Abstract:
The integration of Artificial Intelligence (AI) into geotechnical engineering is revolutionizing the way foundations are designed, enabling more efficient and consistent solutions. This paper presents a comprehensive AI-driven framework that automates the end-to-end process of foundation design— starting from the interpretation of site investigation data, through geotechnical modelling and numerical analysis, to performance assessment and design option comparison.
The workflow begins with the application of machine learning algorithms to interpret complex site investigation data. These algorithms generate robust geotechnical models that characterize subsurface conditions, including soil stratigraphy, material parameters, and groundwater conditions. From these models, the framework identifies appropriate numerical analysis tools suited to the specific foundation layout and loading scenarios, including axial, lateral, and moment loads, as well as their combinations.
Subsequently, the framework performs geotechnical assessments of foundation performance under various load cases, evaluating critical parameters such as bearing capacity, settlement, and overall stability. A central feature of the system is its ability to intelligently explore and compare a range of foundation options— including shallow and deep foundations— while accounting for site constraints, soil-structure interaction, and construction feasibility. By enabling rapid design iterations and reducing reliance on manual procedures, the AI-driven approach enhances decision-making quality, shortens design cycles, and mitigates the risk of human error.
To illustrate its practical application, a worked example is provided as a test module. The proposed methodology can be extended to other geotechnical scenarios, paving the way for developing a comprehensive AI-powered geotechnical engineering assistant.
This research highlights the transformative potential of AI in geotechnical design, providing engineers with a powerful tool to develop high-quality, data-informed foundation systems. The proposed framework marks a step forward in the digitization and automation of foundation engineering practices.