Projets novateurs réalisés

Investigation of Data-Driven Koopman Model Predictive Control for Hydrogen-Diesel Engine Applications

This project explores the application of the Koopman Operator to system dynamics for the application within model predictive control of hydrogen-enhanced diesel engines. Investigating different implementation options and leveraging data-driven machine learning approaches, the aim is to reduce computational effort while enabling system-theoretic analysis of the dynamic system representation. The topic aligns closely with the interests of Prof. Jakob Andert (RWTH) and Prof. David Gordon (University of Alberta), whose expertise in machine learning-based control of energy conversion systems complements this work. Previous collaborations between these institutions have yielded valuable shared data and lab advancements. This project will further integrate RWTH’s machine learning modeling expertise with the University of Alberta’s embedded hardware knowledge, strengthening ongoing research efforts at both universities.

Research on neophytic plants in forest areas in Southern Germany

Investigations on sample introduction systems for spray drying

Optimizing Lightstage Capture for High Fidelity 3D Facial Reconstruction

Ubisoft is one of the world’s largest video game studios, specializing in 3D open-world games that require precise 3D character representations. In particular, achieving high-quality facial features is crucial, as humans are highly sensitive to small details in facial expressions. Currently, creating 3D
facial representations first requires a Lightstage capture pipeline. This process begins with taking highly detailed photographs of actors’ faces, which are then processed through a Multi-View Stereo (MVS) [1] algorithm to generate an extremely dense and irregular 3D mesh. The mesh is then registered and simplified into a regular mesh ready for use in game development. However, this pipeline is costly, requires significant manual labor, extensive memory storage, high computational power, and relies on third-party software. For instance, a 3D sequence of 20 minutes currently requires 150TB of storage, and 8 weeks of processing time with eleven working PCs. This project aims to optimize the MVS pipeline, helping Ubisoft save precious time as well as financial,
material (working PCs, GPUs, storage) and human resources (artists and performers). Specifically, the project will focus on creating improved metrics to measure MVS reconstruction quality of facial features and leverage those metrics to innovate in the field of 3D facial reconstruction.

Recherche de traceurs chimiques dans l’huile isolante permettant de déterminer indirectement la vie résiduelle des transformateurs de puissance

Un problème réel dans l’industrie des transformateurs de puissance est le manque d’outils et de méthodes précises pour déterminer l’état et la durée de vie de l’isolation principale. Ces appareils onéreux, composés d’enroulement de cuivre isolés par du papier et le tout baignant dans une huile isolante, représentent le coeur de tous réseaux électriques. Le but de ces stages est de déterminer des traceurs chimiques, capables de prédire le niveau de dégradation/vieillissement de l’isolation solide des transformateurs de puissance. Les opérateurs pourraient ainsi prendre des décisions appropriées afin d’éviter des pannes majeures. En effet, si l’on eut prévoir les pannes, il devient possible de les éviter! L’évaluation précise de l’état d’un transformateur permet de prévoir son remplacement au moment optimal. Un tel projet présente des intérêts évidents, pour l’avancement de la science et pour les exploitants de réseaux électriques qui disposeraient ainsi d’outils appropriés pour la planification de leurs investissements, dans le cadre du renouvellement des équipements existants.

Human Multi-Robot Interaction using Foundation Models

Hybrid slab solutions: Timber Concrete Composite research

Skeletal Editing of Fluorescent Heterocycles for Biomedical Applications

Fault detection in chemical processes using machine learning

Diffraction of atomic matter waves through 2D membranes

Spectral Imaging for Postoperative Complication Prediction

Design and development of refrigeration test bench for the experimental analysis of evaporators with low-GWP refrigerants