Université de technologie de Troyes open archive
Not a member yet
10722 research outputs found
Sort by
Tungsten-induced atypical multilayer growth in a new amorphous Al-Ti-(W) thin films: A key to improve corrosion resistance in saline environment
No full textInternational audienc
A Multi-Feature Image-Specific Pixel Classification Approach for Weakly Supervised Semantic Segmentation
No full textInternational audienc
Broadening the Plasmonic Spectral Range of Metallic Metasurfaces Using Dual-Material Arrays
No full textInternational audiencePlasmonic nanoparticles support localized surface plasmon resonances (LSPR), exhibiting broad spectral features determined by nanoparticle geometry and material. When periodically arranged, they enable diffractive modes linked to Rayleigh anomalies (RA). The coupling between LSPR and RA produces surface lattice resonances, which combine strong field confinement, enhanced scattering, and high resonance quality with broad applications in sensing and light emission. While monometallic arrays have been extensively studied, bimetallic systems remain largely unexplored. This work investigates checkerboard arrays of gold and aluminum nanoparticles and demonstrates broadband surface lattice modes arising from hybridization of their distinct plasmonic responses. Simulations and experiments show that gold and aluminum dipolar resonances couple via the same RA in the near-infrared, extending beyond aluminum's usual plasmonic range. This previously unreported mechanism opens new routes toward broadband, high-performance plasmonic arrays for sensing, photonics, and telecommunications
Legitimacy of life cycle assessment for sustainable communities
No full textInternational audienceLife Cycle Assessment (LCA) has been developed within companies with high environmental impacts, and is widely used in industrial contexts. More recently, LCA has gained importance in sustainable communities. Thus, in this paper we aim to answer the question: how legitimate is the LCA method for sustainable communities, regarding their values, practices and goal? For this purpose, we conducted an experiment rooted in a sustainable community. We observed the work of a focus group composed of seven residents of a sustainable community, who conducted an LCA with the support of an LCA expert (the researcher conducting the study). Each focus group session was recorded and analysed using a qualitative method (thematic coding). Focusing on moral and cognitive legitimacy, our study highlighted three main themes: the tracking of material flows within a community, the notion of complexity, and the dissemination of results. We identified certain tensions between LCA practices and the residents’ beliefs. Nonetheless, the output of LCA could still be valuable for residents, especially in terms of communicating their alternative way of life. It is also worth noting that the barriers to LCA adoption are similar in both industrial contexts and sustainable communities. We conclude on the need to better understand the role of LCA as a strategy for sustainable communities to upscale their way of life, while also acknowledging the persistent (and necessary) tensions in terms of values and organization
Success probability in Shor's Algorithm
No full textThis paper aims to determine the exact success probability at each step of Shor's algorithm. Although the literature usually provides a lower bound on this probability, we present an improved bound. The derived formulas enable the identification of all failure cases in Shor's algorithm, which correspond to a success probability of zero. A simulation routine is provided to evaluate the theoretical success probability for a given integer when its prime factorization is known with potential applications in quantum resource estimation and algorithm benchmarking
Pricing des options européennes sur devises: modèles classiques, limites statistiques et apports du Machine Learning
No full textLe pricing d'options sur devises constitue un problème central en finance quantitative, mobilisant des outils issus des probabilités, statistiques et de l'optimisation. Les modèles classiques, tels que Black-Scholes et son extension de Garman-Kohlhagen, offrent un cadre théorique élégant et largement utilisé en pratique. Toutefois, ces modèles reposent sur des hypothèses fortes, en particulier l'hypothèse de volatilité constante, qui sont contredites par les données de marché. Ce rapport propose une présentation progressive du pricing des options européennes sur devises, en mettant en évidence les limites des modèles analytiques classiques, puis en établissant les liens naturels avec les méthodes statistiques. Enfin, des pistes d'amélioration basées sur le Machine Learning sont discutées, dans une optique d'ingénierie financière et non de recherche théorique avancée.</div
Conditions d’insertion des opérateurs intermédiaires aux systèmes alimentaires territorialisés de la Grande Région : rôle des acteurs publics et liens aux transitions
No full textNational audienc
Large Language Models to Enhance Multi-task Drone Operations in Simulated Environments
