DOCTORAL DEGREE IN ARCHITECTURAL, BUILDING CONSTRUCTION AND URBANISM TECHNOLOGY

• GONZÁLEZ ESPINOSA, VANESSA: Diseño de materiales cementícios reforzados con fibras vegetales impregnadas con materiales de cambio de fase para mejorar el comportamiento térmico de las cubiertas de los edificios.
  
  Author: GONZÁLEZ ESPINOSA, VANESSA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ARCHITECTURAL, BUILDING CONSTRUCTION AND URBANISM TECHNOLOGY
  Department: Department of Architectural Technology (TA)
  Mode: Normal
  Deposit date: 31/07/2025
  Reading date: 28/11/2025
  Reading time: 11:00
  Reading place: EPSEB (Escuela Politécnica Superior de Edificación de Bcn) - Sala de Actos Campus Diagonal Sur, Edificio P. - Av. Doctor Marañón, 44-50 - 08028 BCN
  Thesis director: CLARAMUNT BLANES, JOSE | LACASTA PALACIO, ANA MARIA
  Thesis abstract: In the current context, sustainable construction prioritises innovative materials that combine energy efficiency, mechanical strength and safety against extreme conditions, such as fire, in order to address the challenges of climate change and human needs. Fibre-reinforced cementitious composites and phase change materials (PCM) are emerging as a promising solution, particularly in raised roof pavements, where thermal regulation is essential. The main objective of this thesis is to develop a cement board reinforced with non-woven vegetable fibres and PCM for raised roof pavements, determining the optimal dosage of cement, fibres and PCM that maximises mechanical resistance and thermal regulation capacity, as well as analysing its response to fire. The research seeks to advance the integration of PCM and plant fibres into cementitious matrices, proposing improvements for their practical application in construction with an environmental focus.The methodology, which is highly experimental in nature, was structured in several stages. First, the state of the art on cementitious composites and PCM was reviewed, identifying trends and challenges. Subsequently, an experimental campaign was designed that included: selection of materials (commercial cements, non-woven fibres and pure and microencapsulated PCMs), formulation of mixtures with different proportions of fibres and PCMs compared to a control without PCMs, evaluation of mechanical properties (flexural strength), thermal properties (conductivity, thermal storage and retardation) and fire behaviour through standardised tests, and statistical analysis to determine the impact of each component.The results show that the incorporation of PCM RT28 through direct impregnation into non-woven fibres in cementitious composites improves thermal properties, achieving a delay in temperature changes comparable to that of microencapsulated PCM mixed into the cementitious matrix. However, the composite made with PCM RT28 has superior mechanical strength, with a modulus of rupture (MOR) approximately three times greater than that of the microencapsulated composite, although both show a decrease in strength compared to samples without PCM. The non-woven fibres, by effectively impregnating the PCM, reinforce the cohesion of the composite and preserve hardening by deformation, partially mitigating the loss of mechanical strength.Although both the vegetal fibres and the PCM used are organic in nature and therefore combustible, the cementitious composites exhibited good fire performance, with low-intensity flames and a high self-extinguishing capacity once the heat source was removed. Overall, the samples with PCM incorporated through fibre impregnation showed better fire behaviour than those formulated with microencapsulated PCM: although ignition occurred slightly earlier, the total heat released (THR), as measured in the cone calorimeter tests, was substantially lower.This combination, which has been little explored, balances thermal efficiency and structural functionality, with direct applications in sustainable buildings. The research provides a detailed analysis of the interaction between plant fibres, PCM and the cement matrix, proposing optimal dosages and strategies to mitigate fire-related risks. The results lay the foundations for future research and practical applications, promoting the development of more efficient and sustainable building materials.
  
• REZK, JOSEPHIN: Sustainability assessment model for post-war reconstruction of blast-damaged RC buildings without structural collapse risk: A multi-criteria decision-making framework applied to the tourism sector in Damascus, Syria
  
  Author: REZK, JOSEPHIN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ARCHITECTURAL, BUILDING CONSTRUCTION AND URBANISM TECHNOLOGY
  Department: Department of Architectural Technology (TA)
  Mode: Normal
  Deposit date: 26/09/2025
  Reading date: 11/12/2025
  Reading time: 11:00
  Reading place: ETSAB (Escuela Técnica Superior de Arquitectura de Barcelona) - Planta Baja - Sala de Grados Av. Diagonal, 649-651 - 08028 - Barcelona
  Thesis director: PONS VALLADARES, ORIOL | MUÑOZ BLANC, CARLOS
  Thesis abstract: Post-war reconstruction presents a series of complex challenges that go beyond restoring physical structures. In conflict-affected regions, decisions regarding the reconstruction of damaged buildings must be made under conditions of extreme resource limitations, infrastructural disruption, and socio-cultural fragility. However, most existing sustainability assessment frameworks are not suited to these conditions. They typically lack the adaptability and practicality required to support decision-making for individual buildings in post-conflict settings. This doctoral thesis addresses this gap through the development of a new decision-support model for assessing the sustainability of reconstruction strategies for reinforced concrete buildings damaged by external blasts but confirmed to be structurally stable.The thesis proposes a novel methodology that integrates the Integrated Value Model for Sustainable Assessment (MIVES) with the Delphi method, forming a multi-criteria decision-making framework capable of evaluating four reconstruction alternatives: refurbishment, demolition, reconstruction with retained identity, and preservation for future work. A key contribution of the thesis is the development of a specialized technical assessment form designed specifically for use in resource-constrained conflict zones. This form facilitates the classification of damage using visual inspection and expert consultation, offering a practical and accessible alternative to laboratory-based evaluations. It serves as an essential tool for determining a building’s eligibility for further sustainability-based assessment using the proposed model.To validate the MIVES-Delphi model, the thesis applies it to three real-world case studies involving reinforced concrete buildings used in the tourism sector in Damascus, Syria. Two of the buildings are located within protected historical areas, while the third is situated in a major tourism corridor. All buildings were confirmed by engineering reports to pose no risk of structural collapse. The results demonstrate that no single reconstruction alternative is universally optimal. In one case, refurbishment proved to be the most sustainable option in economic and social terms, while in another, preservation emerged as the most favorable from an environmental perspective. These outcomes highlight the model’s ability to adapt to contextual variables and support balanced, evidence-based reconstruction decisions.This thesis concludes that the newly developed decision-support model and technical assessment form together provide a transparent, replicable, and context-sensitive framework for guiding sustainable reconstruction in post-war environments. Although the model was validated through case studies in Syria, it is designed to be applicable across a wide range of conflict-affected regions. The work contributes significantly to the advancement of sustainable post-conflict recovery by offering practical tools for stakeholders seeking to align immediate reconstruction efforts with long-term resilience and development goals.
  

DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION

• DELGADO GUERRERO, JUAN ANTONIO: Learning latent structures for robotic assistance in daily manipulation tasks
  
  Author: DELGADO GUERRERO, JUAN ANTONIO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION
  Department: Institute of Robotics and Industrial Informatics (IRI)
  Mode: Normal
  Deposit date: 02/09/2025
  Reading date: 27/11/2025
  Reading time: 11:00
  Reading place: Sala d'Actes Manuel Marta Recober, planta 0, Edifici B6, Facultat d'Informática de Barcelona (FIB), Campus Nord, C/Jordi Girona, 1-3, 08034 Barcelona
  Thesis director: TORRAS GENIS, CARMEN | COLOMÉ FIGUERAS, ADRIÀ
  Thesis abstract: Robotic domestic assistance presents significant challenges due to the complexity of modeling everyday manipulation tasks, especially those involving deformable objects like cloth. Traditional approaches often struggle with high-dimensional state representations, dynamic uncertainties, and the need for safe human-robot interaction. This thesis addresses these challenges by developing novel machine learning methods based on latent variable models to enable efficient, adaptive, and safe robotic manipulation.First, we propose a Gaussian Process Latent Variable Model (GPLVM) framework combined with Bayesian Optimization (BO) to learn high-dimensional robot motion policies with minimal data. This approach reduces the parameter space dimensionality while preserving task-relevant features, achieving faster convergence than other existing model-free alternatives.Next, we extend this framework to contextual learning using Covariate GPLVM (c-GPLVM), allowing robots to adapt to environmental changes (e.g., user preferences, object positions) without retraining. Experiments in feeding and shoe-fitting tasks demonstrate improved generalization with fewer samples compared to state-of-the-art contextual policy search methods.For dynamic cloth manipulation, we introduce the Controlled Gaussian Process Dynamical Model (CGPDM), which embeds control actions into a low-dimensional latent space to predict cloth motion under robot manipulation. Evaluations in simulated and real-world bimanual cloth handling show that CGPDM accurately generalizes to unseen actions, even with limited training data.Finally, we address safety in human-robot interaction by proposing Cartesian control enhancements for redundant manipulators, including error saturation, singularity avoidance, and impedance tuning. These measures mitigate risks during physical interaction, ensuring stable and compliant robot behavior.Together, these contributions advance robotic cloth manipulation by combining data-efficient learning, context-aware adaptation, and safe control, paving the way for practical deployment in assistive and household robotics.
  
• GARCIA CAMACHO, IRENE: Benchmarking cloth manipulation
  
  Author: GARCIA CAMACHO, IRENE
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION
  Department: Institute of Robotics and Industrial Informatics (IRI)
  Mode: Normal
  Deposit date: 25/09/2025
  Reading date: 24/11/2025
  Reading time: 11:00
  Reading place: Aula Capella, Escola Tècnica Superior d'Enginyeria Industrial de Barcelona (ETSEIB), Edifici PI (Pavelló I). Av. Diagonal, 647, Barcelona
  Thesis director: ALENYÀ RIBAS, GUILLEM | BORRÀS SOL, JÚLIA
  Thesis abstract: Benchmarking is a crucial tool in research for fostering progress in a field. It consists on standardized frameworks for evaluation, quantifying the performance of an approach in comparison to previous works to determine the improvements and progress made. Robotic cloth manipulation presents many challenges due to the high deformable nature of textile objects. It is an interdisciplinary field that integrates many components such as control, perception and hardware, among others, to solve cloth manipulation problems with a wide variety of robotic platforms, end-effectors, objects and strategies. This variability makes difficult to design general evaluation procedures that can be adopted by the vast researchers in the field. This thesis addresses the need for standardized benchmarks in cloth manipulation, providing solutions for the key aspects of which a benchmark is composed: setup description, task description and evaluation procedures. The thesis starts with the design of benchmarks for relevant cloth manipulation tasks, proposing clear procedures and metrics to assess the performance quality. Setup standardization is improved with the introduction and distribution of a standardized cloth object set with textile household objects, given that object directly impact on the manipulations required and results obtained. Additionally, we propose a framework to characterize textile objects through its physical and mechanical properties, so we can deal with object's wear and usage over time, maintaining the standardization and extending it to textile objects of other categories. Subsequently, we proposed a scene state definition of cloth manipulation based on its configuration, grasp type and grasp location to represent cloth manipulation tasks and build more informed evaluation metrics. Later delving into state estimation for decision-making and benchmarking. The thesis culminates with the organization of a cloth manipulation and perception competition, done for joining research groups in comparing their systems with equal test conditions and raise awareness of the importance of designing and adopting standardized evaluation processes to foster in the field.In summary, this thesis touches on benchmarking, standardization, task representation, and decision-making in the context of cloth manipulation.
  
