Archives ouvertes Hal IMT Atlantique
Not a member yet
20414 research outputs found
Sort by
Probing of EoS with clusters and hypernuclei
No full textInternational audienceThe study of the nuclear equation-of-state (EoS) is a one of the primary goals of experimental and theoretical heavy-ion physics. The comparison of recent high statistics data from the STAR Collaboration with transport models provides a unique possibility to address this topic in a yet unexplored energy domain. Employing the microscopic n-body Parton-Hadron-Quantum-Molecular Dynamics (PHQMD) transport approach, which allows to describe the propagation and interactions of hadronic and partonic degrees of freedom including cluster and hyper-nucleus formation and dynamics, we investigate the influence of different EoS on bulk observables, the multiplicity, and rapidity distributions of protons, s and clusters up to A=4 as well as their influence on the collective flow. We explore three different EoS: two static EoS, dubbed 'soft' and 'hard', which differ in the compressibility modulus, as well as a soft momentum dependent EoS. We find that a soft momentum dependent EoS reproduces most baryon and cluster observables, including the flow observables, quantitatively
Real-Time Calculation of 3D Computer-Generated Holograms by Phase Space Warping
No full textModeling light propagation from a 3D structure to a plane in a medium with a constant refractive index, according to scalar diffraction theory, is essential in many optical fields, particularly in computer-generated holography for 3D visualization. However, this process is computationally expensive, often scaling with the geometry complexity of the 3D scene or requiring compromises in computational rigor. To overcome this issue, we propose to model this phenomenon in phase space, which represents both spatial and spatial frequency information in a joint domain. By decomposing wavefields into sets of Gaussian beams, our formulation provides a more precise control over light propagation, significantly reducing the dependency on scene geometry. We implemented our method in CUDA, leveraging acceleration with GPU. Our experiments reported up to a 26.1fold speedup compared to the conventional layer-based methods, while maintaining a similarity of 35.1 dB in PSNR
Closing the Loop in Embedded Security: Evolution of an AIOps Framework for Threat Hunting
No full textInternational audienceAs the 6G and IoT eras usher in a massive deployment of bare-metal embedded devices, securing resourceconstrained hardware like the ARM Cortex-M4 remains a critical challenge. Without the protection layers of a traditional operating system, these devices are uniquely susceptible to buffer overflows and memory corruption via standard C library functions (e.g., memcpy, strncat). While AI has emerged as a powerful tool for detecting these vulnerabilities via side-channel data like power and instruction traces, the operational challenge lies in the lifecycle management of these models.This paper presents an AIMLOps-driven approach to embedded security, using detection of vulnerable library functions to demonstrate AI pipeline evolution. We move beyond "one-off" model training to explore how an integrated MLOps framework enables sophisticated data analytics for high-frequency timeseries and real-time execution traces previously inaccessible to security operators. We detail the transition from manual feature engineering to autonomous pipeline stages: data ingestion from hardware probes, model validation for edge deployment, and continuous adaptation to windowing size and sampling frequency. Results show that treating the security model as an evolving asset within a managed lifecycle achieves high detection accuracy and robustness for zero-day vulnerability detection. This work provides a blueprint for managing AI workloads in embedded system time-series telemetry.</div
Conception et réalisation de systèmes de projection rétinienne intégrables dans une lentille de contact
No full textWith recent advances in smart contact lenses, research into a lens capable of projecting light into the eye seems be a natural progression that opens up for new medical applications and novel forms of augmented reality. In this context, this thesis focuses on miniaturising a light guide embedded in a scleral contact lens, enabling simple functions with out compromising the wearer’s vision. The limited available volume and the placement of the components at the periphery introduce several constraints that must be addressed. Initially, the use of diffuse incoherent sources combined with a simple ring guide allows for uniform illumination of the retina, which can help to slow myopia. Two versions of the guide were designed and manufactured on a diamond machining lathe, then encapsulated in a contact lens. The assembly was tested to validate the design and ensure ocular safety and uniform illumination. In the second phase, a more complex ring guide was designed, manufactured, and tested. Paired with a directional VCSEL source and an EOD, this enables the projection of light symbols onto the retinal plane. The EOD was produced by photo-tracing and transferred to the