SES Water successfully pilots water demand prediction tool

SES Water has successfully piloted a water demand prediction tool and is now looking to roll it out across its area.

Launched in 2022, the pilot ran at the Cheam and Woodmansterne treatment work sites. It used Royal HaskoningDHV’s machine-learning algorithm, Opir, which takes multiple data points into account – including average reservoir flow, time of day and weather conditions – to forecast water consumption up to 72 hours in advance.

By learning from various factors such as dry or hot days and comparing them to similar historical data, the technology has allowed SES Water to make precise predictions about water demand. This has increased system efficiency, and enabled consistent distribution and resource management in the pilot area.

SES Water is the only water company with a fully smart network. Once rolled out fully across its area, SES said Opir has the potential to shape the future of operations by automating production responses and anticipating changes in demand. In addition, it could predict and control crucial operational factors associated with water demand, such as energy and chemical usage.

SES Water’s head of operational technology, Luis Lazarte, said: “The successful completion of the Opir Pilot Project demonstrates the art of the possible. It shows us firsthand, in a real case scenario, the potential benefits of leveraging technologies such as Machine Learning and Artificial Intelligence. The challenge now is to develop a common framework, an eco-system, in which such technologies can be implemented and seamlessly scaled up through the organisation.”