In a groundbreaking development for marine infrastructure, researchers have harnessed the power of artificial intelligence to design high-performance concrete (HPC) that could revolutionize the construction of sea-crossing tunnels and bridges. The study, led by Qiling Luo from the State Key Laboratory of Hydraulic Engineering Intelligent Construction and Operation at Tianjin University and the College of Civil and Transportation Engineering at Shenzhen University, marks a significant leap forward in the application of machine learning to real-world engineering challenges.
The research, published in “Case Studies in Construction Materials,” addresses a critical gap in existing AI-based mix-design studies. “Most studies rely on limited lab-scale mixes and often overlook aggregate gradation, which is crucial for real-world applicability,” Luo explains. To bridge this gap, the team developed an AI-driven framework that combines machine learning-based performance prediction with multi-objective optimization. The framework is supported by a large-scale field dataset comprising 939 HPC mix designs collected from major marine infrastructure projects, including the Shenzhen–Zhongshan Link and the Huangmaohai Sea-Crossing Passage.
One of the standout features of this study is the incorporation of full particle-size distribution descriptors, including eight cumulative sieve residue percentages and maximum aggregate size, to capture granular packing effects. This detailed approach allows for more accurate predictions of key properties such as slump, 28-day compressive strength, and 28-day rapid chloride penetration coefficient, with an impressive R² value ranging from 0.90 to 0.95.
The AI-driven framework not only predicts performance but also optimizes mix proportions to minimize cement content and carbon emissions while meeting strength, workability, and durability requirements. Using the NSGA-II algorithm, the team achieved remarkable results. “Our experimental validation of five AI-designed mixes (C40–C80) demonstrated excellent performance, with 28-day strengths ranging from 44.4 to 80.5 MPa and chloride diffusion coefficients as low as 1.4 × 10⁻¹² m²/s,” Luo notes.
The implications of this research are far-reaching, particularly for the energy sector. Marine infrastructure projects, such as offshore wind farms and subsea power cables, require durable and sustainable materials to withstand harsh environments. The AI-driven design of HPC could significantly enhance the longevity and sustainability of these structures, reducing maintenance costs and environmental impact.
Moreover, the study highlights the practical value of integrating field-scale data and explicit aggregate gradation features into interpretable, low-carbon mix design strategies. This approach not only improves the performance of marine concrete but also sets a new standard for AI-driven design in construction materials.
As the world continues to invest in marine infrastructure, the insights gained from this research could shape future developments in the field. By leveraging AI and large-scale engineering data, engineers and researchers can push the boundaries of material science, creating more resilient and sustainable structures for the future.