Universitat Politècnica de Catalunya
UPCommons. Portal del coneixement obert de la UPCNot a member yet
290727 research outputs found
Sort by
Mary Santpere, a comic actress undoing gender during Francoist dictatorship
No full textSpanish comic Mary Santpere played in secondary roles at the beginning of Francoism in the 1940s but drifted towards protagonism in the last decade before the dictator’s death in 1975. The dictatorship’s censorship required feminine characters to be beautiful and passive, which penalised the representation of women with agency. Given that Mary Santpere started her film career in ‘counterpart’ roles before being the star of four films, one of the goals of this article is to assess whether her stardom came at a price. The hypothesis is that more stringent rules of behaviour were applied to actresses playing leading roles given that they were expected to conform to the regime’s model of womanly conduct. The text analyzes her films in order to detect recurring elements of her star persona, such as her imposing height, her exceptional blondeness in a country where brunettes abounded, and her absence of conventional beauty. It explores her secondary comic characters in the 1940s and 1950s and the four main roles of the 1960s that Mary Santpere played to contextualise her star-persona during the Franco dictatorship, and also looks at whether the personal gender performativity Mary Santpere exploited during the Francoist period can be read against the grain.5 - Igualtat de Gènere4 - Educació de QualitatPostprint (author's final draft
A platform for comparative studies of anomaly detection algorithms
No full textLa detecció d'anomalies és una tasca clau en nombrosos dominis de la ciència de dades, com ara la detecció de frau, la monitorització de sistemes o l'anàlisi de sensors. Tanmateix, la gran varietat d'algorismes disponibles, la seva elevada sensibilitat als hiperparàmetres i la manca de criteris homogenis d'avaluació dificulten la selecció de la tècnica més adequada per a un problema concret. Aquest projecte presenta el disseny i desenvolupament d'una plataforma interactiva per a l'estudi comparatiu d'algorismes de detecció d'anomalies, orientada a facilitar una avaluació objectiva, reproduïble i comprensible de diferents mètodes. La solució permet carregar conjunts de dades tabulars propis, configurar pipelines de preprocessament i executar, en una única experiència integrada, diversos algorismes representatius sota un esquema de validació controlat amb mètriques rigoroses. El treball aporta una solució pràctica que dona suport a l'investigador en l'anàlisi comparativa de mètodes de detecció d'anomalies, promovent bones pràctiques experimentals i una comprensió més crítica dels resultats obtinguts.Anomaly detection is a key task in many data science domains, such as fraud detection, system monitoring, and sensor data analysis. However, the wide variety of available algorithms, their high sensitivity to hyperparameters, and the lack of homogeneous evaluation criteria make it difficult to select the most appropriate technique for a given problem. This project presents the design and development of an interactive platform for the comparative study of anomaly detection algorithms, aimed at facilitating an objective, reproducible, and understandable evaluation of different methods. The solution allows users to load their own tabular datasets, configure preprocessing pipelines, and run several representative algorithms within a single integrated experience under a controlled validation scheme with rigorous metrics. This work provides a practical solution that supports researchers in the comparative analysis of anomaly detection methods, promoting sound experimental practices and a more critical understanding of the obtained results.La detección de anomalías es una tarea clave en numerosos dominios de la ciencia de datos, como la detección de fraude, la monitorización de sistemas o el análisis de sensores. Sin embargo, la gran variedad de algoritmos disponibles, su elevada sensibilidad a los hiperparámetros y la falta de criterios homogéneos de evaluación dificultan la selección de la técnica más adecuada para un problema concreto. Este proyecto presenta el diseño