Title Promoter Affiliations "Random and integrable models in mathematical physics (RIMMP)." "Arno Kuijlaars" Analysis "A-DATADRIVE-B: Advanced Data-Driven Black-box modelling." "Johan Suykens" "ESAT - STADIUS, Stadius Centre for Dynamical Systems, Signal Processing and Data Analytics" "Making accurate predictions is a crucial factor in many systems (such as in modelling energy consumption, power load forecasting, traffic networks, process industry, environmental modelling, biomedicine, brain-machine interfaces) for cost savings, efficiency, health, safety and organizational purposes. In this project we aim at realizing a new generation of more advanced black-box modelling techniques for estimating predictive models from measured data. We will study different optimization modelling frameworks in order to obtain improved black-box models. This will be done by specifying models through constrained optimization problems and studying different candidate core models (parametric models, support vector machines and kernel methods) together with additional sets of constraints and regularization mechanisms. Different candidate mathematical frameworks will be considered. Several aspects that are relevant to black-box models will be studied including incorporation of prior knowledge, structured dynamical systems, tensorial data representations, interpretability and sparsity, and general purpose optimization algorithms. The methods should be suitable for handling large data sets and high dimensional input spaces. The final goal is also to realize a next generation software tool that can be generically applied to data from different application areas. The project A-DATADRIVE-B aims at getting end-users connected to the more advanced methods through a user-friendly data-driven black-box modelling tool.Further information: https://www.esat.kuleuven.be/stadius/ADB/" "Advanced modelling and characterization for power semiconductor materials and technologies" "Martin Diehl" "Numerical Analysis and Applied Mathematics (NUMA)" "The proposed project AddMorePower aims to advance X-ray- and electron-probe related characterization techniques to make them quantitative and automated tools for the power semiconductor industry, and to refine modelling (using MODA) and FAIR data-management methods to enhance and efficiently use characterization data (using CHADA). Thereby, AddMorePower will promote the materials integration and development for European power semiconductor technologies, to allow a broader and faster market penetration, while also providing new opportunities for other industries basing themselves on mono- and polycrystalline materials. With the rapid and massive spread of power electronics and power semiconductors to enable the digitalization and the electrification of our society and its supply with sustainable energy, new requirements arise to the conception and integration of semiconductor and interconnect materials. AddMorePower will provide the necessary characterization and modelling techniques that meet the particular needs of the upcoming power semiconductor technology generations:1. The transition to the new semiconductor materials gallium nitride (GaN) and silicon carbide (SiC), mainly limited by defects in the crystal lattice, for which currently no established characterization workflows exist.2. The starting 3D-integration also of power devices, posing severe thermo-mechanical challenges to the involved metals and intermetallic materials, which can only be mastered by understanding gained by predictive modelling.3. The trend towards digitalization and industry4.0 which requires FAIR (findable, accessible, interoperable and reusable) data at all development and production steps.The project brings together renowned research institutes with many years of experience in electron- and X-ray characterization, emerging new research groups and company start-ups and researchers with a track record in multi-physics materials modelling as well as data engineering." "Translational and Transdisciplinary research in Modelling Infectious Diseases" "Niel HENS" "Centre for Statistics" "TransMID focuses on the development of novel methods to estimate key epidemiological parameters from both serological and social contact data, with the aim to significantly expand the range of public health questions that can be adequately addressed using such data. Using new statistical and mathematical theory and newly collected as well as readily available serological and social contact data (mainly from Europe), fundamental mathematical and epidemiological challenges as outlined in the following work packages will be addressed: (a) frequency and density dependent mass action relating potential effective contacts to transmission dynamics in (sub)populations of different sizes with an empirical assessment using readily available contact data, (b) behavioural and temporal variations in contact patterns and their impact on the dynamics of infectious diseases, (c) close contact household networks and the assumption of homogeneous mixing within households, (d) estimating parameters from multivariate and serial cross-sectional serological data taking temporal effects and heterogeneity in acquisition into account in combination with the use of social contact data, and (e) finally the design of sero- and social contact surveys with specific focus on serial cross-sectional surveys. TransMID is transdisciplinary in nature with applications on diseases of major public health interest, such as pertussis, cytomegalovirus and measles. Translational methodology is placed at the heart of TransMID resulting in the development of a unifying methodology for other diseases and settings. The development of a toolbox and accompanying software allow easy and effective application of these fundamentally improved techniques on many