Using Artificial Intelligence for water pipeline inspection.

With Dr Jessica Bohorquez, Postdoctoral researcher. School of Civil, Environmental and Mining. Pipelines are key to provide drinking water. However, pipes monitoring and maintenance are often complicated because they are buried underground. Fluid transient waves have been used for assessing and monitoring the condition of pipelines to detect the presence of anomalies (e.g. leaks, blockages) and the occurrence of abnormal events (e.g. bursts). Nonetheless, existing techniques require information in regards to the properties of the pipe (model-based techniques) or imply a large processing time to obtain results. Artificial intelligence algorithms have proven to have significant potential in complementing existing techniques with the development of data-based pipeline inspection techniques. This talk will present techniques that combine transient pressure waves and custom-designed Artificial Neural Networks (ANNs) for the active and passive inspection of water pipelines. A second application of the use of artificial intelligence for de detection of cracks in pipelines using acoustic methods is presented to highlight how AI can support the operation of essential services such as water supply.