• PUERTO SANTANA, CRISTIAN: Design, implementation, and evaluation of novel fault detection methodologies for time varying industrial mechanical systems
  
  Author: PUERTO SANTANA, CRISTIAN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION
  Department: Department of Automatic Control (ESAII)
  Mode: Normal
  Deposit date: 25/09/2025
  Reading date: 28/11/2025
  Reading time: 11:00
  Reading place: Aula 28.8, Escola Tècnica Superior d'Enginyeria Industrial de Barcelona (ETSEIB), Avinguda Diagonal, 647, 08028 Barcelona
  Thesis director: OCAMPO MARTINEZ, CARLOS AUGUSTO | DÍAZ ROZO, JAVIER
  Thesis abstract: This dissertation addresses the necesity of scalable and robust fault detection methods in industrial mechanical systems operating under complex, noisy, and variable conditions. These systems—comprising components such as rotors, gears, and structural frameworks—play a vital role in industrial operations, and their failures can lead to substantial performance degradation, safety risks, and increased maintenance costs. Motivated by these challenges, the research proposes data-driven methodologies designed to improve the accuracy, reliability, and scalability of condition monitoring in these environments. Unlike traditional model-based methods that rely heavily on system-specific knowledge and simulation, the approaches developed in this work emphasize adaptability, minimal calibration, and real-world applicability. The thesis begins with a detailed examination of industrial mechanical systems and a critical review of existing fault detection methods, particularly those targeting mechanical unbalance, gear defects, and structural resonance. This foundation highlights the limitations of current technologies and sets the stage for the development of six original contributions.A central innovation in this research is a method for automatically identifying transient and stationary regimes in multivariate systems using statistical tools, signal processing, and geomery of spatial curves. This regime classification is essential for segmenting data accurately and applying appropriate fatult detection strategies. Building on this framework, the thesis introduces two novel condition indicators: one for detecting mechanical unbalance using non-intrusive signal processing techniques, and another for identifying gear faults under fluctuating speed conditions via Gaussian mixture models and the Cauchy-Schwarz divergence. Both indicators are designed to function reliably in noisy environments and are validated on experimental platforms that replicate real industrial conditions. Another major contribution extends monitoring capabilities to both rotordynamic elements and structural components within a single system, demonstrated through a power generation setup. This cross-monitoring approach integrates data-driven modeling to detect anomalies that may result from mechanical faults or structural instabilities, offering a more comprehensive diagnostic tool.In addition to enhancing fault detection in rotating machinery, the dissertation explores advanced techniques in structural health monitoring. Using long-term bridge data, the research applies the Hankel alternative view of Koopman analysis to extract meaningful features and assess parameter sensitivity. This is complemented by a new feature extraction strategy that leverages regime classification to enhance resilience to noise and reduce computational costs. The final contribution is a probabilistic modeling framework based on Copula functions, which allows for flexible, noise-tolerant modeling of feature distributions while incorporating prior knowledge. Collectively, these methodologies form a unified framework capable of operating under real-world industrial constraints. The research not only demonstrates high performance across varied fault types and systems but also advances the field by offering tools that can adapt to multivariate, dynamic, and uncertain conditions without requiring extensive system modeling or manual settings.
  

DOCTORAL DEGREE IN BIOMEDICAL ENGINEERING

• JAMES, CHRISTOPHER WILLIAM VINCENT: Biomaterials for Cardiac Regeneration
  
  Author: JAMES, CHRISTOPHER WILLIAM VINCENT
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN BIOMEDICAL ENGINEERING
  Department: Department of Materials Science and Engineering (CEM)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 26/11/2025
  Reading time: 16:00
  Reading place: Sala d'Actes, Edifici Vèrtex, Campus Diagonal Nord, Vèrtex (VX), Plaça d'Eusebi Güell, 6, 08034 Barcelona
  Thesis director: ENGEL LOPEZ, ELISABET
  Thesis abstract: Cardiovascular diseases are the leading cause of death globally, with heart transplantation being the most effective treatment following injury due to the heart’s limited regenerative capacity. In situ tissue engineering has emerged as a promising approach to activate endogenous cardiac repair. This thesis focuses on the design, development, and characterization of injectable biomaterials for in situ cardiac regeneration.Hydrogel scaffolds that physically support damaged tissue, provide inherent bioactivity, or deliver bioactive agents are particularly promising. We first isolated and characterized porcine cardiac extracellular matrix (cECM) hydrogels, demonstrating their compatibility for supporting the growth of cardiac-associated cells. To enhance the viscoelastic properties of cECM without compromising biocompatibility, alginate—a hydrogel currently in clinical trials—was incorporated, resulting in improved mechanical properties.Lactate, traditionally seen as a metabolic by-product, has gained attention for its role in promoting angiogenesis, cardiomyocyte proliferation, and reducing fibrosis. Based on this evidence, we developed a lactate-release scaffold by embedding FDA-approved PLGA nanoparticles into the cECM-alginate matrix, optimizing nanoparticle size and degradation to achieve sustained lactate release.The regenerative potential of stem cell-derived secretomes, comprising bioactive molecules such as growth factors and extracellular vesicles, was also explored. Bone marrow-derived stromal cell (BMSC) secretomes were evaluated for their cardioprotective effects on human cardiac fibroblasts. A novel culture method showed superior outcomes, and the resulting secretome was incorporated into the scaffold either directly or via PLGA nanoparticles for sustained delivery.In conclusion, this work presents several novel injectable biomaterials that show potential for in situ cardiac regeneration through enhanced mechanical support, bioactivity, and sustained delivery of regenerative agents such as lactate and BMSC-derived secretomes. These findings warrant further investigation to optimize the therapeutic efficacy of these platforms.
  
• LÓPEZ GÓMEZ, PATRICIA VICTORIA: Multifunctional hydrogels for advanced regenerative therapies
  
  Author: LÓPEZ GÓMEZ, PATRICIA VICTORIA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN BIOMEDICAL ENGINEERING
  Department: Department of Materials Science and Engineering (CEM)
  Mode: Normal
  Deposit date: 19/09/2025
  Reading date: 11/12/2025
  Reading time: 10:30
  Reading place: Aula A1.13, Edifici A, Escola d'Enginyeria de Barcelona Est (EEBE), Av. d'Eduard Maristany, 16, 08019 Barcelona
  Thesis director: MAS MORUNO, CARLOS | MEHWISH, NABILA
  Thesis abstract: Implant-associated infections remain one of the most critical challenges in the biomedical field. Despite advances in aseptic surgical techniques and antibiotic therapies, the persistence of bacterial colonization on implant surfaces -often involving biofilm formation- continues to compromise clinical outcomes. Conventional treatment strategies, including the systemic administration of antibiotics, local drug delivery systems, and surgical debridement, often fail to effectively eradicate biofilms, particularly those formed by pathogens such as Staphylococcus aureus and Pseudomonas aeruginosa. Concurrently, the field of tissue engineering demands implantable materials that not only support tissue regeneration but also provide active defense against infection. However, most currently available materials fall short of achieving this dual function. In contaminated or high-risk environments, this shortcoming becomes especially critical. To address these limitations, there has been a growing focus on developing next-generation biomaterials that are no longer passive scaffolds but bioactive and dynamic systems capable of interacting with the biological milieu in real time. In this context, biofunctionalization has emerged as a powerful strategy to enhance both regenerative and antimicrobial properties of biomaterials. Among the different bioactive tools available, peptides have shown considerable promise due to their tunable chemistry, modular architecture, and high specificity. This thesis focuses on two peptide motifs with complementary bioactivities: RGD, which promotes cell adhesion, and hLf1-11, a broad-spectrum antimicrobial peptide derived from human lactoferrin. Together, these peptides represent a rational platform for engineering multifunctional materials that address the dual challenge of infection control and tissue integration. Specifically, the present work investigates the integration of RGD and hLf1-11 peptides into three hydrogel-based material systems, each representing a distinct level of biofunctionality and design complexity: • Chapter I examines the modification of alginate, a naturally derived but bioinert polymer, with RGD-hLf1-11 to evaluate stem cell behavior and antimicrobial activity. • Chapter II explores a fully synthetic PEG-based hydrogel system functionalised with these peptides and incorporating a protease-sensitive crosslinker to enable bioactivity and controlled degradation. • Chapter III focuses on self-assembling peptide hydrogels, where both structural assembly and biological function are encoded at the molecular level, enabling the formation of intrinsically bioactive materials without further modification. Collectively, these platforms offer a comparative and progressive approach toward the design of multifunctional hydrogels. This work not only demonstrates the feasibility of dual viii biofunctionalization but also highlights the critical roles of molecular architecture, crosslinking strategy, and material origin in shaping biological responses. Ultimately, this thesis contributes to the development of smart biomaterials that are both cell-instructive and antibacterial, aligning with the growing clinical need for adaptable, multifunctional solutions in regenerative medicine and infection-prone environments.
  

DOCTORAL DEGREE IN BUSINESS ADMINISTRATION AND MANAGEMENT

• SHEN, SHUYI: Chinese Expatriate Employee Perceptions of Talent Management
  
  Author: SHEN, SHUYI
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN BUSINESS ADMINISTRATION AND MANAGEMENT
  Department: Department of Management (OE)
  Mode: Normal
  Deposit date: 15/10/2025
  Reading date: 02/12/2025
  Reading time: 10:30
  Reading place: FIB Sala de Juntes B6-planta 1
  Thesis director: GALLARDO GALLARDO, EVA | FERNANDEZ ALARCON, VICENÇ
  Thesis abstract: In recent years, the global expansion of Chinese multinational corporations has led to an increasing reliance on expatriate assignments. Despite the growing strategic role of these assignees, little is known about how they perceive their status within talent management systems. This study investigates how Chinese expatriate employees are identified by organizations and how they interpret their inclusion—or exclusion—in organizational TM initiatives based on their perceptions of organizational justice.Adopting a longitudinal, qualitative, and inductive research design, this study draws on initial in-depth interviews with 36 early-career developmental expatriates, followed by a second round of interviews with 19 selected participants within a 12-month period. Grounded in theories of identity construction, identity work, and social comparison, the study investigates how perceptions of TM evolve over time and how expatriates make sense of their talent status across global postings.The findings reveal inconsistencies in talent identification practices, including varied selection criteria across acquisition channels and a lack of alignment between organizational and individual definitions of talent. While headquarters’ classification of expatriates as “talent” serves a symbolic sensegiving function, employees perceive distributive justice primarily through the developmental resources they receive, rather than the talent label itself. Conversely, those excluded from talent pools engage in identity reconstruction and adopt adaptive coping strategies.Additional insights highlight the positive impact of inclusive TM practices and expose tensions arising from dual-track systems in host-country subsidiaries. The study also identifies several mechanisms that undermine expatriates’ talent identity construction and increase turnover risk, including excessive role ambiguity, superficial sensegiving, rigid hierarchical cultures, and insufficient intercultural capacity-building.By shifting the analytical focus from expatriate task performance to long-term identity construction and career development, this research offers novel insights into the global talent management practices of CMNCs. It contributes to the intersection of global talent management and organizational justice by contextualizing TM within Chinese organizational environments. The study calls for more transparent, inclusive, and development-oriented TM systems that account for the evolving identity and motivational needs of expatriates in international assignments.
  

