Overview
This project aims to predict coastal embankment failures and direct proactive maintenance interventions using emerging satellite monitoring techniques.
About this opportunity
Project outline
The impact of coastal floods on ageing civil infrastructure makes them vulnerable to failure, especially as more extreme wet-dry cycles push engineered slopes into more destabilising conditions. Whilst sensor-based monitoring systems are prevalent, these are expensive to install, operate and maintain, which limits them to specific locations. Remote satellite monitoring offers a new method for detecting potential failures and enhancing the resilience of civil infrastructure.
This project will explore using remote satellite monitoring data to predict embankment failures. It proposes to use InSAR data from open-access Sentinel-1 ESA data (used for monitoring ground movements), optical multi-spectral data from Sentinel-2 (that can remotely monitor groundwater levels), and knowledge about ground conditions and hydrogeological information to create a new risk model to inform asset owners of the areas at the highest risk of failure. This will be achieved at a network scale using machine learning methods, trained on a series of known previous failures during which all the site-specific data is available, with the intention of scaling the final methods up for application at a national scale.
Training and supervision
The successful application will be jointly supervised by Dr Paul Shepley and Dr Eda Majtan who bring geotechnical, fluid-soil-structure interaction and satellite monitoring expertise into the project. The project work will be supported by stakeholders and other industrial partners with experience with satellite monitoring, data science and critical infrastructure assessments that will benefit from the methods under development during the research. The student will also have access to the Geographic Data Science Lab at the University of Liverpool.
Project structure
A candidate with experience in one part of the project (e.g. geotechnical/geological, fluid-soil interaction or satellite monitoring/machine learning) will be provided with training and support during their first year of studies, including through taught modules or working with industrial partners, where relevant and necessary.
Subsequent years will be purely independent research to be spent combining these disciplines to produce an efficient tool for predicting infrastructure failures using remote sensing methods.
Candidates with any relevant experience are strongly encouraged to apply for the position.
For more information, do not hesitate to contact the project supervisors.