Evaluation framework for smart predictive digital twin for water supply systems: a case study in Portugal

Evaluation framework for smart predictive digital twin for water supply systems: a case study in Portugal Mariana Alão*,**, Ana Luísa Reis*,**, A. Andrade-Campos* * Department of Mechanical Engineering, TEMA - Centre for Mechanical Technology and Automation, LASI - Intelligent Systems Associate Laboratory, University of Aveiro, Campus Universitário de Santiago, 3810-193 Aveiro, Portugal; [email protected] ** SCUBIC, Parque da Ciência e Inovação, Edifício Principal, 3830-352 Ílhavo, Portugal Abstract: Water Supply Systems are critical infrastructures, and it is important that their service is reliable and can adapt to unpredictable situations in real time. In addition, their complexity makes it difficult for humans to operate them efficiently. There is a strong need for support decision tools to help manage water supply systems. Currently, there is reluctance among water systems operators to use these systems. To increase their confidence, this work presents a framework for evaluating the performance of smart predictive digital twins applied to water supply systems. By providing a structured assessment of accuracy, operational performance and efficiency, this framework will increase the trust in smart predictive digital twins and facilitate their adoption. Keywords: Smart Predictive Digital Twin; Water Supply Systems; Performance metrics Water supply systems (WSS) ensure that water reaches our homes in sufficient quantity and quality. Because water is vital to human life, WSS are considered critical systems (Kalyani et al. 2023). Approximately 1.13% of all energy consumed in Europe is used for WSS, of which at least 60% is by the pumps. However, these are complex systems that are almost humanly impossible to manage. Due to the retirement of WSS operators and municipal merges, there is a strong need for support tools to help manage this vital resource (Berglund et al. 2023; Kalyani et al. 2023). In the recent years, Digital Twins (DT) have begun to emerge. They are dynamic digital replicas that can replicate and synchronise the physical world with the virtual in real-time, making them indistinguishable. It consists of three components: the simulator, real-time streaming data and connection between the other two components. Smart Predictive Digital Twins (SPDT) are an advanced version of a DT that incorporate predictive capabilities and optimize the process it replicates. A schematic representation of a SPDT is shown in Figure 1.1. Despite the enormous potential of real time DTs to be applied to WSS, only a few applications are reported (Alzamora, Castro & Vitens 2021; Berglund et al. 2023). Furthermore, water utilities are reluctant to allow real-time remote control of WSS and prefer to control the network using static rules or based on offline support decision tools (Berglund et al. 2023; Castelletti et al. 2023). For managers and operators to build confidence in the operation of WSS with the of DT, it is crucial for them to understand how well these systems are performing (Berglund et al. 2023). The number of studies available that propose metrics for assessing the performance of DT are still limited. The importance of a complete evaluation, using quantitative metrics to understand their accuracy and effectiveness was highlighted by (Sharma et al. 2022), however without mentioning any specific metric. In some domains, domain-specific metrics have already been defined in order to address the unique requirements and challenges of the specific context, such as the ISO 16250:2013 for the evaluating vehicle crashes simulations (Zeng et al. 2023) . With this in mind, the present study proposes a framework for the evaluation of the performance of a SPDT applied to the WSS. The framework encompasses both general and context-specific metrics, which are evaluated not only for the entire system but also for specific components of the DT. These metrics are designed to access the accuracy, fidelity, operational performance and the computational efficiency of the SPDT. The developed metrics will be calculated for a SPDT applied in a real-world WSS and working in real time, by giving orders directly to the SCADA system. The defined metrics provide a structured methodology for benchmarking the capabilities of the DT enabling the comparison between different methodologies and enhancing the water management process. With this methodology, the confidence of operators in decision-support tools will increase, making them more likely to rely on their recommendations and not replace the suggested operations by manual ones, when operating close to the operational limits. This trust leads to an increased system resilience and facilitates the adoption of DT technologies in WSS. Figure 1.1 Smart Predictive Digital Twin and its components. ACKNOWLEDGEMENTS This work is supported by the doctoral grant (Ref. 2024.04917.BDANA) financed by the Portuguese Foundation for Science and Technology (FCT), by the FEDER and Regional Operational Program of the Center Region (CENTRO2030) within project IReTiS-LeaksD&Op nº 17304 (CENTRO2030-FEDER-01177300) and through the Portuguese Foundation for Science and Technology (FCT), supported by the Recovery and Resilience Plan (PRR), within project I-ReTiS-Leaks (2024.07270.IACDC). This paper was also supported by the project UID 00481 Centre for Mechanical Technology and Automation (TEMA). REFERENCES Alzamora, F.M., Castro, M. & Vitens, G., 2021, Digital Twins-A new paradigm for water supply and distribution networks. Berglund, E.Z., Shafiee, M.E., Xing, L. & Wen, J., 2023, ‘Digital Twins for Water Distribution Systems’, Journal of Water Resources Planning and Management, 149(3). Castelletti, A., Ficchì, A., Cominola, A., Segovia, P., Giuliani, M., Wu, W., Lucia, S., Ocampo-Martinez, C., Schutter, B. De & Maestre, J.M., 2023, ‘Model Predictive Control of water resources systems: A review and research agenda’, Annual Reviews in Control, 55, 442–465. Kalyani, C., Pradnya, K., Samkit, C. & Patil, V., 2023, ‘Regular and Equal Water Supply System’, SAMRIDDHI : A Journal of Physical Sciences, Engineering and Technology, 15(01), 34–37. Sharma, A., Kosasih, E., Zhang, J., Brintrup, A. & Calinescu, A., 2022, Digital Twins: State of the art theory and practice, challenges, and open research questions, Journal of Industrial Information Integration, 30. Zeng, J., Li, G., Gao, Z., Li, Y., Sundararajan, S., Barbat, S. & Hu, Z., 2023, ‘Machine learning enabled fusion of CAE data and test data for vehicle crashworthiness performance evaluation by analysis’, Structural and Multidisciplinary Optimization, 66(4).