DOCTORAL DEGREE IN CHEMICAL PROCESS ENGINEERING

• ESPEJO DELGADO, VICENÇ: Analysis and modelling of explosions in gas-fired combustion chambers
  
  Author: ESPEJO DELGADO, VICENÇ
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CHEMICAL PROCESS ENGINEERING
  Department: Department of Chemical Engineering (EQ)
  Mode: Normal
  Deposit date: 28/10/2025
  Reading date: 17/12/2025
  Reading time: 11:30
  Reading place: Escola d'Enginyeria de Barcelona Est (EEBE) Edifici A - Aula A0.02 Campus Diagonal-Besòs (CDB) https://eebe.upc.edu/ca/lescola/com-arribar
  Thesis director: CASAL FABREGA, JOAQUIM | PLANAS CUCHI, EULALIA
  Thesis abstract: Combustion chambers are a common equipment widely used in many industries to retrieve heat from fuels (such as in boilers, furnaces, and other fired heaters). Despite the well-documented explosion hazards associated with this equipment, accidents continue to be reported periodically in the industry. The consequences of such events can be catastrophic, leading to severe damages to the equipment, surrounding structures or other equipment, as well as injuries or fatalities.This thesis focusses on the study combustion chamber gas-fired explosion scenarios as a result of the accumulation of unburned gas inside the firebox until flammable conditions are reached, and ignition occurs. As an initial step, a historical analysis of accidents was conducted to typical accident sequences and to highlight the importance of different contributing causes. The main objective of the work is therefore to study these scenarios and provide insights that may improve safety protection design, risk assessments and engineering practices for gas-fired combustion chambers in industrial applications.Some experimental work was found during bibliographic research for similar geometry enclosures, but limited in size, up to 64 m3. However, industrial combustion chambers can reach volumes of thousands of cubic meters. Full-scale experimentation on such equipment would be costly and would require extensive infrastructure to contain, isolate and monitor the explosions. As an alternative, this research employs simulations with FLACS, a Computational Fluid Dynamics (CFD) software widely validated for explosion scenarios, to study the considered explosion cases.The effects of explosions and their dependence on combustion chamber geometry characteristics were evaluated taking special attention on the influence of explosion panels, internal elements, chamber size, as well as the fuel used in explosion development and maximum peak pressures reached. To assess the external effects on the surroundings, an adaptation of the TNT-equivalency model was developed, providing a method to adjust the model yield factor to the combustion chamber explosion consequences.Finally, fuel dispersion inside the chamber was simulated to characterize how unburned fuel evolves over time when introduced through burners. These scenarios were related with the explosion sequences identified in the historical accident analysis. The influence of furnace duty, fuel premixing with air, and burner configuration (single vs. multiple burners) was investigated. Ignition of the accumulated fuel at different dispersion times resulted in explosions with different fuel amounts and concentrations and was also assessed to capture the impact of explosion onset at different stages of scenario evolution. The outcomes of this thesis highlight the effect of key combustion chambers characteristics on explosion phenomena. The evolution of the scenarios identified a critical “trend shift” period, that corresponded to the the timeframe to reach hazardous concentrations. Sensitivity analyses considering different fuels, air pre-mixing, burner configurations or maximum duty per chamber volume revealed general trends applicable to other combustion chamber designs. Overall, the findings provide valuable insights into explosion phenomena in combustion chambers, and offer practical guidance for safer systems design, as well as safeguards effectiveness criteria to be considered in risk assessments.
  
• RODRÍGUEZ ALEGRE, RUBÉN: Wastewater treatments in real case studies: separation & purification in the framework of the circular economy
  
  Author: RODRÍGUEZ ALEGRE, RUBÉN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CHEMICAL PROCESS ENGINEERING
  Department: Department of Chemical Engineering (EQ)
  Mode: Article-based thesis
  Deposit date: 02/10/2025
  Reading date: 03/12/2025
  Reading time: 11:00
  Reading place: Sala Polivalent I - Edifici IEEBE - Campus Diagonal Besòs https://eebe.upc.edu/ca/lescola/com-arribar
  Thesis director: PEREZ MOYA, MONTSERRAT | GARCIA MONTAÑO, JULIA | YOU CHEN, XIALEI
  Thesis abstract: This thesis explores innovative strategies for resource recovery and water reuse from wastewater streams in three key sectors (agrifood, industrial, and urban) by integrating membrane technologies and chemical precipitation. While membrane and chemical treatments have been widely studied independently, this work is among the first to demonstrate their combined and sector-specific application in real or pilot-scale scenarios, framed within circular economy principles.In the agrifood sector, pig slurry liquid fraction was used as a representative effluent. This thesis is the first to evaluate both acidification and basification strategies for nutrient recovery from SLF using a membrane-based treatment train. Acidification, combined with microfiltration and forward osmosis, enabled >80% recovery of NPK nutrients. In contrast, basification allowed for selective nitrogen recovery using precipitation and membrane-assisted stripping. From these strategies, the basic treatment was validated under real on-farm conditions over two years, demonstrating its seasonal robustness and nitrogen recoveries between 46–56%. Water recovery for reuse in irrigation ranged from 39–75%, confirming the viability of decentralised systems in rural settings and marking a significant advance beyond laboratory-scale approaches.For the industrial sector, this thesis presents a novel circular treatment train for acid mine drainage, integrating reverse osmosis with bipolar membrane electrodialysis and chemical precipitation, a combination not previously applied to this matrix. This enabled the selective recovery of high-value metals (Al, Zn, Cu, Mn, Mg, Ca) with >60% efficiency, along with 97% water recovery and in-situ regeneration of NaOH. This approach shifts the paradigm from pollutant removal to resource valorisation, offering a replicable model for sustainable mining wastewater treatment with chemical reuse loops.In the urban context, microplastic pollution was addressed through the first systematic study on how microfiltration membrane configuration affects recovery efficiency. Using synthetic wastewater and varying spacer geometries and sizes, it was found that smaller diamond or corrugated geometries significantly improved performance, with recovery rates >99% and water reuse at 80%. This work provides practical design guidance for future applications in microplastic retention under real wastewater conditions.Additionally, circularity indicators (resource and water recovery) were applied to each case study, allowing for a quantitative assessment of circular performance. All recovered products and waters were analysed following the current Spanish legislation, confirming their potential reuse in irrigation, industrial, and cleaning applications. Overall, this thesis not only demonstrates the technical feasibility of integrated membrane–chemical systems, but also establishes their role as scalable, circular solutions for wastewater treatment across multiple sectors.
  

DOCTORAL DEGREE IN CIVIL ENGINEERING

• MARTORELL PONS, LLUÍS: IGA application on crashworthiness CAE analysis including advanced plasticity and ductile fracture
  
  Author: MARTORELL PONS, LLUÍS
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CIVIL ENGINEERING
  Department: Barcelona School of Civil Engineering (ETSECCPB)
  Mode: Normal
  Deposit date: 16/10/2025
  Reading date: 01/12/2025
  Reading time: 12:00
  Reading place: Sala Zienkiewich (CIMNE) Building C1, UPC - Campus North, Gran Capitan S/N 08034 Barcelona
  Thesis director: ROSSI BERNECOLI, RICCARDO | BARBU, LUCIA GRATIELA
  Thesis abstract: In the contemporary automotive engineering landscape, increasing demands for efficiency, safety, and sustainability have intensified the pressure on development processes. The reduction of the overall duration of the development of a new vehicle is a key focus for the manufacturers. Despite significant advancements in both Computer Aided Design (CAD) and Computer Aided Engineering (CAE) technologies, these domains remain disconnected, resulting in information loss, workflow inefficiencies, and extended development cycles. While Isogeometric Analysis (IGA) has emerged as a promising approach to bridge this divide by maintaining geometric exactness throughout the analysis chain, its practical implementation in industrial contexts remains low, mainly relegated to investigation on its use but not application in actual production projects.This thesis analyzes the automotive CAD and CAE processes to identify current barriers to IGA implementation by addressing the full spectrum of implementation challenges, from technical limitations to workflow integration and cultural adoption barriers. The focus is on crashworthiness applications but Noise, Vibration, and Harshness (NVH) applications is also explored. The methodology adopted in this work follows a multi-faceted approach. First, a historical analysis of CAD and CAE evolution in automotive applications reveals the fundamental origins of their disconnection. Second, the industrial material testing and modeling for crash simulations is explored, a novel material characterization framework is developed, introducing the Non-Isochoric Plasticity Assessment (NPA) methodology to identify when materials deviate from the traditional assumptions of plasticity, and the use of pressure-dependent plasticity models. Third, numerical investigations identify and analyze the "Cross-Talk effect" in trimmed IGA, culminating in the development of a detection algorithm. Finally, case studies demonstrate the application of IGA in real automotive components, establishing workflow guidelines for industrial implementation.This research aims to reduce the gap between academic IGA research and industrial IGA application by providing insights for both communities. For academic researchers, it highlights the technical, operational, and cultural barriers that currently prevent widespread implementation. For industrial practitioners, it offers a roadmap for incremental IGA adoption that respects existing workflows while leveraging the advantages of the technology. By addressing both theoretical capabilities and practical limitations, this work establishes a foundation for more integrated automotive development processes that can respond to the increasing complexity of vehicle design and performance requirements.
  
• TARIN TOMAS, JUAN CARLOS: Optimización de dispositivos flexoeléctricos.
  
  Author: TARIN TOMAS, JUAN CARLOS
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CIVIL ENGINEERING
  Department: Barcelona School of Civil Engineering (ETSECCPB)
  Mode: Normal
  Deposit date: 29/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: ARIAS VICENTE, IRENE | GRECO, FRANCESCO
  Thesis abstract: This thesis develops a strategy to study the optimization on flexoelectric devices. There are nowadays many electromechanical devices , sensors, actuators and energy harvesters, that rely on the basis of the well-known piezoelectric effect, but not all materials exhibit this effect. The most widely used piezoelectric materials show limitations in terms of fracture toughness, toxicity, biocompatibility and temperature range of operation. A novel alternative is provided by flexoelectricity, which, unlike piezoelectricity, appears in all dielectric materials. Flexoelectricity is a size dependent electromechanical coupling which manifest itself at submicron scales and relies on the generation of field gradients inside the material. It has been recently shown, that the flexoelectric response to field gradients in the materials can be conveniently accumulated to produce a macroscopic effective piezoelectric-like response by material architecture. Through the suitable geometry of a repeating unit, piezoelectric metamaterials can be conceived to produce a net electromechanical response even when built from non-piezoelectric base materials, and thus devoid of some of the above mentioned limitations. The design of such piezoelectric metamaterials exploiting flexoelectricity poses numerous challenges both theoretical and computational. Flexoelectricity is a gradient-mediated property, and thus requires additional physical and engineering intuition beyond the homogeneous setups of piezoelectricity. The governing equations of flexoelectricity are a coupled system of fourth-order PDEs, which require solution methods beyond standard finite elements providing the required continuity. In recent work, these issues have been addressed in detail, identifying the main design concepts for piezoelectric metamaterials and developing suitable solution methods. In the present thesis, we focus on the systematic rational design of piezoelectric metamaterials and devices exploiting the flexoelectric effect. A useful tool towards this goal is topology and shape optimization with multiple and possibly conflicting objectives. An important challenge is the high-computational cost of solving flexoelectric boundary value problems in general geometries. We will thus aim at devising efficient optimization strategies to reduce the computational cost, introducing machine learning techniques to alleviate the need for detailed and accurate simulations for every design in the optimization process.
  