No full textBenefiting from the rapid advancements in large language models (LLMs), human-drone interaction has reached unprecedented opportunities. In this paper, we propose a method that integrates a fine-tuned CodeT5 model with the Unreal Engine-based AirSim drone simulator to efficiently execute multi-task operations using natural language commands. This approach enables users to interact with simulated drones through prompts or command descriptions, allowing them to easily access and control the drone's status, significantly lowering the operational threshold. In the AirSim simulator, we can flexibly construct visually realistic dynamic environments to simulate drone applications in complex scenarios. By combining a large dataset of (natural language, program code) command-execution pairs generated by ChatGPT with developer-written drone code as training data, we fine-tune the CodeT5 to achieve automated translation from natural language to executable code for drone tasks. Experimental results demonstrate that the proposed method exhibits superior task execution efficiency and command understanding capabilities in simulated environments. In the future, we plan to extend the model functionality in a modular manner, enhancing its adaptability to complex scenarios and driving the application of drone technologies in real-world environments
Ordonnancement avec acceptation sous contraintes énergétiques glissantes
No full textInternational audienceOrdonnancement avec acceptation sous contraintes énergétiques glissante
DonUTT : Un outil partagé pour mesurer la soutenabiltié
No full textHow can the sustainability of a university be measured? How can we know whether an institution is evolving within a space that is both ecologically safe and socially just? DonUTT, an adaptation of Kate Raworth’s Doughnut model to the context of the University of Technology of Troyes (UTT), provides a structured response to these questions. The method simultaneously assesses pressures on the environment (the “ecological ceiling”) and the extent to which fundamental human needs are met (the “social foundation”).Two main challenges guided its development. The first was to translate a global framework—the planetary boundaries—into a local, campus-scale context. The second was to rethink the social foundation so as to include psychological and relational well-being, which are essential to understanding the lived experience of members of the UTT community. Measuring these dimensions required explicit methodological choices: selecting boundaries, defining thresholds, collecting data, and constructing indicators aligned with the concept of sustainability.This participatory approach was designed to be reproducible by other institutions. Existing doughnut models served as a source of inspiration while also highlighting the need for a renewed approach that is more consistent with UTT’s specific reality. DonUTT made it possible to identify major trends. The results reveal a contrasted situation: several social boundaries lie within sustainable zones, while certain ecological boundaries—such as climate and land use—exceed the defined thresholds. These findings are not a sanction; they help identify room for improvement and prioritize areas for action.DonUTT thus stands as an evolving and structuring tool, a starting point for guiding future transformations and engaging UTT on a trajectory that is more just, safer, and more sustainable.Comment mesurer la soutenabilité d'une université ? Comment savoir si un établissement évolue dans un espace à la fois sûr sur le plan environnemental et juste sur le plan social ? Le DonUTT, adaptation du modèle du Donut de Kate Raworth au contexte de l’université de technologie de Troyes (UTT), apporte une réponse structurée à ces questions. La méthode évalue simultanément les pressions exercées sur l’environnement (le « plafond écologique ») et la satisfaction des besoins humains fondamentaux (le « plancher social »). Deux défis ont guidé son élaboration. Le premier a consisté à transposer à l’échelle locale un cadre global, celui des limites planétaires. Le second a été de repenser le plancher social pour inclure le bien-être psychologique et relationnel qui sont indispensables pour appréhender le vécu des membres de la communauté de l'UTT. La mesure de ces dimensions a nécessité des choix méthodologiques explicites : sélection des limites, définition de seuils, collecte de données et construction d'indicateurs alignés sur la notion de soutenabilité. Cette démarche participative a été conçue pour être reproductible par d’autres établissements. Les donuts existants ont servi de source d'inspiration tout en mettant en lumière la nécessité d'une approche renouvelée, plus cohérente avec la réalité de l'UTT. Le DonUTT a permis de faire émerger de grandes tendances. Les résultats montrent une situation contrastée : plusieurs limites sociales se situent dans des zones soutenables, tandis que certaines limites écologiques, comme le climat ou l'usage des sols, dépassent les seuils fixés. Ces constats ne sont pas une sanction, ils permettent d'identifier les marges de progression et de hiérarchiser les priorités d'action. Le DonUTT s’inscrit ainsi comme un outil évolutif et structurant, un point de départ pour guider les transformations futures et engager l’UTT dans une trajectoire plus juste, plus sûre et plus soutenable