Predicting Sustainable Concrete Performance Using AI
Keywords:
AI, artificial intelligence; concrete; infrastructure; life-cycle assessment; machine learning; neural network; sustainability.Abstract
Performance prediction of concrete exhibiting extreme average performance, or a sufficiently accurate upper bound on performance, can be achieved with a comparatively small number of high-performance samples. AI-assisted performance prediction thus offers an attractive research avenue for use, for instance, in accelerating the quest for ideal material compositions allowing for reduced emissions while fulfilling performance criteria. When trained on a suitable dataset, phonetic characteristics from AI-driven models can demonstrate how changing a specific variable influences performance outcome, thus providing insight for design. Moreover, when calibrated with the appropriate input parameters, AI-based prediction models can be incorporated into broader Civil Engineering models to account for durability of surface materials used in infrastructures such as bridges or tunnels exposed to specific environmental loading.
The AI approach can be used advantageously when applied in areas where specialised knowledge is limited or when it is difficult to explore a problem space adequately via simulations. Results presented in this study therefore aim to advance these areas with consideration to sustainability by demonstrating a specific AI model capable of increasing the development speed and enhancing the knowledge regarding a considered concrete. Further results will support the analysis of durability of campuses and Civil Engineering assets under specific environmental loading and will pave the way towards a PhD research topic encompassing planning and support for future interventions in Rock Engineering related scenarios.
References
[1]Abbas, A., & Muntean, R. (2025). Environmental impact and CO2 emission analysis of cement-based materials in sustainable construction. Journal of Cleaner Production, 502, 145250. [2]Alghamdi, H. (2025). Decision Tree methods for the categorization and performance classification of high-strength sustainable mix designs. Materials and Structures, 58(2), 45–58. [3]Ansari, S. S., Ibrahim, S. M., Hasan, S. D., & Mohammad, Z. (2025). Effect of perforated fly ash cenospheres and nano-silica on the thermo-structural behaviour of sustainable concrete. Structures, 80, 109800. [4]Aziz, M. A., Ibrahim, S., & Rahimi, A. (2025). Performance benchmarking of hybrid AI approaches for structural engineering: A review of 2020–2024 trends. Results in Engineering, 26, 102340. [5]Barbhuiya, S., & Sharif, A. M. (2024). Machine learning applications in the optimization of supplementary cementitious materials for decarbonization. Construction and Building Materials, 411, 134250. [6]Duan, Y. (2025). AI-powered data mining for the development of sustainable concrete materials [Master's thesis, Massachusetts Institute of Technology]. DSpace@MIT. [7]Feng, Y., Tak, M. S. N., & Mahgoub, M. (2025). Advanced machine learning techniques for predicting concrete compressive strength. Infrastructures, 10(2), 26. [8]Han, J., Zhang, L., & Liu, X. (2024). Multi-model ensemble learning for the compressive strength prediction of high-strength green concrete. Journal of Building Engineering, 82, 108315. [9]Kelechi Enobie, C., Ahmed, A., & Smith, J. (2024). Circular economy principles in construction waste management through blockchain and AI integration. Sustainability, 16(4), 1542. [10]Mansouri, E., Manfredi, M., & Hu, J.-W. (2024). Hybrid machine learning models for the estimation of geopolymer concrete strength using CatBoost and extra trees. Sustainability, 16(12), 4980. [11]Nikoopayan Tak, M. S., Feng, Y., & Mahgoub, M. (2025). Advanced machine learning techniques for predicting concrete compressive strength. Infrastructures, 10(2), 26. [12]Omoregie, A., Alhassan, M., & Ouahbi, A. (2024). Bibliometric analysis of machine learning and deep learning trends in concrete technology. Automation in Construction, 158, 105210. [13]Razzaq, A., & Shah, M. (2025). Deep learning and artificial neural networks in the representation of intricate relationships in construction data. Archives of Computational Methods in Engineering, 32(1), 112–130. [14]Shylaja, N., & Prashanth, M. (2025). Fusion of conventional engineering methods with sophisticated algorithms for resilient concrete prediction. Case Studies in Construction Materials, 22, e01842. [15]Vedrtnam, A., Gunwant, D., & Chaturvedi, S. (2025). Metaheuristic optimization of machine learning models for green concrete durability prediction. Materials Today: Proceedings, 104, 556–564. [16]Wattanapanich, S., Limkatanyu, S., & Sae-Long, W. (2024). Revolutionary instruments in concrete design: A transition from empirical to AI-driven modeling. Structures, 60, 105820. [17]Wu, J., & Zhou, Y. (2025). Hybrid SVR-PSO models for the feature analysis of splitting tensile strength in sustainable high-performance concrete. Civil Engineering Journal, 11(1), 88–104. [18]Xiao, J., Wang, Y., & Li, T. (2025). Global production trends and environmental degradation of contemporary concrete infrastructure. Nature Reviews Earth & Environment, 6(2), 75–89. [19]Xie, Y., Zhang, Z., & Wang, Q. (2025). Enhanced Bayesian Gaussian process regression for predicting the mechanical properties of blended concrete. Frontiers in Materials, 12, 169825. [20]Yasen, A., Al-Ani, M., & Khalid, N. (2024). Extreme Learning Machine (ELM) for strength prediction of eco-friendly cellular concrete. Engineering Applications of Artificial Intelligence, 128, 107450.