DOCTORAL DEGREE IN COMPUTATIONAL AND APPLIED PHYSICS

• RIU I VICENTE, JORDI: Scaling Quantum Optimization Algorithms: Advancing Techniques for Handling Industrial-Level Workloads with Artificial Intelligence
  
  Author: RIU I VICENTE, JORDI
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTATIONAL AND APPLIED PHYSICS
  Department: Department of Physics (FIS)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 26/11/2025
  Reading time: 11:00
  Reading place: Sala d'Actes, Edifici Vèrtex, Plaça d'Eusebi Güell 6, Barcelona
  Thesis director: GARCIA SAEZ, ARTURO | MONRAS BLASI, ALEXANDRE
  Thesis abstract: This thesis integrates traditional optimization methods and artificial intelligence techniques with near‑term quantum algorithms to accelerate the adoption of hybrid protocols for large‑scale, time‑sensitive decision problems. These techniques can manage millions of variables, but they are very expensive computationally. Conversely, quantum approaches such as quantum annealing and QAOA promise more efficient exploration of complex solution landscapes, yet they are hampered by severe hardware limitations and steep training challenges like barren plateaus. By combining these paradigms, the work offloads subproblems suited to classical resources and reserves quantum circuits for the most demanding kernels.We first present the formulation of a relevant use case of a medical‑drone logistics network modeled across Catalonia’s mountainous terrain. An initial discrete‑time MILP with millions of binary variables captures the relevant properties but exceeds the reach of current quantum hardware. A continuous‑time reformulation later reduces variable count. Resource estimates confirm that hardware scaling alone cannot close the gap, underscoring the need for new algorithmic strategies.To address NISQ‑era circuit depth, the thesis introduces RL‑ZX, a reinforcement‑learning–driven quantum compiler built on the ZX‑Calculus framework. Quantum circuits translate into feature‑annotated ZX diagrams, which a Graph Attention Network encodes. An agent trained with proximal policy optimization then applies ZX rewrite rules to minimize two‑qubit gates under device‑specific cost metrics. When evaluated on large, unseen circuits, RL‑ZX outperforms leading heuristics almost universally, yielding shallower, higher‑fidelity circuits.Next, we automate the process of including hard constraints in QUBO formulations. We design a Graph Neural Network and train it to predict penalty magnitudes in a single inference pass, treating variables and constraints as nodes and edges annotated with energy‑difference features. Integrated end‑to‑end with a differentiable QAOA simulator, this approach uses ground‑state sampling probability as its training signal. Experiments on random instances of the knapsack problem, TSP, and assignment tasks demonstrate large improvements in solution probability compared to analytic bounds.The third topic under study is the choice of starting state for VQAs. By measuring coherence via relative entropy, we show that higher‑coherence initial states produce better approximation ratios in fixed‑depth QAOA on small Max‑Cut instances. A tensor‑network imaginary‑time evolution protocol generates Matrix‑Product States approximating pure Gibbs states, that are later mapped to shallow quantum circuits. This initialization strategy suggests a clear path toward improved performance as quantum hardware matures.Finally, two “constraint-satisfying” ansätze are developed to prepare valid solutions directly, eliminating penalty terms. The first builds assignments incrementally using multi‑controlled rotations, while the second creates superpositions of partial assignments before deterministically completing them with ancilla qubits.By uniting these advances—hybrid classical/quantum decomposition, RL‑driven compilation, GNN‑based penalty tuning, coherence‑aware initialization, and feasible‑only state preparation—the thesis lays a scalable, modular foundation for achieving quantum advantage in real‑world optimization.
  

DOCTORAL DEGREE IN COMPUTER ARCHITECTURE

• ALLKA, XHENSILDA: Enhancing Data Quality in IoT Monitoring Sensor Networks
  
  Author: ALLKA, XHENSILDA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 31/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: BARCELÓ ORDINAS, JOSE MARIA | GARCÍA VIDAL, JORGE
  Thesis abstract: In recent years, technological development and an increased number of cars among other factors, have influenced air pollution levels. This increase in levels has also increased the need to monitor them, as they are directly related to human health and the economy. To monitor air pollution, the government has deployed precise monitoring stations, which are expensive to deploy and maintain. Due to their cost, they are not widely distributed. However, since air pollution can change over short distances, the distribution of these stations can be insufficient. Recently, a solution has emerged: the use of low-cost sensors (LCSs), which provide broader coverage at a much lower cost. However, these LCSs have one drawback: the quality of the data they provide is poor.Current research in this field has employed machine learning (ML) models to calibrate these LCSs, thereby enhancing the quality of the data they provide. In an Internet of Things (IoT) monitoring network, the quality of data is closely associated with decision-making processes. This thesis focuses on enhancing the data quality provided by the LCSs from two perspectives: improving calibration performance and detecting anomalies and outliers. The objective of both of these perspectives is to ensure data accuracy and reliability.The first part of the thesis focuses on the improvement of the calibrated data provided by the LCSs and the detection of the concept drift and the update of the parameters of the current calibration model such that it adapts to the new conditions. We are enhancing the quality of the calibrated data by implementing a model pattern-based approach. Our proposed methods, Temporal Pattern Based Denoising (TPB-D) and Temporal Pattern Based Calibration (TPB-C), improve the quality of the calibrated data. Given that environmental conditions are subject to change over time, it is essential to update the parameters of the calibration model. We proposed Window-based Uncertainty Drift Detection and Recalibration (W-UDDR), a system capable of detecting the presence of concept drift (i.e., environmental changes).The second part of the thesis focuses on the reliability of the data. Sensors, regardless of their cost, are often prone to irregularities such as outliers, anomalies, or drift, which can be caused by various factors. It is critical to identify these irregularities, as the data will be incorporated into the training of the model related to other tasks. In this thesis, three distinct scenarios were examined. The first one is related to the detection of outliers in the edge. In this case, we proposed the Edge Streaming Outlier Detection (ESOD) framework. ESOD is a simple and lightweight framework that can identify outliers in the edge with a limited amount of memory. The system offers two approaches: real-time and near real-time. The near real-time approach involves minor delays in decision-making. The second approach is related to the detection of more complex irregularities, such as anomalies in a given sensor. This scenario is distinct from the first one in that it offers offline anomaly detection capabilities. We proposed spatiotemporal correlation recurrent autoencoder anomaly detection (STC-RAAD), which demonstrated satisfactory performance in detecting anomalies in a given sensor. It is worth noting that the third scenario pertains to the detection and localization of anomalies in a network of sensors. This is of particular relevance in scenarios where the identification and precise location of the source of an anomaly are crucial. We hereby propose a pattern-based attention recurrent autoencoder anomaly detection (PARAAD) method. This method is designed to detect and localize anomalies in sensors.
  
• KHABBAZAN, BAHAREH: Improving Memory-centric Architectures for Accelerating Cognitive Computing Workloads
  
  Author: KHABBAZAN, BAHAREH
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 31/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: RIERA VILLANUEVA, MARC | GONZÁLEZ COLÁS, ANTONIO MARIA
  Thesis abstract: The rapid advancements in deep neural networks (DNNs) have led to increasingly complex and memory-intensive workloads, posing significant challenges for traditional computing architectures. Excessive data movement, computational inefficiencies, and energy constraints limit the scalability of DNN accelerators. This thesis addresses these challenges by proposing memory-centric approaches to optimize DNN execution through efficient quantization, in-memory processing, and data movement reduction.We first introduce DNA-TEQ, an adaptive exponential quantization scheme that minimizes memory footprint and eliminates the need for conventional multipliers, significantly enhancing energy efficiency. Experimental results show that DNA-TEQ reduces the memory footprint by 40% on average compared to the 8-bit integer baseline. The hardware processing-near-memory (PnM) accelerator designed to benefit from DNA-TEQ further improves inference latency by 1.5× while maintaining accuracy comparable to full-precision models.Next, we present QeiHaN, a PnM accelerator that employs base-2 exponential quantization and an implicit bit-shifting technique to reduce redundant memory accesses and optimize DNN inference. Our evaluations demonstrate that QeiHaN reduces memory movement by 67%, leading to a 4.2× speedup in execution time and a 3.5× reduction in energy consumption compared to conventional baseline architectures.Lastly, we propose Lama, a lightweight memory access mechanism that enhances lookup table (LUT)-based processing-in-memory (PuM) architectures by enabling parallel, column-independent accesses within DRAM mats, supporting up to 8-bit integer SIMD operations for large-scale models. The experimental results show that Lama significantly reduces memory commands for SIMD operations compared to the state-of-the-art PuM techniques. We further leverage Lama to design LamaAccel, an HBM-based large language model (LLM) accelerator that executes efficiently without modifying DRAM timing parameters. LamaAccel outperforms GPUs by up to 19×, achieving substantial energy savings in low-precision layers.The proposed techniques collectively reduce data movement, optimize memory utilization, and improve computational efficiency. Our findings demonstrate that memory-centric approaches can significantly enhance DNN acceleration, offering scalable and energy-efficient solutions for next-generation AI systems.
  
• OLIVER SEGURA, JOSÉ: Accelerating SpMV on HBM-equipped FPGAs: Hardware-Software Co-design and Collaboration
  
  Author: OLIVER SEGURA, JOSÉ
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 27/10/2025
  Reading date: 28/11/2025
  Reading time: 12:30
  Reading place: Sala C6-E101
  Thesis director: AYGUADÉ PARRA, EDUARD | MARTORELL BOFILL, XAVIER
  Thesis abstract: SpMV is a key linear algebra kernel at the core of many algorithms across multiple knowledge domains. Its memory-bound nature and its low arithmetic intensity make its efficient implementation a challenging problem. Usual mechanisms present in general-purpose microprocessors, such as cache memories, become useless without further data transformation as the size of the problem grows beyond the capacity of the cache. The capability of FPGAs to generate application-specific logic and memory hierarchies results in performant and energy-efficient designs. This has made them an interesting alternative when trying to efficiently implement SpMV. The push by vendors to position them as HPC accelerators and the inclusion of HBM in the last generations of boards have increased this trend. Most SpMV implementations for FPGAs allow to work exclusively using single-precision floating-point arithmetic, while in the context of HPC applications, double-precision floating-point arithmetic is usually required. CSR or slightly modified versions of it are usually used as the basis for these implementations. This limits inter and intra-row parallelism due to conflicts in memory accesses, requiring the implementation to include complex logic such as arbitration or stall/retry mechanisms or to use replicated memories, increasing resource usage and limiting the scalability of the designs. This thesis presents two proposals to leverage the features offered by FPGAs, especially HBM and customizable memory hierarchies, to further improve the achieved performance and, in the case of the second proposal, allowing for a precision-agnostic design that can be synthesized to work with different arithmetic types as required.The first proposal consists of a double-precision FPGA co-designed SpMV accelerator and matrix representation. Instead of using CSR as the basis, the representation and the accelerator are defined considering all the advanced features that FPGAs offer, in a co-design approach. This approach allows maximization of inter-row and intra-row parallelism by allowing simultaneous processing of several matrix values per cycle in a fully pipelined fashion without requiring complex logic or memory replication. The proposed matrix representation allows the easy partitioning of work among different accelerators and the efficient use of HBM bandwidth. The evaluation shows that the proposed implementation outperforms state-of-the-art implementations in terms of absolute, bandwidth-relative, and energy-relative performance.The second proposal builds on the first one, increasing its arithmetic efficiency. It does so in different ways. In the first place, it improves the efficiency of the proposed encoding by reducing the amount of metadata required to process the matrix. In the second place, it increases the useful data ratio of the transformed representation by considering new hierarchical abstractions within the matrix. In the third place, it repurposes zero-padding, when present, to act as a carrier of useful data. This proposal is highly parametrizable, including the possibility of using it to generate designs working with different data types without requiring more changes than setting the desired data type at compile time. The evaluation shows that this proposal significantly improves over the first one in double-precision arithmetic. Single-precision results demonstrate its capability to improve the performance offered by state-of-the-art designs that use much higher bandwidth.
  
