Women in Control

This lunch session is sponsored by EUCA, University Politehnica of Bucharest, IEEE Women in Engineering and IEEE Control Systems Society.
Scheduling: Wednesday June 14th, 2023.  Location: Library. Room: L.2.1. Time: 12:40-13:20.
Organizers: Ionela Prodan, Daniela Lupu
Abstract: A women networking opportunity, whose aim is to give new perspectives about the current role of engineering, science and industry women with personal stories, gender reports, programs and opportunities.
Speakers:
– Antonella Ferrara
– Maria Prandini
– Jacquelien M. A. Scherpen
– Melanie Zeilinger

Siemens Session

This lunch session is sponsored by Siemens and University Politehnica of Bucharest.
Title: Artificial Intelligence in Personalized Medicine: From Medical Image Processing to Real-time Patient-specific Physiological Computations
Scheduling: Wednesday June 14th, 2023.  Location: Library. Room: L.4.1. Time: 12:40-13:20.
Organizer: Lucian Mihai Itu
Abstract: Physics-based physiological models have shown great promise in being able to noninvasively assess cardiovascular diseases like coronary artery disease, from patient-specific anatomical information, e.g., as obtained from computed tomography scans of the heart and the coronary arteries. We have previously developed a Computational Fluid Dynamics (CFD) based method for performing patient-specific coronary hemodynamic computations. In this talk we’ll present the proposed method, which is based on a multi-variable feedback control framework which ensures that the computed mean arterial pressure and the flow distribution matches the patient-specific values. The boundary conditions at hyperemia are derived from the respective rest-state values via a transfer function that models the vasodilation phenomenon. Since CFD models have a high computational demand, we have further developed a machine-learning-based model for computing personalized coronary hemodynamics as an alternative to physics-based approaches. The anatomical information for performing such analyses is extracted from medical images of the patient. Over the last few years, deep learning-based techniques have revolutionized the field of computer and machine vision and have rapidly entered the field of medical imaging. In addition to the theoretical progress, the convergence of two other factors has contributed to this success: the availability of large amounts of curated data, and the access to high performance computing infrastructure for training and testing the machine-learning algorithms. Thus, the talk will further address the processing of imaging data using artificial intelligence (AI), which is performed at four levels of increasing complexity and wider implications. At the examination level, AI aims at improving, simplifying, and standardizing image acquisition and processing. At the reading and reporting levels, AI focuses on automatic detection and characterization of features and on automatic measurements in the images. At the prediction and prescription levels, AI focuses on risk prediction and stratification, as opposed to merely detecting, measuring, and quantifying images. The digital twin is presented as a concept of individualized computational modeling of human physiology, with AI-based CT fractional flow reserve modeling as example. Finally, at the cohort and population analysis levels, the focus of AI shifts from clinical decision-making to operational decisions.

Industry Panel Session

This lunch session is sponsored by Siemens, Thales, MathWorks, Siemens Industry Software and University Politehnica of Bucharest.
Title: The interplay between applications and theory: Opportunities and challenges for control and process optimization in the current data-driven era.
Scheduling: Thursday June 15th, 2023.  Location: Aula. Room: ACH. Time: 11:20-12:40.
Organizers: Maria Prandini, Silviu Iulian Niculescu
Abstract: New applications enabled by technological advancements typically lead to intriguing theoretical problems, while in turn theoretical discoveries can foster previously unforeseen applications. The goal of this panel session is to discuss such an interplay with a focus on the current era, where smart sensing and actuation systems allow to collect real-time data and use them not only for monitoring and prediction purposes but also for control and optimization of dynamical processes. To this aim, both researchers from industry and academics collaborating with industry have been invited  to provide some insight into the opportunities and challenges offered to our control community by this data-driven era, with a special emphasis on the European Union priorities, namely, the energy, health, mobility, and industry sectors.
Speakers:
– Frank Allgower (U. Stuttgart)
– Riccardo Scattolini (Polini)
– Constantin Suciu (Siemens)
– Patrick Panciatici (RTR France)
– Raduinea Catalin- Florentin (Thales)
– Drew Davis (Mathworks)

