University of Exeter – College of Engineering, Mathematics and Physical Sciences
ain supervisor: Dr. Prakash Kripakaran
Co-supervisor: Prof. James Brownjohn
Co-supervisor: Prof. Richard Everson
Many iconic long-span bridges around the world are now being equipped with sophisticated Structural Health Monitoring (SHM) systems with capabilities for recording several types of response (e.g. strain, acceleration), load (e.g. wind speed, vehicle load) and environmental parameters (e.g. temperature) at various bridge locations. The fundamental purpose of having such systems is to enable effective management of these critical infrastructure assets in order to provide a safe and disruption-free transport network. However, at the moment, there are no reliable approaches for linking the large amounts of data from monitoring to structural performance. This research envisions coupling state-of-the-art SHM systems with Big Data technologies for analysing and visualizing the enormous amounts of data generated by monitoring.
The aims are to develop
- a suite of data-driven strategies for data fusion and integrated analysis of heterogeneous measurement histories in order to relate raw measurements to structural performance-related parameters, and
- a visualization platform that integrates recent developments in visualization of multi-dimensional data with Building Information Models (BIM) in order to enable engineers to view and navigate results from data analytics.
The goal is to provide bridge engineers with readily available, open-source data interpretation toolkits that enable analysis of dense spatial and temporal data of bridge environmental, loading and response parameters collected through large-scale bridge SHM systems. The developed techniques will be trialled and demonstrated on real data from the Queensferry Crossing in Scotland. To this end, the student will work actively with major stakeholders in operation and management of the bridge. The student will benefit from the significant expertise and experience within the vibration engineering research group in monitoring of long-span bridges. The student will also have the opportunity to network with leading research groups internationally through various links already existing via the research group.
The project being a multidisciplinary research topic, would require the student to have an undergraduate degree in a quantitative subject such as engineering or computing. The student would also need to have a deep interest and if possible also an academic background in data mining or pattern recognition techniques alongside a strong motivation to apply these techniques for real engineering problems.
The project will begin with a kick-off meeting with industry partners. The student will then undertake a literature review, and create a work-plan with specific objectives to meet the goals of the project. The work will involve iterative development and testing of a data analytics framework using real data from the Queensferry Crossing. The project team will also meet with industry partners periodically for expert evaluation and feedback. Drawing on the support from the academic supervisors and the industry partners, the student will be able to demonstrate a working, proof-of-concept data analytics platform by the end of the third year. The student will also have published research outcomes in leading international journals and presented findings at conferences. By the time of completion of the project, the student will have obtained valuable skills in computing, statistical methods, structural engineering and structural management. The student will also have led research that has had demonstrable impact in a real-life engineering project. All of these are valuable for initiating an exciting academic career or taking on a challenging role in industry dealing with sensing and infrastructure management.
