Johnson, B., Valverde-Pérez, B., Wärff, C., Lumley, D., Torfs, E., Nopens, I., Townley, L., Kapelan, Z., Weisbord, E., Babovic, V., Holloway, T., Williams, J. B., Ouelhadj, D., Yang, G., Martin, B. & Wade, M. (2022), Artificial intelligence, Digital Twins and dynamic resilience, Chapter 3 in: A Strategic Digital Transformation for the Water Industry, Eds Grievson, O., Holloway, T. & Johnson, B., IWA Publishing, 40-61, doi:10.2166/9781789063400.

See: https://iwaponline.com/ebooks/book/860/A-Strategic-Digital-Transformation-for-the-Water

Global digitalisation and the accelerated development of digital technologies and analytical tools have ushered rapid changes in how humans interact with the different sectors of our society. This has had a significant impact on the international water industry, as described in this book. The amount of data we have at our fingertips is typically much more than any person or group can effectively use, so help is needed to improve our decisionmaking based on these large data sets. These vast datasets are generated intensively from instrumentation mounted in water/ wastewater assets (see Chapter 2) leading to vast data silos of individually formatted and difficult to access data. For example, data logged by water companies is typically done at 15 min intervals, equating to 87,000 data points per annum. When multiple parameters are measured and logged for water and wastewater systems, it can result in millions of data points per annum. Processing this data can present a significant challenge for standard spreadsheet-based packages; therefore, new methods are required to convert this information into valuable insights.

The previous chapter discussed the importance of data and how to collect and transform it into information along with the challenges associated with these processes. When the data/information has been generated and stored, it should be used to generate insights that are both communicable and informed. These insights should extract otherwise unidentifiable characteristics, such as hidden dimensions or trends from preexisting data or modelled outputs.

This chapter will present case studies where analytical tools (artificial intelligence, digital twins and dynamic resilience) were used to collect and better understand this available data. These analytical tools can help us understand data in distinct ways.

  • Artificial intelligence (AI) helps us find and make use of obscured hidden patterns in large data sets, whether it be in numerical or visual patterns. This valuable tool enables computer systems to rapidly analyse water-related data in ways that human operators may not.
  • Digital twins (DT) are essentially tools that force the integration of all the available data into a digital simulation of a system. This enables users to make more time-relevant and actionable decisions about operations by providing the bigger picture about how all the data streams are working together and are related to each other.
  • Dynamic resilience (DR) uses machine learning to extract system operating conditions from actual WRRF data. Datadriven simulations are then performed before generated data is transformed into a dynamic resilience heat map of system stresses to represent how a water resource recovery facility (WRRF) system reacts to extreme events or stressors.

    • While the three tools are distinct from each other, they can be combined in some very powerful ways, as the case studies in this chapter show.

    Home

    Copyright © 2025 by Lloyd Townley
    Last revised: 30 June 2025