ring using a copying system with a mask aligner. Finally, a discussion of the limitations of these devices is presented, along with opportunities for further research.Avec les récents progrès des lentilles de contact intelligentes, la recherche sur une lentille capable de projeter de la lumière dans l’il apparaît comme une évolution naturelle qui ouvre la voie à de nouvelles applications médicales et de nouvelles formes de réalité augmentée. Dans ce contexte, ce travail de thèse traite de la miniaturisation d’un guide de lumière embarqué dans une lentille de contact sclérale, permettant d’assurer des fonctions d’imagerie simples sans impacter la vue du porteur. Le faible volume disponible et le positionnement des composants en périphérie engendrent un certain nombre de contraintes. Ceci sera illustré, dans un premier temps, où l’utilisation de sources diffuses incohérentes associées à un guide annulaire simple permettra une illumination uniforme de la rétine favorable au ralentissement de la myopie (Repeated Low-level Red Light Therapy). Deux versions de guide ont été conçues, réalisées sur un tour à usinage diamant, puis encapsulées dans une lentille de contact. L’ensemble a été testé pour valider la conception et les critères de sécurité oculaire et d’uniformité d’éclairement. Dans un second temps, un guide annulaire plus complexe a été conçu, réalisé et testé. Associé à une source VCSEL et à un élément optique diffractif (EOD), il permet d’imager des symboles lumineux dans le plan rétinien. L’EOD a été réalisé par phototraçage et reporté sur l’anneau en utilisant un système de recopie avec un aligneur de masque. Enfin, une discussion sur les limites de ces dispositifs est proposée ainsi que des pistes d’amélioration
Source term distribution and mobility – WP3 results of EU project SAMOSAFER
No full textInternational audienceThe aim of the EU Horizon 2020 founded project SAMOSAFER was to develop and demonstrate new safety barriers and a more controlled behavior in severe accidents of the Molten Salt Reactor (MSR). The aim of work package 3 was to develop and validate models for tracing the nuclides carrying the radiotoxicity and decay heat, i.e. the source term and their chemical form and mobility during nominal and accidental conditions. The source term distribution at the beginning of an accident depends on the foregoing nominal operation mode. Hence, the liquid salt reprocessing and gaseous and insoluble Fission Products (FPs) separation was modeled before the severe accident simulation were accomplished. Because the source term assessment is a complex topic, only selected phenomena were addressed. As a reference system the MSFR concept was adopted from the previous EU project SAMOFAR. The study included benchmarking of the burnup tools, thermal-hydraulics simulations to confirm the removal rates to the off-gas system. Chemical experiments and calculations of the methods used in the reprocessing unit, simulation of severe accident conditions and finally calculation of the source term distribution. The major outcomes are the benchmark for the nuclide inventory tracing tools, application of multi-physics tools on He-bubbling for gaseous and solid FPs removal rates simulations, refinement of the reprocessing scheme, and first simulation of simplified severe accident
All-sky Searches for Continuous Gravitational Waves from Isolated Neutron Stars in the Data from the First Part of the Fourth LIGO-Virgo-KAGRA Observing Run
No full textInternational audienceWe present results from an all-sky search for continuous gravitational waves, using three different methods applied to the first eight months of LIGO data from the fourth LIGO-Virgo-KAGRA Collaboration s observing run. We aim at signals potentially emitted by rotating, non-axisymmetric isolated neutron star in the Milky Way. The analysis spans a frequency range from 20 Hz to 2000 Hz and accommodates frequency derivative magnitudes up to Hz/s. No statistically significant periodic gravitational wave signals were detected. We establish 95% confidence-level (CL) frequentist upper limits on the dimensionless strain amplitudes. The most stringent population-averaged strain upper limits reach 9.7 near 290 Hz, matching the best previous constraints from 250 to 1700 Hz while extending coverage to a much broader spin-down range. At higher frequencies, the new limits improve upon previous results by factors of approximately 1.6. These constraints are applied to three astrophysical scenarios: 1) the distribution of galactic neutron stars as a function of spin frequency and ellipticity; 2) the contribution of millisecond pulsars to the GeV excess near the galactic center; and 3) the possible dark matter fraction composed of nearby inspiraling primordial binary black holes with asteroid-scale masses
Conception et dimensionnement d’un réseau de transport sous incertitudes
No full textNational audienceNous étudions un problème de localisation d'entrepôts avec stocks multi-commodités sous incertitudes. Nous adoptons une approche stochastique basée sur des scénarios avec contraintes probabilistes. Deux méthodes de résolution sont proposées : un modèle intégré et une décomposition de Benders. Une première approche est mise en place pour améliorer les performances de la méthode de décomposition. Les résultats préliminaires montrent que la décomposition fournit rapidement de bonnes solutions