y desarrollo de una plataforma interactiva para el estudio comparativo de algoritmos de detección de anomalías, orientada a facilitar una evaluación objetiva, reproducible y comprensible de distintos métodos. La solución permite cargar conjuntos de datos tabulares propios, configurar pipelines de preprocesado y ejecutar, en una única experiencia integrada, varios algoritmos representativos bajo un esquema de validación controlado con metricas rigurosas. El trabajo aporta una solución práctica que apoya al investigador en el análisis comparativo de métodos de detección de anomalías, promoviendo buenas prácticas experimentales y una comprensión más crítica de los resultados obtenidos
Diseño e implantación de un sistema de gestión del mantenimiento en una planta de tratamiento de residuos sólidos urbanos
No full textEl proyecto tiene como objetivo implementar un nuevo Modelo de Gestión de Mantenimiento Industrial en la Planta de Tratamiento de Residuos de Orís, con el fin de mejorar la planificación y estandarización de las tareas de mantenimiento correctivo y preventivo, y de aumentar su eficiencia operativa. Además, se pretende demostrar que dicho modelo puede aplicarse de forma homogénea en todas las plantas de tratamiento de residuos de la empresa, independientemente del tipo de residuo gestionado. A lo largo del trabajo se analizará el impacto de la implantación del nuevo modelo, evaluando aspectos como el incremento de la disponibilidad de la planta, la mejora de la fiabilidad de los equipos y la reducción de incidencias repetitivas. Además, se llevará a cabo un estudio de los costes asociados al mantenimiento y una valoración del impacto ambiental, con el fin de minimizarlo mediante prácticas más eficientes y una gestión más estructurada de los recursos. La finalidad del proyecto es mejorar la funcionalidad, la seguridad y la sostenibilidad de la planta, garantizando un sistema de mantenimiento más robusto, ordenado y alineado con las necesidades operativas de la organización
Environmental and behavioural factors influencing myopia onset: literature review to develop an evidence-based infographic
No full textPurpose: The aim of this literature review is to examine current evidence on environmental and behavioural factors influencing myopia onset and development in children and teenagers. Based on the findings, an informative infographic representing key risk factors and potential preventive interventions for myopia onset was developed.
Methods: A systematic literature search was conducted in PubMed and on the online literature directory of the International Myopia Institute, including articles published within the last 5 years. Studies focusing on myopia from early childhood to late teenage years were evaluated. In total, 7 articles were reviewed in detail. Various study designs were considered and the selection of studies was not restricted to a specific geographical region.
Results: The reviewed literature indicates that time spent outdoors and near work are the primary environmental and behavioural factors associated with myopia onset and development. Increased outdoor exposure is consistently associated with a reduced risk of developing myopia and appears to have a protective effect because of daylight exposure. In contrast, near work is inconsistently associated with a higher prevalence of myopia. Additional factors, including education, digital screen time, socio-economic status and urban/rural differences, were also identified as influencing factors.
Conclusion: Environmental and behavioural factors play a significant role in the development of myopia in children and teenagers. These findings underline the importance of preventive interventions that can be integrated into everyday life to reduce the risk of myopia onset. Further research is required to better understand the underlying mechanisms and the role of confounding factors
Com contribuïm a la contaminació de l’aire? Una activitat col·laborativa per a l’ensenyament de l’estadística a secundària
