The UrbanFlood project has created an early warning system framework that can be used to link sensors via the internet to predictive models and emergency warning systems. The data collected from the sensors is then analysed using Artificial Intelligence (AI) to assess the condition of the embankments and the likelihood of failure; different models can be used to predict the failure mode and subsequent potential inundation in near real time.
A number of pilot sites were selected across Europe to install the sensors:
Sensor equipment was selected for installation at the various sites across Europe. The variety of sensors can measure pore pressure, temperature, inclination, soil strain, and three dimensional soil deformations.
Based on the previously known or investigated ground conditions, calculations and geotechnical expert appraisal the most likely failure mode for each site was identified and the sensor networks deigned in such a way to pick up this failure mode. HR Wallingford undertook a series of analysis using the sensors to determine the instantaneous probability of failure.
The real time analysis can happen in two ways:
If there is an indication of impending failure a cascade of HR Wallingford developed models can be started:
HR Wallingford worked on this project as part of a consortium of partners from very different disciplines, to develop what we call the ‘Early warning system of the future’.
In Autumn 2012 HR Wallingford participated with the UrbanFlood project in the final IJkdijk experiment in Groningen, the Netherlands, the “All-In-One-Sensor Validation Test”. Before the experiment no one know the failure mechanism by which the two test dikes, both with sensors installed, would collapse. A first analysis of the results showed the tested sensor systems indicated the two embankment breaches well in advance.
HR Wallingford worked on this project as part of a consortium of partners from very different disciplines, to develop what we call the ‘Early warning system of the future’.