The usage of natural resources has been largely increased due to construction activities and conventional concrete causes harmful effects to the environment. To reduce these impacts, shifting to sustainable construction materials in concrete technology is essential. Utilization of construction demolition wastes like ceramic tiles and ceramic powder is a better strategy to reduce solid waste footprint. In this study, waste ceramics and ceramic powder are used to partially replace coarse aggregates (0-100%) and cement (0-35%) respectively. Experimental evaluation of workability and mechanical properties was conducted for M20 grade concrete, followed by interpretable machine learning model to analyze the influence of different parameters. Sustainability–Performance Index (SPI) was calculated to integrate mechanical performance with environmental impacts, by quantifying the proposed concrete’s carbon emission. Results indicate that while 28-day compressive strength shows less reduction by approximately 2.5% at 35% cement replacement, embodied CO₂ emissions reduction was achieved by nearly 32%. This indicates a maximum SPI improvement of about 44% compared to conventional concrete. The comparison of different ML models and parameters analysis by SHapley Additive exPlanations (SHAP) showed that cement related parameters influence strength development greatly, whereas aggregate replacement showed comparatively minor influence. Additionally, feasible ranges of ceramic waste utilization that maintain acceptable mechanical performance were identified using scenario-based predictions. The findings demonstrate that substantial carbon reduction can be achieved without compromising structural requirements, providing a robust decision-support framework for resilient infrastructure.
