Soft Computing Solutions for Reducing the Carbon Footprint of Fly Ash Based Concrete

Paul O. Awoyera, Joshua Adetola, Mohammed Nayeemuddin, Hiren Mewada, Olaolu George Fadugba

The construction industry significantly contributes to environmental degradation, with many structures exhibiting high carbon footprints throughout their construction processes and lifespans. Activities such as cement hydration and other common construction practices substantially influence environmental conditions over time, necessitating a critical evaluation of material and design choices. This study reported the environmental impact of fly ash (FA), which is largely used to enhance concrete strength. A prediction of two endpoint indicators, that is, global warming potential (GWP) and CO 2 emission using soft computing methods are presented, which are particularly effective for handling complex, nonlinear relationships in environmental data. To achieve this, two machine learning approaches, the random forest (RF) and decision tree (DT) models, are employed to assess the environmental impact of structural materials and designs. Two datasets were obtained from reputable databases, including ResearchGate, ScienceDirect, Semantic Scholar, and Mendeley Data. The models are trained to explore the potential for optimizing structural designs and material selections to minimize environmental impacts. Feature importance is analyzed using Shapley values, providing insights into the most influential factors affecting GWP and CO 2 emission Model performance is evaluated using R 2 and root mean square error (RMSE) metrics. Notably, the RF model achieved an R 2 score of 91% for GWP and 97% for CO 2 emission, demonstrating superior predictive accuracy compared to the DT approach. The findings demonstrate the effectiveness of these machine learning techniques in enhancing the sustainability of construction practices, offering a pathway for informed decision‐making. This study highlights the urgent need for innovative approaches in the built environment to support sustainable development and mitigate the carbon footprint associated with structural engineering.