• RODRIGUEZ FERRANDEZ, IVAN: Mixed software/hardware-based fault-tolerance techniques for complex COTS system-on-chip in radiation environments
  
  Author: RODRIGUEZ FERRANDEZ, IVAN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 01/12/2025
  Reading time: 09:00
  Reading place: C6-E101
  Thesis director: KOSMIDIS, LEONIDAS | TALI, MARIS
  Thesis abstract: This thesis titled “Mixed Software/Hardware-based Fault-tolerance Techniques for Complex COTS System-on-Chip in Radiation Environments" explores the challenges and solutions for integrating high-performance Commercial Off-The-Shelf (COTS) System-on-Chip (SoC) technologies, specifically GPUs, into space applications. These automotive-grade systems offer significant computational capabilities but face unique challenges in radiation-intense environments typical of space. The research investigates these challenges and proposes solutions to enhance the reliability of such systems. A key component of the thesis involves the comprehensive evaluation of modern embedded GPUs under space-like conditions, including exposure to proton and heavy-ion radiation. This analysis identifies vulnerabilities such as Single Event Effects (SEE) , which can cause faults like Single Event Upset (SEU), Single Event Functional Interrupt (SEFI), and Single Event Latch-up (SEL). To support these evaluations, the author develops the OBPMark suite, an open-source benchmark tailored for assessing the performance and efficiency of GPUs in space-specific computational tasks. To address the faults identified, the thesis proposes innovative software-based fault mitigation strategies. These include the design of fault-tolerant GPU kernels and middleware solutions that improve error detection and recovery. The effectiveness of these methods is demonstrated through both simulation and radiation testing. Additionally, the research presents hardware-level innovations, such as the development of application-specific integrated circuits (ASICs) and specialized printed circuit boards (PCBs), to enhance system resilience without compromising performance. This work significantly contributes to the field of space computing by creating a robust framework for evaluating and mitigating radiation effects in complex COTS SoCs. It bridges the gap between the cost-effectiveness and performance of commercial technologies and the reliability demands of space-grade applications. The findings of this thesis pave the way for the adoption of high-performance embedded GPUs in future space missions.
  
• SERRACANTA PUJOL, BERTA: Accelerating the Cloud: An Application-Agnostic Approach to Network and Compute Optimization
  
  Author: SERRACANTA PUJOL, BERTA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 19/09/2025
  Reading date: 21/11/2025
  Reading time: 10:00
  Reading place: Sala 103 - Aula Teleensenyament - Edif. B3 - Planta 1
  Thesis director: CABELLOS APARICIO, ALBERTO | RODRÍGUEZ NATAL, ALBERTO
  Thesis abstract: This thesis explores how to enhance the performance of cloud applications by addressing inefficiencies across both the network and compute layers of modern distributed systems. As cloud-native applications grow more complex and are deployed across heterogeneous, geographically distributed infrastructures, traditional abstractions, though foundational for scalability and modularity, have begun to constrain opportunities for global coordination and responsiveness. To overcome these limitations, this work introduces two complementary approaches: one that improves network resource utilization without requiring developer involvement, and another that enhances compute-side elasticity through a smarter, more proactive autoscaling mechanism. Both approaches are guided by a common design philosophy: introducing context-aware intelligence in a minimally disruptive way, maintaining full compatibility with existing architectures, infrastructure, and developer workflows.The first part of the thesis focuses on Network-Application Integration (NAI), specifically targeting performance improvements in inter-datacenter communication. To this end, it proposes an application-agnostic solution based on extended Berkeley Packet Filter (eBPF) and eXpress Data Path (XDP) technologies. By dynamically identifying and separating short and long Transmission Control Protocol (TCP) flows at the network ingress, the system enables differentiated routing through distinct network tunnels, thereby mitigating queuing delays and reducing flow completion times. A key advantage of this approach is that it operates transparently, requiring no modifications to applications or developer-provided annotations, making it highly deployable within existing environments. Testbed experiments demonstrate that this technique significantly reduces latency and improves resource utilization in hybrid, multi-datacenter scenarios.The second part of the thesis turns to the compute domain, focusing on autoscaling mechanisms in Kubernetes-managed microservice environments. Recognizing the limitations of existing reactive scaling strategies, the work develops a control-theoretic model of the Kubernetes Horizontal Pod Autoscaler (HPA), formally analyzing its stability and responsiveness. Based on these insights, a new context-aware HPA is introduced, which incorporates upstream CPU metrics from the application’s service graph to anticipate downstream load changes. This proactive strategy enables more efficient and stable scaling decisions, improving responsiveness and reducing latency during traffic spikes. Notably, it achieves these gains without relying on complex performance models or machine learning, preserving simplicity and compatibility with standard Kubernetes tooling.Overall, the two approaches presented in this thesis offer practical methods for improving the performance and efficiency of distributed cloud applications, with a focus on compatibility with existing systems and workflows. Rather than proposing disruptive architectural changes, both solutions extend current abstractions to introduce additional context-awareness where it can be most effective. The results suggest that incremental, deployable enhancements to orchestration and networking layers can help address emerging challenges in scalability and responsiveness, making them suitable candidates for integration into real-world cloud-native environments.
  

DOCTORAL DEGREE IN COMPUTING

• PONTÓN MARTINEZ, JOSE LUIS: Learning Data-driven Character Animation for Avatars in Virtual Reality
  
  Author: PONTÓN MARTINEZ, JOSE LUIS
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTING
  Department: Department of Computer Science (CS)
  Mode: Normal
  Deposit date: 27/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: ANDUJAR GRAN, CARLOS ANTONIO | PELECHANO GOMEZ, NURIA
  Thesis abstract: The accelerating trend of remote interaction, driven by globalization and digital communication, underscores the need for richer, more immersive virtual collaboration. While current 2D video platforms enhance communication, Virtual Reality (VR) offers the unique potential for truly natural 3D interaction. Accomplishing this, however, critically depends on accurately representing human motion and achieving presence within virtual environments.This thesis addresses the challenge of achieving real-time, high-fidelity, and perceptually natural full-body self-avatar animation within VR environments using consumer-grade tracking devices. Accurate self-avatars are fundamental for inducing a strong Sense of Embodiment and enabling effective non-verbal communication, yet current methods often struggle with the inherent sparsity and variability of available sensor data.We first address fundamental aspects of animation fidelity and perceptual realism, and introduce methodologies for precise avatar skeleton adjustment, which significantly mitigate issues arising from mismatches between a user's physical proportions and their virtual representation. We also study various interaction metaphors to minimize visual discrepancies between real controllers and virtual hands, thereby enhancing user embodiment and task performance. These studies underscore the importance of accurate animation and lay the groundwork for learning-based approaches to achieve natural and temporally coherent motion from sparse inputs, overcoming the limitations of traditional inverse kinematics.Building upon these insights, the thesis explores the development of data-driven reconstruction methods that can handle diverse and ambiguous sensor inputs. We propose a novel deep learning-based system that accurately reconstructs full-body poses from minimal consumer-grade VR trackers, effectively addressing the underdetermined nature of this problem. Recognizing the inherent one-to-many mapping problem in sparse input, where a single input can correspond to multiple plausible poses, we then explore the potential of generative AI. Our work demonstrates how Variational Autoencoders (VAEs) can enable fine-grained control and adaptability to variable sensor configurations through latent space optimization, while diffusion models facilitate multimodal reconstruction from novel sensor types, such as pressure-sensing insoles.
  

DOCTORAL DEGREE IN CONSTRUCTION ENGINEERING

• RAMIREZ PEREZ, ALEXIS JOHARIV: Comportamiento a flexión y cortante de un tablero continuo de vigas pretensadas con tendones de polímeros reforzados con fibras (FRP)
  
  Author: RAMIREZ PEREZ, ALEXIS JOHARIV
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CONSTRUCTION ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 22/10/2025
  Reading date: 12/01/2026
  Reading time: 11:00
  Reading place: C1-002
  Thesis director: OLLER IBARS, EVA MARIA | MARI BERNAT, ANTONIO RICARDO
  Thesis abstract: The durability of reinforced concrete structures is mainly compromised by steel corrosion, which generates high maintenance costs and reduces structural safety. Fiber-reinforced polymers (FRP) represent an alternative of great interest, as they provide high specific strength and are not susceptible to corrosion. However, their application as active reinforcement in continuous prestressed members is still very limited, due to the scarce experimental research on their structural performance and the absence of specific design guidelines.The main objective of this dissertation is to analyze the flexural and shear behavior of a two-span continuous bridge at 1/3 scale, built with precast prestressed girders and a cast in situ reinforced concrete slab, using carbon carbon fiber composite cables “CFCC” tendons as active reinforcement. The research was organized into three phases: (1) characterization of carbon fiber (CFRP) bars, glass fiber (GFRP) bars, and CFCC tendons, with the latter selected for prestressing due to their suitability; (2) a flexural test on span 1, with a concentrated load applied at midspan, to study the global flexural behavior at the serviceability and ultimate limit states; and (3) a shear test on span 2, with a concentrated load applied 1.6 m from the end support, to evaluate shear strength, effectiveness of GFRP stirrups, and the influence of CFCC prestressing. The results were compared with numerical simulations using the CONS program and with the CCCM analytical model adapted to FRP tendons. The experimental tests showed that CFCC tendons reached 62–76% of their ultimate strength without anchorage slip in the flexural test, confirming their reliability as active reinforcement. Failure was governed by shear-off at the girder–slab interface. In shear, failure occurred after a characteristic diagonal cracking pattern and progressive redistribution of stresses between spans, while shear-off failure was avoided through a reinforcement added after the flexural test.The overall contribution of this dissertation lies in providing the first comprehensive experimental, analytical, and numerical evidence on a continuous bridge prestressed with CFCC tendons. The findings strengthen confidence in the use of FRP in concrete structures, and open new research avenues aimed at optimizing transverse reinforcement and moving towards the codification of this technology.
  
