Bridging Physical and Digital Realms: The Role of Digital Twins in Geotechnical Engineering-Paper

Paper

Pdf Summary

The paper examines how digital twins—dynamic virtual replicas continuously updated with data from sensors, IoT systems, BIM, and other sources—can improve geotechnical engineering. By linking physical assets to physics-informed computational models, digital twins can support real-time monitoring, performance prediction, safety and maintenance planning, and better project decisions across tunnels, infrastructure, and renewable-energy developments.<br /><br />The proposed modelling framework has four main stages: quantify uncertainty in ground and design parameters; simulate asset behaviour using finite-element or finite-difference models; optimise calculations, including through surrogate models when simulations are costly; and update predictions as new evidence becomes available, using Bayesian methods.<br /><br />Two soft-ground tunnelling case studies demonstrate the approach. In a Tehran roadway tunnel, sensitivity analysis reduced 28 input variables to 10 influential parameters, while a neural-network surrogate reduced computational demands. Monitoring data were used to update settlement predictions. In the more complex Rio Piedras subway-station case, modelling spatial variability in soil properties was computationally intensive. The study reduced costs by focusing random-field modelling on the soil’s zone of influence and identifying an adequate number of simulations.<br /><br />The authors emphasise that such models become genuine digital twins when developed and updated throughout construction and operation using time-sequenced, real-world data and integrated with BIM and other project systems. Implementation barriers include limited expertise, complex data management, industry conservatism, information misalignment, cybersecurity risks, and cloud-computing costs. Successful deployment depends on high-quality, standardised data, effective human interpretation, and consultative leadership that promotes multidisciplinary collaboration. Overall, digital twins can improve geotechnical safety, efficiency, and decision-making by accounting for uncertainty and incorporating new observations as projects evolve.

Keywords

digital twins

geotechnical engineering

sensor data

BIM integration

uncertainty quantification

finite element models

surrogate models

Bayesian updating

tunnelling case studies

real-time monitoring