No full textPresentem una proposta col·laborativa basada en la indagació per a l’ensenyament de l’estadística, la qual va ser implementada a l’escola Nelson Mandela de Berlín durant el curs 2023-24. Va ser realitzada en els cursos 7, 9 i 11, equivalents a 1r. i 3r de l’Educació Secundària Obligatòria (ESO), i un curs pont entre l’ESO i el Batxillerat. L’activitat es va dissenyar de tal manera que cada curs treballava en el que marcava el currículum sobre l’ensenyament de l’estadística en el seu nivell. Addicionalment, els alumnes grans actuaven de guies dels petits i validaven les tasques presentades. El resultat final de l’activitat incorporava les tasques realitzades per cada curs
Safety aspects of torrefied biomass: prosperous and dangerous process for green transition
No full textIn this paper we present an analysis on safety aspects of torrefied biomass. Torrefaction is increasingly deployed to upgrade biomass for co-firing and as a precursor to advanced biofuels, yet the same pretreatment can increase dust generation and modify self-heating behaviour, elevating fire and explosion risks. This paper synthesises the comparable 20-L sphere explosibility datasets for raw and torrefied wheat straw, rape straw and vine shoots, a harmonised interpretation of reported Kst, Pmax and (dp/dt)max under standardised test conditions, and a keyword co-occurrence network (1984–2025) interpreted through Perrow's Normal Accident Theory to connect material hazards with system-level drivers. Across the three residues, torrefaction increased (dp/dt)max by ~20–25% while remaining in the St1 class, implying ~17–26% larger vent areas when translated to a representative reduced explosion pressure target. The bibliometric layer indicates a persistent separation between renewable-energy narratives and discourses on explosion prevention or explosion protection and functional safety, with only limited bridging emerging in recent years. We therefore position torrefaction mid-way on a complexity–coupling continuum between conventional biomass handling and the more tightly coupled high temperature pyrolysis and high pressure hydroprocessing used in sustainable aviation fuel (SAF) production. We argue that treating accidents as “normal” in such coupled systems could motivate design- and governance-level measures (decoupling, monitoring, barrier management) that complement standard prevention and certification practices.Postprint (published version
RV.S3.C1.V4 Vigas
No full textVídeos docents de l'assignatura Representació Arquitectònica II (REVIT) impartida per Isidro Navarro, PDI de l'ETSA
Near-future prediction of pipe failures in water supply networks: a key determinant for water pipe renewal policies through a machine learning approach
No full textThis study explores the development and deployment of a predictive model for pipe failures within Barcelona's Water Distribution System. Using the XGBoost algorithm, refined through systematic analysis of explanatory variables, machine learning algorithms, and hyperparameter configurations, the model predicted up to 30.2% of expected pipe failures with a 5% annual renewal rate and 10.32% with a 1% renewal rate. Material-specific responses to predictive variations were observed, with ductile iron and HDPE pipes showing significantly different behaviours compared to non-cylinder reinforced concrete and fibre-cement pipes. The model development process was heavily guided by domain expertise, as reflected in the custom-built dataset, which was meticulously created considering local system conditions, strategic hyperparameter tuning, and optimisation using business-focused metrics. A preliminary exploration employing SHAP (Shapley Additive exPlanations) assessed the importance of local explanatory variables, which varied across pipe materials. This research advances understanding of how specific materials within water systems respond to predictive modelling of pipe failures, emphasising the vital role of extensive historical data in enhancing predictive accuracy and informing infrastructure planning and management.The authors gratefully acknowledge Aigües de Barcelona (ABEMGCIA) for providing the data and work environment necessary for model development and validation. The authors also thank the Àrea Metropolitana de Barcelona (AMB), the Universitat Politècnica de Catalunya (UPC), the Barcelona Supercomputing Centre (BSC-CNS), and the Water Research Centre (CETAQUA) for their support.
