Digital transformation bears fruit for the water sector

Several experts shared their vision on the most recent trends in hydroinformatics on the occasion of the first hydroinformatics knowledge exchange of this year. Thus, Edo Abraham (TU Delft) outlined the many challenges facing the global water sector, including the development of proactive leakage management for the faster detection, localisation and repair of leakages in the buried water infrastructure. To this end, model-based leakage localisation techniques are a promising option, in which the water network is divided into smaller segments (District Metered Areas, or DMAs) to monitor pressure and inflow. A new stochastically active, model-based leakage localisation makes use of Bayesian probability distributions, so as to integrate uncertainties in the measurements (under the influence of leakages). Leakage and water demand scenarios can provide the basis to determine which DMA management regime allows for the best estimate of the leakage location. The results show that this new approach is faster and more exact than conventional methods.

Uncertainty in optimisation of urban systems

Another subject discussed concerned taking account of uncertainties in the optimisation of urban water systems, which is something that does not happen under the traditional, deterministic approach. The consequence is that, under realistic conditions, such customary solutions frequently do not perform as forecast. Dimitri Solomatine (IHE Delft) showed how robust optimisation (ROPAR) helps in effectively integrating uncertainties through the use of multi-objective optimisation. The approach was explained by using a hydrodynamic modelling of the discharge and sewer system in an urban area. A model of a discharge network calculates the flood and restoration costs on the basis of the rainfall. A robust optimisation of the model requires taking uncertainty into account, for instance by simulating the many (chances of) showers and the resulting flooding and associated costs.

Schematic representation of system uncertainties.

Fiware4Water: from research to practice

The importance of the simpler integration of new hydroinformatics applications – such as optimisation and uncertainty developments – at the water utilities, was emphasised by Christos Makropoulos (National Technical University of Athens, KWR). Most water utilities often already possess a complex architecture of hardware and software systems. But the introduction of new innovations into this environment is frequently complex, and this stands in the way of these practical applications. The European FIWARE standard can offer a solution to this problem. FIWARE – an open-source IT platform – provides what is known as a ‘context broker’, which can connect different platforms. This assures the data flow. The implementation of a European standard like FIWARE makes digital solutions possible in cases where applications use different domains and platforms. Makropoulos showed how in the Fiware4Water project hydroinformatics solutions can, without time loss, be connected to existing systems through the use of the FIWARE standard. For instance, the project includes a test at the Amsterdam West WWTP, in which the influent and treatment process will be monitored in real time, and the process managed using artificial intelligence (AI), which will result in reduced nitrous oxide emissions and energy consumption.

Amsterdam West WWTP – Fiware4Water demo case.

Making hydroinformatics potential manifest

The knowledge exchange meeting was closed by Dragan Savić (KWR, University of Exeter). In his presentation, Savić stressed that we live in a period of exponential data growth: 90 percent of today’s data worldwide was created in the last two years. The big challenge is to transform this data stream into usable information. This also applies to the water sector. All water utilities dispose of gigantic quantities of data and digital systems, but struggle in realising their hidden potential. And the promising technologies on the horizon will render this challenge even greater. In the years ahead we will see an increased application, among others, of artificial intelligence, remote sensing, digital twins and robotics. Attention has to be paid to the integration of this technology in practice. In doing so, we can learn from the successes and failures in other sectors. For instance, accidents with self-driving vehicles or aircraft automatic pilots show that it is essential that humans maintain ultimate control. However, humans need to be well trained in dealing with these digital tools. Security, and specifically cybersecurity, is one element that should not be underestimated in this context.

Commit to expertise

The conclusion arising from this year’s first knowledge exchange meeting of the Hydroinformatics Platform is that the water sector must strengthen its commitment to this field of research. The world around us is digitalising. If we implement the new technology in the right way, it will not only produce a lot more information but also further optimise our operational processes. This requires the development of specific expertise.