MathWorks Session

This lunch session is sponsored by MathWorks and University Politehnica of Bucharest.
Title: REIMAGINE Lab: Scaling Control Labs with Digital Twins and Extended Reality Technologies
Scheduling: Friday June 16th, 2023.  Location: Library. Room: L.2.1 Time: 11:30-12:30.
Speaker: Professor Aleksei Tepljakov, TalTech University
Abstract: Control systems are crucial in various industries, but poorly tuned controllers can lead to decreased performance and wasted energy. Many real-world control systems suffer from this issue due to the complex plants and processes involved in manufacturing and production, as well as a lack of scalable and sustainable hands-on training for control engineers. Enter ReImagine Lab, a novel approach to delivering control engineering education that combines various control lab modes, including virtual, remote, and hands-on experiments. ReImagine can be an effective tool for engineering education in both higher education and vocational schools, as well as for company personnel training. ReImagine is powered by Digital Twins created with MathWorks products and Extended Reality environments built with Unreal Engine 4/5. ReImagine Lab has already had a trial run in TalTech University. The students appreciated the possibility to directly interact with digital twins, and not just with mathematical models, and as a result, the understanding of related control concepts was easier to transfer as part of experiments designed using the ReImagine platform. In this talk, we shall discuss the ReImagine approach to control system education. The participants will learn:
– About the ReImagine framework, its origins and overall design;
– How MATLAB/Simulink can be used for implementing digital twins from the perspective of mathematical modeling and system identification, as well as control design;
– How the Unreal Engine 4/5 side of digital twins is implemented and connected with MATLAB/Simulink;
– How Extended Reality contributes to enhancing the learning processes with Unreal Engine and various VR and AR headsets.

Siemens Industry Session

This lunch session is sponsored by Siemens Industry Software and University Politehnica of Bucharest.
Title: Executable Digital Twin. From simulations to reality.
Scheduling: Friday June 16th, 2023.  Location: Library. Room: L.3.1. Time: 11:30-12:30.
Speaker: Raluca Raia
Abstract: While simulation models are still the basis of modern product design engineering, executable digital twin (xDT) expands the usefulness of a traditional digital replica through the product life-time cycle. This feature allows to run real-time simulations and to make real-time predictions, as the system state of health, increasing the reliability of industrial systems by early fault diagnosis. xDT is a physics-based virtual model and uses only a few data coming from sensors to determine the system real-time performance and then to predict the whole behavior of the system at given virtual sensors locations. It evolves and updates to reflect any change in the real system, becoming smart and self-adaptive by including Kalman filters to estimate the states or the parameters of the system. All approaches must be carefully chosen so that the used algorithms fit the requirements of high-fidelity nonlinear models of industrial complexity. Join us on a journey that takes you from modelling a high-fidelity digital replica of an industrial system to its interaction with the real physical one. We will show you how to:
– Use Simcenter Amesim within the framework proposed by IMOCO4.E
(https://www.imoco4e.eu/) to build a high-fidelity virtual model of an industrial system.
– Improve the physics-based model and used algorithms to match reality by using XiL (Model and Software in the Loop).
– Augment the xDT with new data using Kalman filters
– Perform predictive maintenance, discover effects early and push the device to the limit to make sure it is safe and reliable.
– Install the developed xDT on the edge, in the cloud.

Thales Session

This lunch session is sponsored by Thales and University Politehnica of Bucharest.
Title: Air Traffic Management: Practical applications for tracking algorithms
Scheduling: Friday June 16th, 2023.  Location: Library. Room: L.4.1. Time: 11:30-12:30.
Speaker: Hristea Emanuel
Abstract: The main objectives of Air Traffic Control (ATC) are the regulation of air traffic, preventing collisions between aircraft, obstructions on the ground and expediting and maintaining the orderly flow of traffic. A part of this complex system is managed by the ATC automation center which integrates data sent by all the implemented surveillance sensors. Processing and unifying all types of surveillance data is done by the tracking system in order to provide clear, fused information for visualization and safety nets systems. Enhanced results are obtained based on a multi-sensor tracking system which combines multiple received data pertaining to a single aircraft into a single surveillance track, utilizing the best contribution from each surveillance source and eliminating the influence of their respective drawbacks. In this talk we shall discuss the evolution and application of tracking algorithms to the air traffic management problem. The participants will have the opportunity to step into the:
– Surveilance problem, a key pillar for efficient ATM
– Topsky ATM system, and its overall design
– How tracking algorithms are integrated in our solution and their technological evolution in time
– How Simulation and Validation tools can improve the quality of service of our solution