39936 research outputs found
Sort by
Análisis de Datos CabinPaq y Evaluación de Modelos Predictivos de Índice de Calidad del Aire basados en Redes Neuronales
No full textEste trabajo tiene como objetivo el de analizar la naturaleza, estructura y composición del conjunto de datos CabinPaq, así como preprocesar, evaluar y seleccionar el mejor modelo predictivo basado en redes neuronales para predecir el índice de calidad del aire (ICA) del día siguiente. Con la finalidad de cumplir este objetivo, se han llevado a cabo múltiples métodos entre los que destaca la extracción y análisis exploratorio de datos inicial sobre el conjunto de datos en diferentes niveles de granularidad. Posteriormente, se ha desarrollado una metodología “dummy” sobre datos validados del Ayuntamiento de Madrid con el objetivo de asegurar la realización de un procedimiento correcto, que posteriormente se va a aplicar sobre el conjunto de datos de estudio a nivel de punto de medición (DCI). Finalmente, se han evaluado los resultados presentados por cada modelo predictivo estudiado, siendo estos RNN Simple, LSTM, GRU y Transformer, seleccionando aquel que es capaz de presentar un mejor ajuste.
Los resultados obtenidos del desarrollo del proyecto han derivado en un análisis exhaustivo de ambos conjuntos de datos presentando las principales inconsistencias y los métodos empleados para solventar dicha situación. A su vez, se han presentado múltiples herramientas intermedias, procedimientos y modelos que han permitido comprender la naturaleza de los datos y el tratamiento correcto de los mismos, lo cual ha permitido a su vez, generar en cada metodología de trabajo un dataset consistente y robusto para el modelado. Por último, se ha podido observar el gran desempeño predictivo que presentan los modelos, precisamente los basados en redes neuronales recurrentes, donde se ha seleccionado en concreto LSTM y GRU debido a su capacidad de retención de información de memoria tanto en el corto plazo como en el largo plazo.
Para concluir, en términos generales a nivel de proyecto, se ha podido observar el gran potencial que presentan diferentes tratamientos de datos y modelos predictivos para determinar valores futuros; así como el impacto positivo que un buen desarrollo de esta metodología puede presentar tanto a nivel social como económico.Doble Grado en Ingeniería Informática y Administración de Empresa
The new ITER baseline, research plan and open R&D issues
Get PDFA new baseline (NB) has been proposed by the ITER Project to ensure a robust achievement of the Projects' goals, in view of past challenges including delays incurred due to the Covid-19 pandemic, technical challenges in completing first-of-a-kind components and in nuclear licensing. The NB includes modifications to the configuration of the ITER device and its ancillaries (e.g. change from beryllium to tungsten as first wall material, modification of the heating and current drive mix, etc.) as well as additional testing of components (e.g. toroidal field coils) or phased installation (start with inertially cooled first wall before later installation of the final actively water-cooled components) to minimise operational risks. In the NB, the ITER research plan (IRP) will be divided into three main phases: (a) start of research operation, with 40 MW of ECH and 10 MW of ICH, which will focus on the demonstration of 15 MA operation in L-mode, commissioning of all required systems, including disruption mitigation, and the demonstration of H-mode plasma operation in deuterium; (b) DT-1, with 60-67 MW of ECH, 33 MW of neutral beam injection (NBI) and 10-20 MW of ICH, which will demonstrate robust operation in high confinement H-mode plasmas in DT up to Q >= 10 and for burn durations of 300-500 s within an accumulated neutron fluence of 1% of the ITER machine's lifetime total, and; (c) DT-2, with up to 67 MW of ECH, up to 49.5 MW of NBI and up to 20 MW of ICH, with the ITER tokamak and ancillaries in their final configuration to demonstrate routine operation in DT plasmas at high Q and the Q>=5 long-pulse and steady-state scenarios to the final neutron fluence and to perform R&D on nuclear fusion reactor issues. The logic, physics basis, modelling and experimental evaluations carried out to support the NB and the associated IRP are described. These include the impact of the tungsten wall on plasma scenarios and associated risk mitigation measures, as well as the optimisation of the tokamak components and ancillaries to minimise Project risks. Open R&D issues related to these evaluations and mitigation measures are also described together with experimental, modelling and validation activities required to address them.The authors wish to thank a large number of colleagues (too many to be cited individually) in the fusion communities of the ITER Members for constructive discussions and criticisms in the elaboration of the new ITER baseline and of its Research Plan