• VALVERDE BURNEO, DAVID ENRIQUE: Desarrollo de nuevos materiales cementicios multifuncionales
  
  Author: VALVERDE BURNEO, DAVID ENRIQUE
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CONSTRUCTION ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Article-based thesis
  Deposit date: 10/10/2025
  Reading date: 19/01/2026
  Reading time: 11:00
  Reading place: C1-002
  Thesis director: SEGURA PEREZ, IGNACIO | GARCIA TRONCOSO, NATIVIDAD LEONOR
  Thesis abstract: This doctoral thesis focuses on the development of multifunctional cementitious materials, combining structural strength with self-sensing capabilities through piezoresistivity, as well as integrating deformation energy dissipation through auxetic structures. The research explores the integration of conductive and structural fibers in cementitious matrices, coupled with the use of advanced manufacturing techniques such as 3D printing and the use of flexible silicones to obtain molds with complex architectures. The objective is to obtain cementitious materials that in addition to possessing structural capacity, have added function capabilities. It is expected that these materials can be used in buildings with self-monitoring, damage prevention, stress sensing pavements, structural elements with higher impact resistance and energy dissipation capabilities. The research begins with an exhaustive bibliographic review, from which the most promising materials have been selected to achieve the proposed objectives. The experimental campaign and data treatment/analysis have been defined. The work continues with the realization of the planned experiments, the analysis of the results, the optimization of the composition and properties of the new cementitious materials, the development of prototypes testing the potential applications.From the achievements obtained in this doctoral thesis we have the following: the research and publication of a cementitious composite reinforced with recycled carbon fibers to obtain a piezoresistive conductive concrete, which presents a variation of the electrical conductivity with respect to the unitary deformation quite evident when the fiber addition contents are around 1% in volume. This makes it an ideal sustainable cementitious material for strain and/or stress detection. This publication can be found in the journal Construction and Building Materials.Another research focuses on the mechanical characterization of cellular auxetic cementitious cementitious composites (which achieve their auxeticity through the presence of ellipsoidal holes in their structure) reinforced with recycled steel fibers. This research successfully characterizes the influence of fiber content on the mechanical response to compression and deformation energy dissipation, while demonstrating the feasibility of using recycled resources. Within this same publication, a family of functions was presented that successfully fit the mechanical response curves (stress-strain, energy dissipated by deformation) that were obtained experimentally. This publication can be found in the Journal of Building Engineering.A third article achieved in this thesis deals with the development of a new type of piezoresistive concrete with auxetic capacity. This material, obtained by combining cellular auxetic cementitious cementitious composites and recycled carbon fibers, is capable of detecting deformations from very low to high levels. Its potential applications in structural monitoring are promising, and the results of this research have been published in Case Studies in Construction Materials.
  

DOCTORAL DEGREE IN EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS

• TIRADO GUTIERREZ, RODOLFO JAVIER: EVALUATION OF STRUCTURAL RELIABILITY. REVIEW OF DESIGN METHODOLOGIES AND SEISMIC PERFORMANCE EVALUATION OF REINFORCED CONCRETE STRUCTURES
  
  Author: TIRADO GUTIERREZ, RODOLFO JAVIER
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 25/09/2025
  Reading date: 20/11/2025
  Reading time: 16:00
  Reading place: ETSECCPB.UPC, Campus NordBuilding C1. Classroom: 002C/Jordi Girona, 1-308034 Barcelona
  Thesis director: VARGAS ALZATE, YEUDY FELIPE | GONZALEZ DRIGO, JOSE RAMON
  Thesis abstract: Seismic risk and structural reliability are fundamental concepts in the design and evaluation of safe and resilient buildings. It should be noted that most of the casualties, injuries and economic losses during an earthquake are associated with damage to civil structures. In this context, there is an urgent need to improve the design and evaluation methodologies of this type of structures. From a numerical perspective, several strategies can address this need, ranging from improvements in modeling and advanced analysis methods, to the probabilistic analysis of the variables involved (such as seismic hazard), including the review of configuration and material properties of the systems under study. Therefore, this research proposes a methodology to evaluate structural reliability using a probabilistic approach, validated through nonlinear dynamic analysis and a statistical cloud study. This methodology constitutes a robust and powerful tool, applicable not only at the building level but also at urban and regional scales. For this purpose, it is proposed to study a set of reinforced concrete buildings with variable configurations in both plan and elevation. These models represent real buildings, that were recently designed and constructed, located in areas of high seismicity in Colombia, for which the most modern seismic-resistant design standards have been followed. This thesis is divided into three main sections: 1) Calculation of improved intensity measures, in terms of efficiency and steadfastness, to derive more accurate fragility curves; 2) Development of a probabilistic analysis methodology to estimate, with high statistical accuracy and reduced time, the dynamic response of tall buildings by using transfer functions; and 3) Evaluation of structural reliability, based on the probability of exceeding different damage indices and thresholds at different levels of seismic intensity. The first part focuses on identifying and developing optimal and improved intensity measures, based on the efficiency and steadfastness they show as correlated with the structural response of complex systems. An optimal seismic intensity measure enables the development of more accurate fragility curves, which are essential for assessing the probability of damage in a structure under different seismic intensity levels. The second part focuses on the development of a structural analysis method based on the transfer function (TF) concept. This mathematical model establishes the relationship between the response of a system and the input excitation. The proposed approach allows probabilistic estimation, while maintaining statistical accuracy, of the nonlinear dynamic response. It aims to overcome the limitations of the high computational cost associated with nonlinear dynamic analysis. Finally, the third part focuses on calculating the structural reliability of two real buildings located in a high seismicity zone, evaluating the probability of exceeding different damage thresholds. The results obtained show that intensity measures based on velocity present a higher correlation with the structural response, regardless of whether they are analyzed as a whole. This will make it possible to evaluate risk scenarios in large areas by means of fragility curves that adequately represent different structural typologies, facilitating a better characterization of urban environments. Likewise, it is observed that the developed method, by using a reduced number of seismic records, allows obtaining reliable results in terms of the principal statistical moments of the structural response of complex systems, and importantly, in a considerably shorter time. In conclusion, this research presents a series of advanced numerical tools that allow the calculation of damage scenarios, while optimally accounting for seismic hazard, and using methodologies that significantly reduce the time required to estimate the structural reliability of a set of buildings.
  

DOCTORAL DEGREE IN ELECTRONIC ENGINEERING

• DE LA VEGA HERNÁNDEZ, JOAQUÍN: Advanced modelling and forecasting methods for electric vehicle batteries based on data analysis with realistic operating conditions.
  
  Author: DE LA VEGA HERNÁNDEZ, JOAQUÍN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ELECTRONIC ENGINEERING
  Department: Department of Electronic Engineering (EEL)
  Mode: Normal
  Deposit date: 22/10/2025
  Reading date: 04/12/2025
  Reading time: 11:00
  Reading place: Sala de Juntas TR5, UPC Terrassa
  Thesis director: ORTEGA REDONDO, JUAN ANTONIO | RIBA RUIZ, JORDI ROGER
  Thesis abstract: This PhD thesis addresses the challenges of modeling, monitoring, and forecasting the behavior of lithium-ion batteries at both the single-cell and pack levels, with a particular focus on improving robustness under realistic operating conditions. The work is motivated by the growing deployment of battery systems in electric mobility and stationary storage, where performance, lifetime, and safety critically depend on reliable state estimation and degradation prediction.At the single-cell level, the thesis introduces two complementary modeling approaches: a parametric voltage–capacity model for constant-current discharge, and a deep learning framework for partial-charge data. These methods enable early detection of degradation trends and real-time estimation of capacity fade, providing interpretable health indicators (HI) and accurate remaining useful life (RUL) forecasts.At the pack level, the thesis explores the additional complexities that arise from cell-to-cell variability, imbalance, and data acquisition issues. A hybrid data imputation methodology based on the Unscented Kalman Filter (UKF) is proposed to reconstruct missing voltage signals at both the cell and branch levels, ensuring continuity of BMS functions such as balancing, SoC/SoH estimation, and fault detection. The method is benchmarked against neural networks, highlighting the trade-off between data-driven accuracy and model-based adaptability. Building on this, this work expands upon the imputation framework by studying how reconstructed signals impact the accuracy of forecasting models. Four reconstruction strategies of increasing complexity (ZOH, ARIMA, UKF, and GRU) were compared, and their outputs were fed into recurrent neural networks (LSTM and GRU) developed for this purpose. These networks were used to predict the remaining time to depletion (RTD) of individual cells under driving conditions. The results demonstrate that the quality of signal reconstruction directly impacts forecasting performance.The contributions are supported by multiple datasets, including public repositories (NASA, Sandia National Laboratories) and a custom experimental testbench capable of executing standardized drive cycles and controlled CC–CV protocols. Together, these datasets provide a rigorous foundation for validation across chemistries and cycling conditionsOverall, the thesis demonstrates how the integration of signal processing and filtering, and machine learning can enhance the reliability of battery models, both for immediate diagnostic tasks and long-term prognostics. The findings contribute toward more robust and practical battery management systems, bridging the gap between academic models and real-world applications in electric mobility and energy storage.
  
• KUMAR, DILEEP: Deep Learning for Improving Resilience of the Sensors in Mars Exploration Missions
  
  Author: KUMAR, DILEEP
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ELECTRONIC ENGINEERING
  Department: Department of Electronic Engineering (EEL)
  Mode: Normal
  Deposit date: 22/10/2025
  Reading date: 24/11/2025
  Reading time: 11:00
  Reading place: Aula de Teleensenyament, edifici B3, campus nord
  Thesis director: DOMINGUEZ PUMAR, MANUEL MARIA | PONS NIN, JOAN
  Thesis abstract: In space exploration missions, scientists have been putting efforts to achieve targeted scientific goals, such as environmental study of Mars through various exploration missions. In recent years, multiple missions have been sent to study the atmosphere of Mars with advance sensing systems. In the process of developing various sensing systems, worst- case scenarios are also considered and their possible solutions become crucial to be implemented to deal with adversities. During the operation of a mission, critical system components such as sensors could face complete or partial damage, and because of that, the mission might fail to send data to the monitoring station. This is a known fact in the case of wind sensors deployed with Mars 2020 Perseverance and Curiosity Rover missions those suffered partial damage. Moreover, at some point, TWINS wind sensor with InSight lander mission also faced power issues. Such scenarios create hurdles in the scientific study of a planet. Various space sensors related problems can be caused by various adversities, such as dust devils at Mars, adaptation of operating points in sensors themselves, and obstacles around the sensing systems. Furthermore, it is not possible to repair or replace a sensor on Mars. Thus, it is crucial to address space sensor problems with remedial techniques to achieve the target objectives.Considering problems occurred in Mars wind sensors, this thesis is focused on investigating data-driven approaches to improve resilience of Mars wind sensors deployed in the last two Mars missions of NASA, namely TWINS (InSight Mission, 2018) and MEDA (Perseverance Mission, 2020) instruments. Various Machine Learning (ML) and Deep Learning (DL) models have been investigated to enhance the resilience of the aforementioned wind sensors in the case of partial failure. These data-driven algorithms are investigated to develop a soft or virtual sensor for Martian wind sensors. Furthermore, Transfer Learning (TL) based approach has been investigated to deal with data scarcity scenarios. These methods have yielded promising results in recovering the data in the event of partial failure. In TWINS investigation, RMSE for velocity is reduced by a factor between 2.43 and 4.78; and for wind angle by a factor between 1.74 and 4.71, compared to the case where only two wind sensing transducers are functioning. For MEDA, the investigated algorithms allowed to recover variables of the wind sensing boards with errors similar to TWINS instrument and in some cases have achieved slightly better results. With the TL approach, the multivariate predictions improve with the RMSE percentage between 10.21% to 22%. In summary, various data-driven methods investigated have illustrated the efficacy and potential in recovering data and dealing with adverse scenarios related to Mars wind sensors.
  

