Proactive Measures for Preventing Highway Embankment Failures on Expansive Soil: Developing an Early Warning Protocol

Efficient data use for early warnings is a critical component of failure management, which encompasses activities such as vulnerable zone mapping, prediction, warning elements, prevention, planning, and action. This study proposes an early warning protocol (EWP) against highway embankment (HWE) failures constructed on expansive soil, implementing in filed situ/lab testing, instrumentation, geophysical testing, 2D/3D finite element method (FEM) analysis, development of machine learning-based predictive models, and analysis of the slope vulnerability index (SVI) factor. Six referenced HWEs were investigated, instrumented, monitored, and considered to measure their evaluative parameters (soil in situ, soil index properties, and factor of safety). The field-recorded data were validated using 2D electrical resistivity imaging. The 2D/3D FEM numerical models were developed based on the field-recorded rainfall volume to analyze the flow and coupled flow deformation of the HWEs' slopes. Predictive models were implemented to analyze the SVI, and the EWP was developed. The action plan was found to be the main component of the proposed EWP. Observing, planning, deciding, and acting are the proposed EWP's key predecessors. The developed EWP for embankment slope failure mitigation could benefit various public and private transportation agencies.