Geotechnical Backanalysis, Plaxis, Monitoring Data, Digital Transformation

Abstract

In the realm of construction, the influence of ground conditions on structural design is paramount, highlighting the importance of accurate soil characterization. Traditional approaches, predominantly laboratory and in-situ testing, are often marred by significant uncertainties. Advancing the estimation of soil parameters, backanalysis, or inverse analysis, emerges as a potent methodology. This process involves iteratively adjusting the input parameters of a conceptual model (for instance, a Plaxis Model) to ensure its outcomes align with the empirical data gathered from monitoring the physical system. This keynote introduces DAARWIN, a pioneering solution that integrates machine learning algorithms to streamline the backanalysis process. The application of DAARWIN across various construction projects underscores its effectiveness in refining soil parameter estimations, thereby mitigating geotechnical uncertainties and propelling the construction industry towards greater sustainability, efficiency, and safety.