DOCTORAL DEGREE IN ENGINEERING, SCIENCES AND TECHNOLOGY EDUCATION

• MIRÓ MEDIANO, ÀLEX: Defining, Modelling, and Sequencing Complexity in Secondary Mathematics
  
  Author: MIRÓ MEDIANO, ÀLEX
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ENGINEERING, SCIENCES AND TECHNOLOGY EDUCATION
  Department: Institute of Education Sciences (ICE)
  Mode: Normal
  Deposit date: 23/10/2025
  Reading date: 27/11/2025
  Reading time: 11:00
  Reading place: Sala de Juntes . Edifici B6 1ª planta - Campus Nord
  Thesis director: ALIER FORMENT, MARC | MORA SERRANO, FRANCISCO JAVIER
  Thesis abstract: Learning mathematics at secondary school is difficult for many students. In the Catalan education system, national and international evaluations have shown little to no improvement in mathematical competency by the end of secondary education. Instruction design plays a critical role in how this learning happens. The central hypothesis of this thesis is that learning outcomes in secondary mathematics could be significantly better when learning tasks are sequenced from simple to complex. However, determining what makes a mathematical task complex is not straightforward and lacks academic consensus. The main goal of the thesis was to investigate precise definitions, reliable measurement approaches, and effective sequencing strategies of learning complexity in mathematics, specifically within secondary education.The research was framed under cognitive load theory, which explores the educational implications of human cognitive architecture. This architecture includes three memory systems—sensory memory, working memory, and long-term memory—that process and store information. Learning begins when information enters sensory memory, is consciously processed in working memory (which has limited capacity), and is stored in long-term memory as interconnected schemas. These schemas can later be retrieved and used without overloading working memory.Cognitive load theory proposes that the complexity of learning is determined by *element interactivity*, a central construct in the thesis. Element interactivity refers to the number of informational elements that interact or need to be managed simultaneously within a task. The greater the number of elements, the higher the cognitive demand on working memory, increasing the risk of overload and hindering learning.Using this definition, the research began with the development of the *Mathematical Knowledge Matrix* (MKM), a tool to assess the complexity of learning sequences and tasks. Initial analyses revealed limitations in the MKM, as it did not account for other relevant sources of complexity affecting mathematics learning. A literature review showed a knowledge gap regarding the range of complexity factors influencing element interactivity, prompting the need for more precise definitions of mathematical complexity. The first study addressed this by generating a taxonomy of complexity sources and examining whether element interactivity could explain all of them.Although element interactivity was a valuable construct for explaining various forms of mathematical complexity, clear guidelines for using it to assess different sources were lacking. The second study developed methods to measure complexity arising from multiple sources—providing guidelines for applying element interactivity more precisely—and tested their reliability against real students’ perceptions. Students’ perceived complexity data were also used to evaluate the relative influence of each source and to explore how these sources collectively accounted for overall task complexity.Findings from the second study showed that element interactivity arising from the *knowledge* and *operations* of tasks had the strongest influence on perceived complexity. Therefore, the third study aimed to design effective task sequencing strategies considering these two variables separately, in comparison to a control group. Both sequencing approaches improved learning outcomes, but only the *operativity-based* sequence produced statistically significant results.The investigation concluded that element interactivity is an effective construct for measuring mathematics complexity when considered from multiple sources. It can be measured to inform simple-to-complex task sequences that enhance learning. Moreover, the results from the third study suggest that disregarding operational complexity may lead to reduced learning gains, even in already proven didactic strategies.
  

DOCTORAL DEGREE IN ENVIRONMENTAL ENGINEERING

• SAVADKOOHI, MARJAN: An Advanced Control Strategy for Optimizing HVAC System Performance in Non-Residential Buildings
  
  Author: SAVADKOOHI, MARJAN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ENVIRONMENTAL ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 10/09/2025
  Reading date: 24/11/2025
  Reading time: 11:00
  Reading place: Place: ETSECCPBUPC, Campus NordBuilding C1. Classroom: 002C/Jordi Girona, 1-308034 Barcelona
  Thesis director: CASALS CASANOVA, MIQUEL | MACARULLA MARTÍ, MARCEL
  Thesis abstract: This PhD research addresses the scientific and practical challenges in implementing advanced control systems (ACS) for HVAC optimization in non-residential buildings. While adaptive and predictive strategies, especially model predictive control (MPC) and neural network (NN)-based methods, have shown promise in research, real-world deployment remains limited. Barriers include insufficient building historical data, technical limitations of HVAC systems, lack of building energy management systems (BEMS) standardization, and low institutional readiness. To address these, this thesis uses a dual-method approach combining empirical analysis and simulation-based experimentation.First, a survey of 676 non-residential buildings evaluates BEMS practices, focusing on HVAC control, data storage, and predictive control adoption. While smart metering and sensors are increasingly common, predictive control is reported in only 0.6% of buildings. Key barriers identified include a lack of environmental data logging, obsolete HVAC systems that do not support integration with predictive control technologies, limited technical expertise among building operators, and insufficient investment frameworks, particularly in the public sector.To respond to data and implementation challenges, the second part develops and evaluates NN-based predictive controllers using a calibrated building energy model. Eight NN models are trained on varying amounts of historical data to assess impacts on prediction accuracy and HVAC performance. Validations use consistent KPIs for thermal comfort and energy efficiency. Results show that 1–4 months of data are needed for acceptable performance, reaching a performance threshold after two years. Data preprocessing helps in data-limited cases (<100 instances), but adds little value with larger datasets, suggesting a context-specific role.Further analysis explores operational and climatic sensitivities. In cold climates and post-HVAC shutdown periods (e.g., Monday mornings), models struggle due to sparse training data. Performance improves mid-week and in warm zones, highlighting the need for diverse and climate-adapted training data. Compared to rule-based scheduling, NN controllers consistently improve energy use and comfort, especially when supported by adequate data and system configuration.This thesis offers novel insights into deploying intelligent HVAC control systems. It identifies data thresholds for effective predictive control, clarifies preprocessing roles, and provides guidance on model adaptation to climate and operations. It also highlights broader needs such as standardizing data acquisition, training energy professionals, and fostering public-private collaboration to reduce implementation risk. The findings support scalable predictive control in practice and contribute to the goals of energy efficiency, smart building management, and decarbonization.
  

DOCTORAL DEGREE IN GEOTECHNICAL ENGINEERING

• BENHAMMADI, RIMA: Convective mixing in heterogeneous porous media
  
  Author: BENHAMMADI, RIMA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN GEOTECHNICAL ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 22/10/2025
  Reading date: 25/11/2025
  Reading time: 11:00
  Reading place: ETSECCPB. UPC, Campus Nord Building C1. Classroom: 002 C/Jordi Girona, 1-3 08034 Barcelona
  Thesis director: DENTZ, MARCO | HIDALGO GONZÁLEZ, JUAN JOSÉ
  Thesis abstract: This thesis seeks to advance the understanding of convective mixing in heterogeneous porous media, a topic that remains comparatively underexplored compared to its homogeneous counterpart. Through the combination of high-resolution numerical simulations and laboratory experiments, we explore how spatial variability in permeability influences the onset, development, and efficiency of convective mixing processes, with applications to thermal convection, CO$_{2}$ dissolution and reactive transport.First, we begin by investigating thermal convection in the classic Horton-Rogers-Lapwood (HRL) configuration, where permeability fields are modeled as two-dimensional, log-normally distributed random fields with varying variance and correlation lengths. These serve as quantitative measures of the underlying heterogeneity. Our conducted parametric study shows that increasing the variance and/or the correlation length of the log-permeability field enhances segregation, sharpens thermal interfaces, and leads to more irregular flow structures. While the dissolution flux decreases with Rayleigh number in both homogeneous and heterogeneous systems, its sensitivity to permeability variance becomes more pronounced at longer correlation lengths. In highly heterogeneous cases, high-permeability zones near boundaries coincide with stagnation points that influence the formation of temperature plumes and localised strain rates, while the interface width decreases, indicating enhanced stretching and deformation due to the underlying structure.Next, we study CO$_{2}$ convective dissolution in heterogeneous Hele-Shaw cells, via a combined experimental-numerical approach. Heterogeneity is introduced through variations in the cell gap width, corresponding to a log-normal distribution of permeability with fixed variance and correlation lengths. Results show that heterogeneity advances the onset of instability, increases the amplitude and growth rate of convective fingers, and causes more distorted and dispersive flow patterns. However, the dimensionless wavenumber of the instability remains similar to that in homogeneous cells. A comparison of the autocorrelation functions of the fingering patterns and the permeability field shows that heterogeneity increases the dimensionless correlation length of the fingering pattern, which in turn slows down its growth once the finger size becomes comparable to the heterogeneity scale.Finally, we investigate reactive convective dissolution involving the bimolecular chemical reaction \( \A + \B \rightarrow \C \), across four permeability configurations: homogeneous, horizontally layered, vertically layered, and multi-Gaussian log-normally distributed fields. Key metrics such as product mass, reaction rate, front position and width and mixing length are all substantially affected by the structure of the permeability field. Vertically layered and log-normal configurations promote more efficient mixing and faster front progression. Overall, when horizontal correlation length is small relative to the vertical, convective transport and mixing efficiency are maximised.Collectively, these findings demonstrate that it is not simply the presence of heterogeneity, but the specific structure of the permeability, particularly its variance and spatial correlation, that fundamentally governs convective behaviour. The insights gained show the necessity of incorporating geologically realistic heterogeneity into the predictive models.
  
• DAWI, MALIK ALI A: Process-Based Numerical Models to Assess Hydrogeochemical Effects of Microbial Biofilms in Porous Media
  
  Author: DAWI, MALIK ALI A
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN GEOTECHNICAL ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 16/10/2025
  Reading date: 19/12/2025
  Reading time: 10:00
  Reading place: ETSECCPB. UPC, Campus Nord Building C1. Classroom: 002C/Jordi Girona, 1-308034 Barcelona
  Thesis director: SANCHEZ VILA, FRANCISCO JAVIER | STARNONI, MICHELE
  Thesis abstract: Microorganism presence and spatial distribution over time in natural porous media, such as soils, sediments, and aquifers, play critical roles in mediating geochemical processes, influencing contaminant fate, and maintaining ecosystem functionality. In particular, microbial communities in the form of biofilms mediate complex biotransformation reactions, significantly altering the hydraulic properties of the host porous system. These dynamics, coupled with multiscale physical and chemical interactions, present major challenges for the predictive modeling of microbial processes in porous media. This thesis aims at developing a suite of computational models that integrate microbial growth and activity within flow and transport frameworks. The work is structured around three main contributions. First, a hybrid pore-scale model is developed which couples a micro-continuum representation of biofilms with a particle-based transport solver, enabling detailed analysis of how biofilm morphology and structure influence conservative solute transport. Second, this framework is extended to simulate the dynamic biofilm development and its interaction with groundwater flow, introducing a cohesive microporous model for biofilms that incorporates growth, attachment, spreading, and flow-induced detachment. A novel dimensionless number is introduced to characterize the interplay between hydrodynamic forces and biofilm cohesion. Third, we revisit Monod kinetics by proposing a mechanistic two-step reaction scheme that linearizes the growth rate expression, facilitating its integration into particle-based reactive transport models. This formulation is validated against batch experiments and applied to simulate microbial degradation in porous media. Finally, the thesis synthesizes these findings and outlines future directions for model development and experimental integration. By combining theoretical insights with computational advances, this work contributes to a deeper understanding of microbially mediated processes in porous media and provides modeling tools to support both hypothesis testing and experimental research in environmental and engineered systems.
  

