Project Description

This is a project within the Call for IRI MdM Internal Projects 2019, a call of R&D projects addressed to IRI early career researchers, under the María de Maeztu Unit of Excellence Programme.

Research on critical infrastructures has been addressed in numerous aspects in modern cities, such as water/energy supply/distribution, wastewater treatment, communication and transportation, health and financial services, etc. Disruptions or malfunctions of these critical infrastructures may put at risk the functioning of the societies and their economies. Therefore, as aligned with the H2020 programme for critical infrastructure, reliable and intelligent operation of critical infrastructures to reduce their vulnerabilities and increase their resilience is very important.
Up to now, there are plenty of studies for resilient operation of critical infrastructures. It is also one of research lines of the Advanced Control Systems Group (SAC) at the Institute of Robotics and Industrial Informatics (IRI), who addresses optimal and robust operation of large scale infrastructures through advanced real-time supervisory control, fault diagnosis and fault-tolerance control techniques, specifically to the water and energy fields. Operational optimization processes are mostly based on control-oriented simplified models, which represents the main system dynamics but involves some uncertainty, so that the effect of the produced control strategies on the real system applications may not be optimal.
In this context, a digital twin is a virtual representation (model) of the physical processes involved in the infrastructure with enough accuracy and fidelity to predict operational outcomes of different control actions and disturbances, so as to allow testing and assessment without affecting the daily function of the real systems. In recent years, the concept of digital twin has emerged as one of the main tools and are being used to optimize the operation and maintenance of industrial systems.
In particular, supervision of critical infrastructure systems may be tackled using a digital twin to explore a wide variety of normal and fault behaviors. However, up to now, none of existing studies have applied this research yet and we are motivated to address this issue through this project. The main contribution of this project is proposing an advanced real-time supervision approach of critical infrastructure, using a digital twin, data analysis and machine learning for early detection of faults or attacks, as well as for service reconfiguration after the detected event.