BabloXR: An Authoring Tool for Developing WebXR Educational Applications
Get PDFDifferent areas of education have undergone a significant transformation due to the rapid progress of extended reality in recent years. Nonetheless, existing eXtended Reality (XR) authoring tools often require advanced programming skills, limiting their accessibility to educators and students. To address this challenge, we introduce BabloXR, an intuitive authoring tool that leverages Blockly to enable visual, code-free programming for creating educational XR environments. BabloXR was evaluated through a user study involving teachers and students who successfully designed their own XR learning environments. The results indicate that users of BabloXR were able to create complete and detailed scenarios, with usability testing revealing high ease of use, positive user experience, and strong interest in extended reality
Dynamic responses to smoking bans: Evidence from young adults in a developing country
Get PDFSmoking bans have been widely implemented, despite mixed evidence on their effectiveness in reducing smoking prevalence. This paper provides novel insights into the dynamic impacts of smoking bans in the context of Brazil, a large developing country with 18.4 million regular smokers in 2013. Our estimation strategy exploits the staggered implementation of comprehensive smoking bans in Brazilian state capitals using an event-study framework. We also leverage the variation in policy enforcement across cities. Our results indicate that bans reduced smoking prevalence by 18% among young adults, particularly when rigorously enforced. This effect is primarily driven by smoking cessation, while the impact on initiation is relatively modest. This policy change prevented roughly USD 87 million in healthcare costs in the treated cities.This project was funded by the Sao Paulo Research Foundation (FAPESP), Brazil, under grants #2016/18683-5 and #2017/11879-4. Paula Carvalho Pereda also thanks the Institute of Economic Research Foundation (FIPE) for the financial support, as well as the National Council for Scientific and Technological Development (CNPq) and FAPESP research grants #304221/2022-8 and #2014/50848-9 respectively. The opinions, hypotheses, and conclusions or recommendations expressed in this paper are the responsibility of the authors and do not necessarily reflect the views of FAPESP
AI-Driven Orchestration of 6G-Enabled Services Across the Computing Continuum
Get PDFProceedings of: 2025 8th International Balkan Conference on Communications and Networking (BalcanCom), Piraeus, 17-20 June 2025The emergence of 6G mobile communications promises transformative advancements powered by Artificial Intelligence (AI) and Machine Learning (ML). However, these capabilities introduce significant challenges, mainly due to the heterogeneous nature of 6G-enabled environments. This paper presents an AI-driven orchestration framework designed to address the complexities of future 6G infrastructures. The proposed framework orchestrates microservice-based applications across diverse layers of the computing continuum and performs key tasks such as service placement, computation offloading, horizontal scaling, and live migration, leveraging AI to enhance performance, energy efficiency, and automation. Additionally, it incorporates Distributed Ledger Technology (DLT) to enable federated service management when centralized control is impractical. We validate this approach through experimental trials with a latency-sensitive 6G application, analyzing intra- and interdomain orchestration, proactive extreme edge migration, and inter-cluster communication technologies. A set of representative workflows are provided as open-source implementations.
