Exploring sustainable construction through experimental analysis and AI predictive modelling of ceramic waste powder concrete

Abstract

This research tackles the one of the major environmental issues raised by the ceramic industry’s production of ceramic waste powder (CWP) during the cutting and polishing stages, Which requires a different strategy to reduce pollution and cut landfill usage. The main goal of this study is to make use of AI model for CWP concrete by applying Machine Learning (ML) techniques to effectively estimate the mechanical property of the concrete. A thorough investigation is carried out on 54 different concrete mixes, using CWP in place of cement in percentages of 10%, 20%, 30%, 40%, and 50%. The baseline is the compressive strength of plain concrete. A supervised machine learning (ML) techniques such as gradient boosting, Random forest and Regression analysis are used to predict the compressive strength of CWP concrete (CWPC). Important metrics including R2, Mean Absolute Error (MAE), and Mean Square Error (MSE) are used to assess the performance of the model. The results show that Random forest performs better than the other models (\(R^2\) = 0.95, MSE = 5.25 KN/mm\(^2\), MAE = 1.05 KN/mm\(^2\)). The study emphasizes how using CWPC as a building material might help reduce water pollution and land degradation. It also emphasizes the efficiency advantages that may be attained by using ML approaches for concrete characteristic estimate, which will ultimately result in time and resource savings for researchers in the construction industry.