author of the study, states that some of the new recipes are already showing real potential. They not only reduce emissions but also maintain strength and can be applied in the industry.
This Is Just the Beginning
Despite the advancements, researchers emphasize that there is still a long way to go. The discovered formulas need to undergo many practical tests. The current study is merely a proof of concept.
The use of artificial intelligence has enabled the discovery of promising recipes without the need for thousands of physical experiments. This may speed up the process and lead to the creation of more sustainable construction materials.
Other projects are also under development. In one, researchers transformed clay into cement. In another, 3D printing with graphene was used to create stronger concrete.
Cement with lower carbon levels is still not ready for large-scale use. However, the use of AI demonstrates that there are faster and more efficient ways to get there.
-
1 person reacted to this.