Development of separation method for difficult-to-measure radionuclides from nuclear decommissioning
No full textInternational audienc
La Fabrique des Jumeaux Numériques: verrous scientifiques, technologiques et perspectives du programme EDT
No full textNational audienceThe current transition from traditional physical systems to software-driven ecosystems is profoundly transforming engineering practices and increasing the demand for accessible and scalable technologies. Digital twins (DTs) are emerging as a foundational concept, enabling the separation of high-level decision-making from physical operations while maintaining a close connection to the real system. By combining deductive models from engineering and inductive models derived from data, they provide powerful capabilities for analysis, simulation, prediction, and optimization of systems. Despite this potential, their implementations often remain fragmented, domain-specific, and costly to deploy and maintain. This article synthesizes the scientific and technological challenges associated with digital twin engineering and outlines the objectives of the Engineering Digital Twins (EDT) program, which aims to establish a rigorous and reliable approach to their development, deployment, and operation.La transition des systèmes physiques traditionnels vers des écosystèmes pilotés par logiciel transforme profondément les pratiques d'ingénierie. Les jumeaux numériques (JN) émergent comme un concept structurant permettant de dissocier la prise de décision de haut niveau des opérations physiques tout en maintenant un lien étroit avec le système réel. En combinant des modèles déductifs issus de l'ingénierie et des modèles inductifs dérivés des données, ils offrent des capacités puissantes d'analyse, de simulation, de prédiction et d'optimisation des systèmes. Malgré ce potentiel, leurs implémentations restent souvent fragmentées, spécifiques à un domaine et coûteuses à déployer et à maintenir. Cet article propose une synthèse des verrous scientifiques et technologiques associés à l'ingénierie des jumeaux numériques et présente les objectifs du programme Engineering Digital Twins (EDT), qui vise une approche rigoureuse et fiable de leur développement, de leur déploiement et de leur exploitation
Differentiable forward modeling and inverse lithography for two-photon lithography: application to voxel-by-voxel, diffractive optical element, and spatial light modulator systems
No full textInternational audienceBackground: Two-photon polymerization (TPP) enables additive manufacturing with sub-micron resolution, but fabrication accuracy is strongly limited by nonlinear exposure dynamics, radical transport, and chemical proximity effects, especially in parallelized writing setups. Simplified threshold models neglect the resist response, whereas compact models often omit key physical and time-dependent processes. Moreover, such models are typically non-differentiable, which prevents their use in solving inverse design problems.Aim: We address these challenges by developing and validating a time-dependent, differentiable forward model of the TPP process and by demonstrating how it enables optimization approaches for inverse problems across different exposure conditions. Approach:We implement the forward model in PyTorch, combining a precomputed point spread function, radical generation, oxygen quenching, diffusion, polymerization kinetics, and resist development. Based on this model, we establish two optimization strategies: (i) a gradient-based point-by-point backpropagation method to optimize power along scan paths and (ii) a hybrid approach where a neural network predicts an initial exposure strategy that is refined by the physics-based model. These methods are applied to three experimental configurations: point-by-point writing, diffractive optical element (DOE)-based parallel exposure, and spatial light modulator (SLM) projection. Results:We demonstrate that the forward model, combined with inverse optimization, derives exposure distributions directly from target structures and improves fabrication accuracy in all three cases. In point-by-point writing, backpropagation adjusts power to control feature height and connectivity in three-dimensional structures. In DOE-based parallelization, the framework predicts line merging and height loss at the start and end of lines caused by oxygen inhibition and compensates for them through local dose control. In SLM projection with a low-contrast resist, the U-Net-assisted workflow produces optimized exposure distributions that enable the fabrication of microlens arrays and freeform features consistent with the design targets within the material's resolution limits. Conclusions:The presented time-dependent, differentiable forward model enables gradient-based optimization of exposure conditions across different TPP setups. This provides a scalable tool for dose correction and lays the foundation for physics-based inverse lithography in TPP.</div