Enhancing the predictive accuracy of recycled aggregate concrete’s strength using machine learning and statistical approaches: a review

Recycled aggregate concrete (RAC) has emerged as a sustainable alternative in the construction industry, reducing environmental impacts. However, predicting the mechanical properties of RAC using traditional experimental methods is challenging due to material variability and the complex interactions within the concrete matrix. This review paper explores the application of machine learning (ML) techniques for predicting the engineering properties of RAC, with a focus on compressive strength (CS), split tensile strength (STS), and durability. Various ML models, including artificial neural networks (ANN), support vector machines (SVM), and ensemble methods, are examined for their effectiveness in handling high-dimensional data and modeling non-linear relationships. The paper emphasizes the critical role of input parameters such as the water-to-cement ratio (W/C), aggregate replacement ratio, and curing period in determining RAC strength. It also discusses the advantages of ML over conventional statistical methods in predicting RAC properties, demonstrating enhanced accuracy and predictive reliability. Recommendations for future research include adopting hybrid ML approaches and further exploring feature importance analysis to optimize RAC mix designs. This comprehensive review highlights the potential of ML to revolutionize material property predictions and promote the informed use of recycled materials in construction.

Graphical Abstract