The Challenge The client was looking for innovative solutions to help reduce interruptions to supply and to provide a better customer experience. The key challenge was how to identify customers most at risk of experiencing long interruptions to water supply when an incident took place. Once these “at-risk” customers had been identified, the results needed to be communicated to non-technical staff in a visual, informative manner. The Project Jumping Rivers utilised modern machine learning techniques to predict when an interruption event would last beyond the threshold for a customer area.