This work was supported by the "Agència de Gestió d'Ajuts Universitaris i de Recerca" (AGAUR) (https://agaur.gencat.cat/en/) under the Industrial Doctorate Plan of the Secretariat for Universities and Research of the Department of Business and Knowledge of the Government of Catalonia (reference: DI 093–2021). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.Peer ReviewedPostprint (published version
Assessment of pilot fatigue based on flight routes: Towards the determination of operational limits
No full textThis Bachelor's Thesis investigates pilot fatigue in long-haul operations with the specific objective of assessing the feasibility of operating the Miami-Barcelona (MIA-BCN) return flight with two pilots instead of the current three-pilot configuration. The study uses the BCN-MIA and MIA-BCN routes as a case study to analyse how crew composition, duty timing, rest opportunities and circadian effects influence fatigue and alertness levels. The analysis combines predictive alertness modelling, roster-level fatigue assessment and subjective pilot feedback. A biomathematical fatigue model implemented through the Crew Alert tool is used to evaluate predicted alertness at Top of Descent under different crew configurations and departure times. In parallel, a survey-based study collects subjective data on fatigue, sleepiness and alertness before, during and after flight. In addition, a comparative analysis of fatigue accumulation is performed at both flight and roster level to assess the impact of operating the MIA-BCN return flight with two pilots instead of three. The results show that reducing crew composition on the return flight leads to a decrease in fatigue and alertness margins, increasing exposure to fatigue-related hazards, particularly in the presence of circadian misalignment and long duty periods. While subjective and predictive indicators generally remain within acceptable operational limits under the current three-pilot configuration, the analyses indicate that a two-pilot operation would significantly reduce operational margins and increase sensitivity to schedule variations and delays. The study concludes that, although operating the MIA-BCN route with two pilots may be feasible under certain conditions, a definitive assessment would require the implementation of a fully validated Fatigue Risk Management System supported by biomathematical fatigue modelling. The findings highlight the importance of data-driven fatigue management in long-haul operations and provide operational insight to support informed decision-making regarding crew scheduling and fatigue mitigation strategies
Microstructural effects on the fatigue crack growth behavior of y-phase containing WC-Co cemented carbides: Mechanics, mechanisms and fatigue sensitivity
No full textThe partial substitution of tungsten carbide by cubic refractory ones (¿-phase) represents an accessibility-driven approach for the microstructural design of hard ceramic-metal composites, offering an alternative to WC-Co hardmetals by reducing dependence on tungsten as a critical raw material. However, successful implementation of this approach requires knowledge and deeper understanding of microstructural effects on mechanical integrity, beyond simple hardness – indentation fracture toughness correlations, for these ¿-phase containing- WC-Co cemented carbides. In this study, a systematic and detailed investigation addressing the influence of ¿-phase carbides – as third phase – on crack growth resistance of WC-Co hardmetals, under monotonic and cyclic loading, is conducted. Materials studied include two ¿-phase containing grades with submicron and fine grain sizes, as well as two reference WC-Co systems with matching microstructural features. Fatigue crack growth behavior and fracture toughness are assessed by testing through-thickness pre-cracked specimens. The mechanical study is complemented by an extensive characterization of cracking paths and fractographic features. Independent of microstructural assemblage, crack propagation under variable loading is found to be dominated by static failure modes rather than pure cyclic ones. Meanwhile, quantitative analysis of crack-microstructure interactions reveals an increased frequency of transgranular cracking through the ¿-phase carbides in both submicron- and fine-grained grades. This is more pronounced in the former, significantly reducing the relative prominence of binder-related crack paths. Hence, despite exhibiting higher crack growth rates, the submicron-grained three-phase cemented carbide is found to have a lower fatigue sensitivity relative to its reference counterpart. This behavior reflects a microstructure-dependent trade-off and should be interpreted within a tailored application framework. However, such behavior is not kept as microstructure gets coarser, because this yields higher and lower proportions of ductile binder fracture and transgranular crack paths within ¿-phase carbides, respectively. Nanoindentation measurements revealed significant differences in hardness and modulus between WC, ¿-phase and binder regions, further validating the observed failure micromechanisms. The experimental findings and their corresponding analysis underscore the critical influence of microstructural assemblage — particularly the contiguity and distribution of the ¿-phase carbides — in controlling fracture and fatigue behavior in multielement cemented carbide systems.The research work was conducted within an Industry-Academy collaboration between Hyperion Materials & Technologies and Universitat Politècnica de Catalunya. The study was partly supported by the Spanish Government (Ministerio de Ciencia, Innovación y Universidades and Agencia Estatal de Investigación) and European Union (FEDER) through the projects PID2021-126614OB-I00 and PID2022-137274NB-C32 and it is part of the María de Maeztu grant CEX2023-001300-M, all of them funded by MICIU/AEI/10.13039/501100011033Peer ReviewedPostprint (published version