infectious diseases and in different geographic contexts, which should maximize TransMID's impact on public health in Europe and beyond." "Exploring Duality for Future Data-driven Modelling" "Johan Suykens" "Dynamical Systems, Signal Processing and Data Analytics (STADIUS)" "Future data-driven modelling is increasingly challenging for many systems due to higher complexity levels, such as in energy systems, environmental and climate modelling, traffic and transport, industrial processes, health, safety, and others. This requires powerful concepts and frameworks that enable the design of high quality predictive models. In this proposal E-DUALITY we will explore and engineer the potential of duality principles for future data-driven modelling. An existing example illustrating the important role of duality in this context is support vector machines, which possess primal and dual model representations, in terms of feature maps and kernels, respectively. Within this project, besides using existing notions of duality that are relevant for data-driven modelling (e.g. Lagrange duality, Legendre-Fenchel duality, Monge-Kantorovich duality), we will also explore new ones. Duality principles will be employed for obtaining a generically applicable framework with unifying insights, handling different system complexity levels, optimal model representations and designing efficient algorithms. This will require taking an integrative approach across different research fields. The new framework should be able to include e.g. multi-view and multiple function learning, multiplex and multilayer networks, tensor models, multi-scale and deep architectures as particular instances and to combine several of such characteristics, in addition to simple basic schemes. It will include both parametric and kernel-based approaches for tasks as regression, classification, clustering, dimensionality reduction, outlier detection and dynamical systems modelling. Higher risk elements are the search for new standard forms in modelling systems with different complexity levels, matching models and representations to system characteristics, and developing algorithms for large scale applications within this powerful new framework." "Dynamic MOdelling of REsilience" "Inez Germeys" "Contextual Psychiatry" "The project DynaMORE (Dynamic MOdelling of REsilience) exploits advanced mathematical modelling for the promotion of mental health and well-being. DynaMORE will generate the first personalised in-silico model of mental health in the face of adversity, that is, stress resilience. The model will be based on and validated against unique multi-scale longitudinal real-world empirical data sets. The development and testing of the model will substantially deepen our scientific understanding of the mechanisms of resilience. These insights will allow us to create mechanistically targeted interventions for the primary prevention of stress-related disorders. On this basis, DynaMORE will develop an entirely new mHealth product that will include model-based prognostic tools for monitoring of at-risk subjects and for automatised decision-making about timed, personalised interventions. The DynaMORE model will be openly available, whereas the interventions and surrounding application solutions (data analysis tools, information and communication technology (ICT)) will have a strong potential for commercial exploitation, which will be actively prepared within DynaMORE.These objectives will be achieved through the interdisciplinary collaboration of world-leading experts in the fields of (1) computer-modelling and simulation and multi-scale data integration (project area MODELLING), (2) bio-psycho-social research on mental health and resilience (area HUMAN LIVES), and (3) mHealth technology, including ambulatory monitoring and intervention, and ICT (area TECHNOLOGY). These will be supported, and results will be exploited, by an experienced management and impact maximisation team (area IMPACT).The overall aim of DynaMORE is to improve human lives by preventing stress-related mental health problems. DynaMORE’s health-focussed strategy will promote individual well-being and reduce healthcare demands and indirect economic costs, by contributing to a healthy and productive workforce." "Innovation in modelling Placenta for Maternal and Fetal Health" "Jan Deprest" "Urogenital, Abdominal and Plastic Surgery" "The placenta is the least understood human organ but arguably one of the most important”-NIH Human Placenta Project . Innovation in Modelling Placenta for Maternal and Fetal Health (iPLACENTA) is a European Training Network (ETN), and will act as a springboard for promoting international, intersectoral and multi/inter-disciplinary training, career development and collaboration of fifteen early-stage researchers (ESRs) in Maternal and Fetal Health. iPLACENTA will improve our ability to study, model and visualise the placenta. iPLACENTA focuses on doctoral-level training and will be delivered by eleven participating universities located in ten different European countries. It coordinates research and training collaboration among world-leading academic institutions in Europe, providing a new combination of in-depth international expertise. iPLACENTA’s unique network aims are to improve our ability to study the placenta through in vitro and mathematical modelling (WP1&2). Whilst enhancing visualisation and assessment of the placenta in animal models and the clinic, thus enhancing investigation and prognosis of complicated pregnancies. To link the research to industrial exploitation, the network brings together four different businesses; two established companies as beneficiaries Mimetas (MIM) (organ-on-a-chip developer) and Moor, a clinical-technology specialist, together with two partners that are industrial global brands Samsung (SAM) and