Intelligent Geotechnical Digital Twins

Monitoring and Predicting the Performance of Soil–Structure Systems

Understanding how foundations, reclaimed land, and other geotechnical structures perform as conditions evolve.

Geotechnical structures, including foundations, retaining systems, embankments, and reclaimed ground, interact with complex and variable soil environments. Changing loads, groundwater conditions, and environmental influences can affect their long-term performance.

Traditional assessment draws on site investigation, laboratory testing, field monitoring, and numerical modeling. Limited observations and uncertain ground properties, however, can make it difficult to assess settlement, deformation, and stability throughout the life of an infrastructure asset.

Our proposed Geotechnical Digital Twin research aims to connect field observations, physics-based soil–structure models, and AI-assisted prediction to support evolving performance assessment and risk-informed engineering decisions.

Proposed Capabilities

  • Sensor-informed condition assessment: Combine site investigation and monitoring data to evaluate changing geotechnical conditions.
  • Soil–structure interaction modeling: Represent how ground conditions, foundations, and supported infrastructure interact under varying loads and environmental conditions.
  • Predictive performance assessment: Estimate settlement, deformation, pore-pressure response, and other relevant indicators where supported by models and data.
  • Physics-audited predictions: Check AI-enabled predictions against defined geotechnical and constitutive requirements.
  • Uncertainty-aware assessment: Quantify model and data limitations and identify cases needing additional observations or engineering analysis.
  • Closed-loop model updating: Use field observations to refine predictions and feed results back into monitoring and engineering decisions.

Our Vision

Move beyond isolated geotechnical measurements toward connected, data-informed assessment of the changing performance of ground-supported infrastructure.

Geotechnical Digital Twins - Illustrative uncertainty-aware soil-property prediction component