Led by Sheffield University and funded with a £7.7m EPSRC Programme Grant, the ROSEHIPS project brings together experts from academia and industry to solve the challenge of safely and economically safeguarding current and future infrastructure. ROSEHIPS runs until 2027 and involves Queen’s University Belfast, the Universities of Cambridge and Exeter, plus industry partners including Northern Ireland’s Department for Infrastructure, Translink, Arqiva, Cellnex (UK) and Siemens Gamesa.
In 2019 the cost of clearing the UK’s backlog of maintenance works was valued at £6.7bn. ROSEHIPS (Revolutionising Operational Safety and Economy for High-value Infrastructure using Population-based SHM (structural health monitoring)) aims to solve the UK’s infrastructure asset management problem through research to automate health monitoring.
Instead of expensive scheduled inspections, diagnoses can be provided economically by permanently installed sensors that continuously collect structural data and interpret it via computer algorithms that will help inform the ROSEHIPS framework.
“A real challenge to monitoring large infrastructure is the lack of data from these structures, particularly damaged state data,” said Dr David Hester of Queen’s University Belfast. “Even when there is data available on a given structure it is not obvious whether it can be used to inform…decision support on repairs improvements on another broadly similar structure.”
Prof James Brownjohn, Exeter University, added that it is possible to use engineering judgment to identify potential similarities between structures, but subjectivity can lead to a lack of consistency between different analysts. A significant element of ROSEHIPS seeks to redress this with the development and implementation of a formal mathematical approach for calculating a ‘similarity score’ between structures.
Structures will be represented in mathematical graph form and graph theory will be used to calculate a similarity score between them. Dr Hester explained that for structures with a high similarity score it is likely that data from one will be helpful in managing the other.
Prof Elizabeth Cross at Sheffield University continued: “This systematic approach for comparison is a huge step in the field, as for the first time it allows the limited data that is available to be leveraged, by exploiting it to manage other structures.”
For the project’s sensing element, the ROSEHIPS framework will accept whatever sensor data has been collected on a given structure.
SHM has successfully employed machine learning (ML) for certain condition monitoring and structural health monitoring applications, but there have been few examples of ML effectively managing large civil Infrastructure and even fewer - if any - examples for populations of structural types, said Sheffield’s Prof Keith Worden.
“So, the question is how can you most effectively exploit and develop machine learning, or particular aspects of machine learning such as transfer learning, for managing large civil infrastructure,” continued Prof Worden. “The combination of machine learning/artificial intelligence with physics can offer a very novel and powerful approach, which will enable informative and robust predictions in the health monitoring setting.”