Data-driven forecasting of embankment settlement on soft soils using CRS tests data

Geotechnical engineers continually seek accurate methodologies for predicting the time-dependent settlement of embankments on soft soils. This paper introduces a novel data-driven approach that utilises Constant Rate of Strain (CRS) test data to forecast embankment settlements. The proposed framework circumvents the requirement to subjectively assign soil parameters used in traditional forecasting models by directly integrating CRS test data into an automated analytical process, thus significantly reducing manual effort and user dependant variation in forecast results. The accuracy of the method is demonstrated through applications to real-world trial embankments in Australia and the Netherlands. All data used in this study are freely available in digital format from Datamap.