Real-time Anomaly Detection in the Behaviour of Structures

Civil structures such as bridges, dams, buildings, and tunnels are major components of infrastructure networks which establish a foundation for the economic growth of a country. Infrastructure deterioration is associated with a negative economic impact caused by the direct maintenance costs, as well as the indirect costs for production and transport of goods and services. One long-term solution to mitigate the effects of deterioration, to prolong the service life, and to minimize life-cycle cost of ageing infrastructure is to adopt preventive and predictive maintenance strategies. All structures experience deterioration and degradation during their lifetime and thus, early detection of anomalies in the rate of change of degradation could be used as a trigger for preventive maintenance and interventions. Yet, there is currently no generic data-interpretation methodology capable of detecting anomalies in real time without also being adversely affected by false alarms.