Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
Magnetotactic Bacteria in the Medical Field versus Magnetic Nanoparticles
No full textInternational audienceMagnetotactic bacteria (MTB) is a heterogenous group of prokaryotic organisms with a ubiquitous distribution and some common, specific attributes. Magnetosomes are a distinctive feature of magnetotactic bacteria from other prokaryotic organisms and contain membrane-encased magnetic crystals arranged in chain-like structures that allow the cell to passively align itself along magnetic fields (magnetotaxis) [1,2]. Magnetic nanoparticles have a high popularity among nanotechnological processes and applications due to their structural, physicochemical properties and magnetic composition. They have a wide range of biotechnological uses in applications such as biomedicine, drug delivery, MRI imaging, cancer theranostics, biosensors, catalysis and bioseparation. Their use is limited by concerns related to biocompatibility risks – toxicity, mutagenicity and immunological rejection, low stability and not enough extended functionality in numerous therapies, treatments or devices [3]. Magnetic nanoparticles from magnetotactic bacteria (MTB-NPs) are much more promising compared to MNPs mainly due to their improved characteristics in terms of biocompatibility, functionalization and sustainability. Among the most important advantages we mention biocompatibility, reduced risk of toxicity and immunological rejection, increased stability and extended functionality in numerous therapies, treatments or devices [4]. Considering these aspects, MTB have multiple applications in the medical field, such as precised, targeted and personalized medicine thanks to reduced side effects, improved treatment efficacy and long-term survival rates, treatment for hard-to-treat diseases and development of new drug classes [5]. As part of our review, we performed a comparative analysis of recent studies highlighting the specific mechanisms through which MTB contribute to targeted drug delivery, and proposed a conceptual framework that integrates current findings with potential clinical applications
Failure Detection Under Sensing Uncertainty in Vehicular Systems
No full textInternational audienceThe advancement of autonomous vehicles depends on the integration of sophisticated sensors that perceive their environment in real-time. However, these sensors are prone to uncertainties that, if not properly managed, could undermine the vehicle’s decision-making process and elevate road safety risks. In response to this challenge, we present an embeddable solution designed to filter uncertain data using a machine learning model. Our solution not only resists to inconsistencies in sensor data but also quantifies the degree of anomaly in sensor outputs. This capability is essential for enabling effective predictive maintenance by identifying potentially failing systems before they compromise the overall performance. Through simulations, we demonstrate that the anomaly system detector enhances the vehicle’s resilience against discordant data
Directional light scattering in Mie-resonant Si particles with ultra-thin plasmonic shells
No full textInternational audienceMetamaterial research has sought to create nanostructures with strong directional optical scattering to control light propagation at the nanoscale. Core-shell architectures comprised of both resonant cores and resonant shells have been suggested as candidate particles in which the spectral overlap of the electric and magnetic dipoles can be controlled to create strong directional scattering. In this study, we present Au-decorated Si core-shell (Si@Au) particles. These were synthesized by first creating Si particles through the thermal disproportionation of hydrogen silsesquioxane (HSQ), which were then decorated with ∼ 4 nm diameter Au nanoparticles. We characterized the resonant behavior of the core-shell particles using electron energy-loss spectroscopy mapping and optical single-particle scatter spectroscopy. These observations were supported by T-matrix simulations and Mie theory calculations of the scattering spectra, which show that compared to Si, Si@Au particles demonstrate a dampened magnetic dipole resonance for smaller Si core diameters (100 -130 nm) and an enhanced magnetic dipole resonance for larger Si core sizes (150 -200 nm). However, we show that to significantly improve forward scattering intensity, continuous plasmonic shells of ~12 nm thickness are needed
Uncertainty-aware digital twin of business processes via Bayesian calibration and posterior-predictive simulation
No full textInternational audienceEvent logs are finite, partial views of a latent stochastic process. We present a Bayesian digital twin based on probabilistic, random-effects, event-by-event generators that utilize historical logs and propagate uncertainty. After calibration with Hamiltonian Monte Carlo, each posterior draw is a parameter vector that defines a complete simulator: using that we generate or continue event logs by sequentially sampling the next activity, the inter-event time, and new case arrivals conditional on history and crowding (congestion). Computing KPIs (cycle time, cost, directly-follows counts, etc.) on the simulated logs and aggregating over all posterior draws yields posterior-predictive KPI distributions. Validation compares these distributions to bootstrap baselines from the observed log using distributional distances. The result is a scenario-ready process twin that reports outcomes as distributions, enabling risk-aware decisions
Uncertainty-aware Digital Twin of Business Processes via Bayesian Calibration and Posterior-predictive Simulation
No full textInternational audienceEvent logs are finite, partial views of a latent stochastic process. We present a Bayesian digital twin based on probabilistic, random-effects, event-by-event generators that utilize historical logs and propagate uncertainty. After calibration with Hamiltonian Monte Carlo, each posterior draw is a parameter vector that defines a complete simulator: using that we generate or continue event logs by sequentially sampling the next activity, the inter-event time, and new case arrivals conditional on history and crowding (congestion). Computing KPIs (cycle time, cost, directly-follows counts, etc.) on the simulated logs and aggregating over all posterior draws yields posterior-predictive KPI distributions. Validation compares these distributions to bootstrap baselines from the observed log using distributional distances. The result is a scenario-ready process twin that reports outcomes as distributions, enabling risk-aware decisions
Contre le brouillard de la guerre : agir en connaissance de risques
No full textInternational audienc
Order acceptance scheduling under Time-Of-Use and energy constraint
No full textInternational audienc
Deep learning-driven spectral exploration for metal-insulator-metal plasmonic metasurfaces
No full textInternational audienceMetal-insulator-metal plasmonic metasurfaces exhibit intricate spectral responses arising from the interplay among localized surface plasmon polaritons, surface lattice resonances, and Fabry-Pérot cavity modes. However, traditional characterization methods relying on iterative electromagnetic simulations and manual spectral analysis face inefficiencies in handling complex parameter spaces and measurement-condition heterogeneity. Here, we present a deep learningdriven framework to analyze the spectral behaviors of metal-insulator-metal metasurfaces by integrating experimental fabrication, finite-difference time-domain simulations, and data-driven spectral classification and regression. Gradient-parameter metasurfaces with varying insulator gaps (20-200 nm), nanostructure geometries (disc/ring), and periodicities (500-1500 nm) are fabricated via electron-beam lithography and optically characterized under reflection/transmission configurations. Numerical simulations reveal the interplay of hybridized modes and surface charge dynamics. Leveraging convolutional and recurrent neural networks (CNNs, LSTMs, GRUs) and Transformers, we achieve robust classification of 24 structural categories and spectral regression for inverse design. Notably, LSTM models attain superior classification accuracy (>99.2%), while CNN demonstrates superior time efficiency. This work establishes a data-physics-integrated paradigm for rapid MIM metasurface characterization, and the proposed methodology bridges the gap between complex optical responses and deep learning-driven spectral exploration, advancing applications in label-free biosensing and tunable photonic systems