FilesThis work was supported in part by the European Commission through the Horizon Europe/JUSNS Projec tHexa-X-II under Grant 10109575
Rapid fatigue life estimation of drilled CFRP laminates and titanium stacks using thermography
Get PDFIn this work, we estimate the fatigue life of drilled Carbon Fiber Reinforced Polymer (CFRP) laminates making use of the temperature evolution of the component. As the self-heating effect due to cyclic loading causes a temperature change in the initial stages of the loading, the slope of the temperature variation ΔT versus the number of cycles can be used to estimate the remaining life. We have validated the applicability of the procedure for drilled laminates under different cutting conditions, and for drilled Ti-CFRP stacks. Infrared thermography was used to obtain the temperature variation during cycling loading. The results proved that the obtained adjusted model can effectively estimate the fatigue life of CFRP laminates with delamination damage or thermal damage with an error of 15%, proving the usefulness of this methodology for drilled and stacked laminates.This research is part of the R + D + i project PID2020-118480RB-C21,-C22, and PID2023-151610OB-C21,-C22, funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe”. Grant PRE2018-086277 funded by MCIN/AEI/10.13039/501100011033 and by “ESF Investing in your future”. The funding of the Generalitat Valenciana, Programme PROMETEO 2021/046 and project GV/2021/060 is also acknowledged
The critical role of the anode porous transport layer/catalyst layer interface of polymer electrolyte membrane water electrolyzers: A parametric analysis
Get PDFReducing the dependency of proton exchange membrane water electrolyzers (PEMWE) on precious metals, such as iridium (Ir), is necessary to develop a widespread green hydrogen system. This challenge requires a careful design of the interface between the anode porous transport layer (PTL) and the catalyst layer (CL). A comprehensive numerical analysis of relevant parameters that govern the behavior of the anode PTL/CL interface is presented. Calculations are also combined with an experimental characterization of the thickness and electrical conductivity of an unsupported CL as a function of Ir loading. The results show that the in-plane electrical resistance at the anode PTL/CL interface plays a critical role in cell performance. Reaching an acceptable electrical resistance at low Ir loading (Ir ≃ 0.1 mgIr , cm−2) can be accomplished through the incorporation of a micrometer-sized microporous layer (MPL) onto the PTL or the preparation of bimodal CLs with a secondary conductive phase. Further reduction of the Ir loading to the ultra-low regime (Ir ≲ 0.1 mgIr , cm−2) may require the use of nanometer-sized MPLs with unsupported CLs or micrometer-sized MPLs with bimodal CLs. Furthermore, the decline of the volume reactive area at ultra-low Ir loading needs a maximization of the exchange current density and the specific electrochemical surface area, and a decrease of the catalyst oxygen coverage factor in the anode CL.This work was supported by project TED2021-131620B-C21 of the Spanish Research Council. The authors acknowledge the use of facilities and instrumentation at the UC Irvine Materials Research Institute (IMRI), which is supported in part by the National Science Foundation through the UC Irvine Materials Research Science and Engineering Center (DMR-2011967). SEM work was performed using instrumentation funded in part by the National Science Foundation Center for Chemistry at the Space-Time Limit (CHE-0802913)
Los derechos digitales, una nueva generación de derechos humanos. Fundamentación y positivización en el plano internacional, europeo y español
Get PDFPrograma de Doctorado en Estudios Avanzados en Derechos Humanos por la Universidad Carlos III de MadridPresidenta: Ana María Manero Salvador.- Secretario: Migle Laukyte.- Vocal: Vanesa Morente Parr
Elliptical ejecta of asteroid Dimorphos is due to its surface curvature