FujifFilm VisualSonics (FUJ). Together they will work with academics and clinicians to develop new placenta-on-a-chip technology (WP1), in silico placenta modelling (WP2), new modalities of laser technology to visualise the placenta in vivo (WP3), improve maternal-cardiovascular assessment and validate novel ultrasound tools for diagnosis of complicated pregnancies (WP3). Cross-sectorial training delivered by Business and Law Schools, Industry, Clinical specialist and European leaders in OpenScience-OpenInnovation." "Computer modelling and experimental validation of plasmas and plasma- surface interactions, for a deep insight in cryogenic etching (Cryoetch)." "Annemie Bogaerts" "Plasma Lab for Applications in Sustainability and Medicine - Antwerp (PLASMANT)" "Microchips have caused a revolution in electronics over the last few decades. Following Moore's law, much effort has been put into continuously shrinking electronic feature dimensions. Indeed, typical feature sizes of semi-conductors decreased from 10 μm in 1971 to 14 nm in 2014. With the shrinkage of feature sizes, plasma etching plays a more and more important role due to its anisotropy during surface processing. However, to go beyond 14 nm features, current state-of-the-art plasma processing faces significant challenges, such as plasma induced damage. Recently, one such novel process with limited plasma damage is cryogenic etching of low-k material with SF6/O2/SiF4 and CxFy plasmas. In this project, the fundamental mechanisms of the plasma, and its interaction with the surface, for these gas mixtures, will be studied to improve cryogenic plasma etching. For this purpose, numerical models (a hybrid Monte Carlo - fluid model and molecular dynamics model) will be employed to describe (i) the plasma behavior for SF6/O2/SiF4 and CxFy gas mixtures applied for cryogenic etching, and (ii) the surface interactions of the plasma species with the substrate during etching. Furthermore, cryogenic etch experiments will also be conducted to validate the modeling results." "A Synergetic Training Network on Energy beam Processing: from Modelling to Industrial Applications" "Patrick De Causmaecker" "Computer Science, Kulak Kortrijk Campus" "With the use of more advanced, but difficult-to-cut materials, on ever-more sophisticated products, the need to further develop and utilise the particular capabilities of the energy beam (EB) processing techniques seem to become a key enabler for the European industry. Although they are of various nature, a set of key communalities can be identified among EB methods when considered as dwell-time dependent processes; this allows the treatment of EB processes under a unitary technology umbrella.In this context, and based on a multidisciplinary pool of knowledge, the STEEP ITN aims to establish a European research training platform to enable a holistic approach of the EB processing methods. A number of 28 academic/research and industry partners with multidisciplinary & complementary expertise will set the first common European training programme that will take the technology from the modelling & validation of its key aspects (i.e. beam footprint) to the development of simulation tools (i.e. beam path simulator) and the demonstration (e.g. on various EB workstations) by generating freeform surfaces.This wide breath of topics will be the vehicle to train European researchers in complementary (e.g. maths – material processing – computing – machine simulation/control) areas and environments (academic, industrial) of EB processes so that a sustainable evolution of this group of technologies is achieved." "Generally Accepted Reliability Principle with Uncertainty modelling and through probabilistic Risk assessment." "Dirk Van Hertem" "ESAT - ELECTA, Electrical Energy and Computer Architectures, Research Centre of Energy, Transport and Environment, Leuven" "Power system reliability management means to take decisions under increasing uncertainty (for instance, related to renewable generation). It aims to maintain power system performance at a desired level, while minimizing the socio-economic costs of keeping the power system at that performance level. Seven TSOs (Belgium, Bulgaria, Czech Republic, Denmark, France, Iceland, Norway), together with eleven RTD performers, propose the four year GARPUR research project. GARPUR designs, develops, assesses and evaluates new system reliability criteria and management while maximizing social welfare as they are progressively implemented over the next decades at a pan-European level. The new management methodologies encompass multiple business activities (system development, asset management, power system operation) that, in turn, ensure coherent decision-making at the respective time horizons. These methodologies also involve mathematical and computational models to predict the location, duration and amount of power supply interruptions. Five alternatives to improve reliability management of the pan-European power system are studied. After practical validation by the TSOs, these alternatives are analysed with the help of a quantification platform. Pilot tests of the new proposed reliability criteria are performed by individual TSOs or (when appropriate) a group of TSOs using this quantification platform, either in a given control zone or (where appropriate) throughout the pan-European system. Reliability criteria are compared and presented to the TSO community and regulatory authorities who establish the robustness of the results. Dissemination activities of the new reliability criteria are supported by a Reference Group of TSOs and address all the key electricity market stakeholders. An implementation roadmap is delivered for the deployment of the resulting technical and regulatory solutions to keep the pan-European system reliability at optimal socio-economic levels."