In the quest for sustainable construction materials, researchers have turned to industrial waste, seeking to transform byproducts into valuable resources. A recent study led by Sagar Paruthi from the School of Architecture and Design at K.R. Mangalam University has made significant strides in this area, demonstrating how machine learning can optimize the use of foundry sand (FS) and coal bottom ash (CBA) in concrete production. Published in *Scientific Reports* (which translates to *Nature Research Reports*), the research offers a promising approach to predicting the compressive strength of concrete mixes that incorporate these industrial byproducts.
The scarcity of natural sand and growing environmental concerns have driven the construction industry to explore alternative materials. FS and CBA, typically considered waste, have emerged as potential partial replacements for natural sand in concrete. However, predicting the compressive strength of these mixes has been challenging due to the complex, nonlinear interactions among their components. Paruthi and his team addressed this challenge by developing a novel machine learning framework to accurately forecast the strength of concrete incorporating FS and CBA.
The study compiled a dataset of 172 mix designs from published literature and evaluated nine machine learning models, including traditional regressors and ensemble methods. The Extreme Gradient Boosting (XGBoost) model stood out, achieving remarkable accuracy with an R² value of 0.983, a root mean square error (RMSE) of 1.54 MPa, and a mean absolute percentage error (MAPE) of just 3.47%. “The XGBoost model’s performance was exceptional,” Paruthi noted. “It not only provided highly accurate predictions but also identified key factors influencing the compressive strength of the concrete mixes.”
Feature importance analysis revealed that curing duration, superplasticizer dosage, cement content, and the water-to-cement ratio were the most significant predictors of compressive strength. This insight could guide engineers and researchers in optimizing mix designs, reducing the need for extensive trial-and-error experiments.
The implications of this research extend beyond the construction industry. By leveraging artificial intelligence, the proposed method accelerates decision-making, reduces costs, and supports the circular economy by encouraging the use of industrial byproducts. “This approach minimizes waste and promotes sustainability,” Paruthi explained. “It’s a win-win for both the environment and the industry.”
The study’s findings could shape future developments in sustainable concrete mix design, offering a scalable and efficient solution for incorporating industrial waste into construction materials. As the energy sector increasingly focuses on sustainability, this research provides a valuable tool for optimizing resource use and reducing environmental impact.
In an era where technology and sustainability are converging, Paruthi’s work highlights the transformative potential of machine learning in the construction industry. By harnessing the power of data and advanced algorithms, researchers are paving the way for a more sustainable and efficient future.