Approximately 21% of treated drinking water in the UK is lost due to leakage in water distribution networks. Therefore, pinpointing the locations of these leaks is crucial for water companies. This study focuses on creating innovative data fusion techniques to autonomously evaluate the precision and effectiveness of diverse sensing technologies in practical clean water pipe networks, employing autonomous robotics to minimise human intervention.

One goal of this project is to leverage pressure and acoustic sensors in a water network, capitalising on both the hydraulic and acoustic characteristics of a leak in the system. The application of graph theory and machine learning algorithms will be instrumental in achieving the research objectives.