Get PDFDocumento escrito por un elevado número de autores/as, solo se referencia el/la que aparece en primer lugar y los/as autores/as pertenecientes a la UC3M.Kinetic deflection is a planetary defense technique delivering spacecraft momentum to a small body to deviate its course from Earth. The deflection efficiency depends on the impactor and target. Among them, the contribution of global curvature was poorly understood. The ejecta plume created by NASA’s Double Asteroid Redirection Test impact on its target asteroid, Dimorphos, exhibited an elliptical shape almost aligned along its north-south direction. Here, we identify that this elliptical ejecta plume resulted from the target’s curvature, reducing the momentum transfer to 44 ± 10% along the orbit track compared to an equivalent impact on a flat target. We also find lower kinetic deflection of impacts on smaller near-Earth objects due to higher curvature. A solution to mitigate low deflection efficiency is to apply multiple low-energy impactors rather than a single high-energy impactor. Rapid reconnaissance to acquire a target’s properties before deflection enables determining the proper locations and timing of impacts.This work was supported by the DART mission, NASA Contract No. 80MSFC20D0004. M.H. acknowledges NASA/HERA-PSP (80NSSC24K1434). This work was supported by the Italian Space Agency (ASI) within the LICIACube project (ASI-INAF agreement n. 2019-31-HH.0 and its extension 2019-31-HH.1-2022). This work is partially supported by NASA through grant HSTGO-16674 from the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., under NASA contract NAS 5-26555. Portions of this work were performed by Lawrence Livermore National Laboratory under DOE contract DE-AC52-07NA27344. LLNL-JRNL-853920. S.D.R. and M.J. acknowledge support from the Swiss National Science Foundation (project number 200021 207359). Work of E.G.F., S.P.N., and S.R.C. was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration (80NM0018D0004). R.M. acknowledges funding from a NASA Space Technology Graduate Research Opportunities (NSTGRO) award, NASA contract No. 80NSSC22K1173. H.F.A. was supported by the French government through the UCA J.E.D.I. Investments in the Future project managed by the National Research Agency (ANR) with the reference number ANR-15-IDEX-01. P.M. acknowledges funding support from the French Space Agency CNES and The University of Tokyo. G.T. acknowledges financial support from project FCE-1-2019-1-156451 of the Agencia Nacional de Investigación e Innovación ANII and Grupos I + D 2022 CSIC-Udelar (Uruguay). The work by J.O., S.D.R., and G.S.C. was supported by Grant PID2021-125883NB-C22 by the Spanish Ministry of Science and Innovation/State Agency of Research MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe.” J.O., I.H., S.D.R., and M.J. were supported by the Spanish Research Council (CSIC) support for international cooperation: I-LINK project ILINK22061. S.R.S. acknowledges support from the DART Participating Scientist Program, grant no. 80NSSC22K0318. R.L., D.G., and T.S. acknowledge support by the NASA/GSFC Internal Scientist Funding Model (ISFM) Exospheres, Ionospheres, Magnetospheres Modeling (EIMM) team, and the NASA Solar System Exploration Research Virtual Institute (SSERVI). R.L. and D.G. acknowledge work done through the Center for Research and Exploration in Space Science and Technology (CRESST-II) supported by NASA award number 80GSFC24M0006
Minimum-Cost Design of Auto-Scaling Server Farms Providing Reliability Guarantees
No full textAs next-generation mobile networks increasingly rely on virtualized infrastructure to deliver critical services, ensuring both the efficiency and reliability of server farms becomes essential. These infrastructures must meet stringent reliability guarantees to support time-sensitive applications in emerging 5G and beyond networks. In this paper, we address the design of auto-scaling server farms–specifically, selecting the most suitable server type and corresponding number of servers–by considering both service requirements and associated operational and infrastructure costs. To this end, we develop an optimization algorithm that combines (i) a queueing-theoretic model to estimate the resources needed to meet reliability constraints, and (ii) a general cost model that captures both capital and operational expenditures. We validate our approach through extensive simulations, comparing it against classical queueing-based methods and exhaustive numerical searches: our proposal reduces costs by 22% as compared against the benchmark, with solutions that are within 3% of numerical searches at 10% of the computational complexity, offering a new scalable and cost-effective methodology for designing reliable server farms.This work was supported in part by MCIN/AEI/10.13039/501100011033/ through the Project 6GINSPIRE under Grant PID2022-137329OB-C42; in part by the NEC Laboratories Europe Student Research Fellowship Program of 2021; and in part by the Spanish Ministry of Economic Affairs and Digital Transformation and the European Union-NextGenerationEU through the UNICO 5G I+D SORUS Project. The work of Jaime Garcia-Reinoso was supported in part by the Spanish Ministry for Science and Innovation through the ADMINISTER Project under Grant TED2021-131301B-I00