DOCTORAL DEGREE IN MARINE SCIENCES

• ANGELINI, RICCARDO: Coastal environment monitoring through satellite, terrestrial and airborne remote sensing
  
  Author: ANGELINI, RICCARDO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN MARINE SCIENCES
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 24/11/2025
  Reading time: 10:30
  Reading place: Civil and Environmental Engineering Department, University of FlorenceVia di Santa Marta 3, Florence
  Thesis director: MASIERO, ANDREA | LUZI, GUIDO | RIBAS PRATS, FRANCESCA
  Thesis abstract: Coastal areas are increasingly threatened by sea-level rise due to climate change and anthropogenic pressures, calling for robust and scalable monitoring tools. The first phase of this thesis implements a comprehensive methodology for semi-automatic shoreline extraction through the use of multispectral satellite imagery (Sentinel-2 and PlanetScope). The extracted shorelines are validated using in situ GNSS surveys and high-resolution orthomosaics along three Mediterranean sandy beaches. The shoreline extraction tool works with several spectral indices tested with thresholding and unsupervised clustering segmentation methods. A high coastline extraction performance is achieved using Sentinel-2, with an average sub-pixel accuracy of 4 m (Mean Absolute Deviation, MAD) obtained from a 10 m pixel. A MAD of 2 m is achieved from imagery at 3 m pixel resolution of PlanetScope. In the second phase, the obtained multispectral satellite shorelines are used to characterize megacusps shoreline undulations with alongshore wavelengths of hundreds of meters and cross-shore amplitudes up to a few tens of meters that can significantly affect beach usability. Subsequent validation with reference data proves satellite-derived shorelines can robustly and accurately describe megacusp parameters such as amplitude and wavelength. Moreover, megacusp evolution can be effectively characterized by combining different types of satellite imagery (Sentinel-2 and PlanetScope), enabling the identification of periods of growth, decay, and migration, even at weekly timescales. This can be a useful tool to manage the impact of these features on Mediterranean beaches.Another phase involves evaluating and correcting the extracted shorelines to tide excursions and wave setup. Corrections based on tide gauge and buoy data show that although absolute displacements are limited, assessing them helps eliminate a potential source of error, justifying their integration into high-accuracy workflows. This would also allow applying the developed methodology to meso- or macro-tidal beaches.Finally, the research incorporates Synthetic Aperture Radar (SAR) imagery (Sentinel-1 and TerraSAR-X) to expand the shoreline extraction tool to periods without light or with clouds, including two other Mediterranean beaches characterized by gravel sediment. The SAR-based shoreline extraction module includes advanced extra-denoising filtering and an outlier detection module, but it maintains the core methodology used in the first phase, demonstrating the flexibility of the developed approach. The results demonstrate unprecedented accuracy and stability for gravel beaches (approximately 6 m of MAD), and for sandy beaches (approximately 7 m of MAD), starting from a 10 m pixel size. In a comparative assessment between the two TerraSAR-X images available and the closest Sentinel-1 in terms of time, the first achieves higher accuracy in terms of MAD (2.5 m) compared to the second one (6.5 m), but only in the image with good meteorological conditions (no differences are obtained in the other date). The study also investigates the influence of radar parameters, such as polarization, wavelength, acquisition geometry, and environmental conditions, on SAR shoreline accuracy. The results show that higher land–sea backscatter contrast, typical of gravel beaches, and moderate wind conditions enhance detection reliability. Conversely, high wind and wave activity reduce contrast and increase errors.Overall, this work offers robust, scalable tools for high-resolution coastal morphology monitoring in Mediterranean beaches using satellite data. These methodologies could be extended to beaches with significant tides and other related fields such as flood mapping. The developed algorithms could be incorporated into operational platforms like early warning systems or interactive WebGIS applications, potentially significantly contributing to local authorities' adaptive coastal zone management.
  

DOCTORAL DEGREE IN MECHANICAL, FLUIDS AND AEROSPACE ENGINEERING

• MASSAROTTI, GIORGIO PAOLO: New Dual Steering System in a Compact Tractor
  
  Author: MASSAROTTI, GIORGIO PAOLO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN MECHANICAL, FLUIDS AND AEROSPACE ENGINEERING
  Department: Department of Mechanical Engineering (EM)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 28/11/2025
  Reading time: 10:00
  Reading place: Sala de Juntes, Ed. TR5, ESEIAAT
  Thesis director: GAMEZ MONTERO, PEDRO JAVIER | CODINA MACIA, ESTEBAN
  Thesis abstract: In order to achieve optimal controllability in a dual-steering tractor (a four-wheel, iso-diametric tractor equipped with a dual-hydraulic steering system), this thesis proposes a coordinated approach that combines experimental testing (using a special agricultural tractor) with numerical analysis of the entire vehicle, developed in Bond Graph-3D. After an exhaustive review of the scientific literature, it is observed that the compact tractor with dual steering, has not yet been thoroughly analysed. In this thesis, in chapter 1 it is possible to identify the reasons that led to the realization of this long work and the objectives that were set at the beginning. These objectives were born from the understanding of the state of the art relating to double steering in the off-road sector, focusing particularly on the case of a vineyard tractor. All starting from the basics, from the steering which occurs smoothly and through the variants that can be found on the market today. In light of the machine construction information, the model of the studied tractor was introduced, searching the literature for the methods and models that could describe its dynamic behavior. In order to detail the description, the hydraulic circuit chosen based on the requirements listed in chapter 4 was introduced and an analysis to its modeling combined with the dynamic model of the tractor using Bond-Graph was provided. At the same time, experimental tests were carried out with a prototype tractor which incorporated the hydraulic circuitpreviously described, together with the dynamic model of the tractor, also obtained through modeling from the physical machine. The numerical analysis provided results that match very well with the experimental data, providing the key to the "salient" factors that characterize the tractor's steering capacity. A threshold can be set, relative to the vehicle speed, to disable dual-steer mode when a certain speed is exceeded. Based on experimental data, this threshold is set around 8.5 km/h. In conclusion, there are possibilities for future development that would lead the system described to a new circuit capable of appreciating not only the factors that determine the drivability of the tractor, but also of managing possible dangerous conditions for the user.
  

DOCTORAL DEGREE IN NETWORK ENGINEERING

• BAZÁN GUILLÉN, ALBERTO: Contribution to smart charging for electric vehicles in urban environments
  
  Author: BAZÁN GUILLÉN, ALBERTO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN NETWORK ENGINEERING
  Department: Department of Network Engineering (ENTEL)
  Mode: Normal
  Deposit date: 28/10/2025
  Reading date: 25/11/2025
  Reading time: 15:00
  Reading place: Aula C3-304 Aula Seminari, campus nord
  Thesis director: AGUILAR IGARTUA, MONICA | BARBECHO BAUTISTA, PABLO ANDRES
  Thesis abstract: The transition toward sustainable urban mobility and the urgent need to reduce greenhouse gas emissions have positioned electric vehicles (EVs) as a key element in the transformation toward a cleaner, more efficient, and smarter transportation system. The progressive increase in the number of EVs makes it essential to develop a robust and flexible charging infrastructure capable of meeting the growing energy demand while minimizing the impact on the power grid. In this context, coordinated charging planning becomes essential to optimize both user comfort and the efficiency of the electric and urban systems, particularly in dense environments and Mobility Hubs.This thesis addresses two fundamental and complementary challenges within the field of sustainable urban mobility: realistic traffic generation and optimal EV charging scheduling. To support decision-making in urban planning, DesRUTGe (Decentralized Realistic Urban Traffic Generator) has been developed — a new simulation framework that integrates Deep Reinforcement Learning (DRL) techniques with the SUMO simulator to generate high-fidelity, time-varying traffic profiles over 24 hours. Its main contribution lies in the incorporation of Decentralized Federated Learning (DFL), where each traffic detector and its associated area act as autonomous nodes that train local models with minimal historical data and exchange knowledge with nearby nodes. This strategy produces more accurate and realistic traffic patterns than those generated by centralized methods or conventional tools such as RouteSampler, achieving a better representation of daily variations and congestion peaks.Building upon this realistic simulation environment, the thesis proposes an intelligent EV charging scheduling framework designed for urban Mobility Hubs. The scheduler considers key factors such as time-varying electricity prices, vehicle charging time windows, initial and target State of Charge (SoC), and the possibility of bidirectional operation (vehicle-to-grid) when applicable. The work analyzes and compares two main approaches: Mixed-Integer Linear Programming (MILP), which provides optimal solutions in small-scale scenarios but has limited scalability, and Reinforcement Learning (RL) methods based on Proximal Policy Optimization (PPO), which demonstrate robust, adaptive, and efficient performance in more complex and dynamic environments. Additionally, reverse charging strategies are explored, allowing EVs to return energy to the grid during peak demand periods, thereby generating economic incentives for drivers and improving grid stability.Overall, this research makes three main contributions: (i) It introduces a decentralized and highly realistic simulation platform for generating urban traffic profiles; (ii) it develops a scalable and adaptive framework for optimal electric vehicle charging management, balancing user comfort, grid stability, and environmental objectives; and (iii) it demonstrates how both methodologies can be integrated into smart city planning, enabling coordinated design of sustainable mobility services and more efficient energy management. The results, validated with real-world data from the city of Barcelona, demonstrate the feasibility and potential impact of the proposed solutions for developing more sustainable and resilient urban mobility systems.
  
• PALOMARES TORRECILLA, JAVIER: Enabling collaborative Intelligence in Heterogeneous Edge to Cloud Continuum.
  
  Author: PALOMARES TORRECILLA, JAVIER
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN NETWORK ENGINEERING
  Department: Department of Network Engineering (ENTEL)
  Mode: Normal
  Deposit date: 21/10/2025
  Reading date: 25/11/2025
  Reading time: 12:00
  Reading place: Aula C4 - 028 -2 de l'EETAC de Castelldefels
  Thesis director: CERVELLO PASTOR, CRISTINA | CORONADO CALERO, ESTEFANÍA | SIDDIQUI, MUHAMMAD SHUAIB
  Thesis abstract: The growing convergence of the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and Multi-access Edge Computing (MEC) is reshaping service orchestration across the Edge-to-Cloud Continuum. These paradigms enable intelligent and flexible applications in heterogeneous, resource-constrained environments. A representative use case is multi-Automated Guided Vehicle (AGV) systems in smart factories, where moving agents must make time-sensitive decisions while adapting to variable connectivity, shifting computational loads, and coordinating with remote control logic. These scenarios demand orchestration mechanisms that dynamically integrate mobility, computation, and communication across infrastructure tiers. However, such systems exhibit highly dynamic behavior, driven by factors like device mobility and fluctuating resource availability, which creates significant challenges for scalable, fair, and autonomous orchestration under strict service guarantees.To address these demands, this thesis proposes a novel architecture unifying Information Technology and Operational Technology, extending MEC orchestration to the factory floor. It introduces an enhanced MEC orchestrator (MEO) and extended service descriptors to capture hardware-aware, location-sensitive, and non-computational constraints, enabling precise, constraint-driven placement across the Continuum. The approach is validated through a multi-AGV coordination use case.To optimize service placement, two orchestration mechanisms based on Deep Reinforcement Learning (DRL) are introduced. The first, a DRL-based Multi-Task Scheduling (DRL-MTS) strategy, minimizes end-to-end delay by distributing tasks across nodes while respecting resource availability. The second, the Intelligent Placement Algorithm (IPA), enables hardware-aware deployment by accounting for resource granularity and spatial constraints. Both integrate context-aware scheduling for adaptive and efficient placement.For decentralized coordination, this thesis proposes multi-agent learning strategies. The Multi-Agent Collaborative Protocol for Dynamic Resource Allocation (MACP-DRA) uses DRL to ensure fair and efficient placement under competition and resource constraints. The Multi-Agent Dynamic Bandwidth Environment (MADBE) manages bandwidth across heterogeneous service tiers with diverse communication demands, enabling real-time adaptation through policy specialization and aggregation. To further reduce bottlenecks in collaborative training, an Explicit Congestion Notification–based gradient compression mechanism for federated learning lowers bandwidth overhead while preserving accuracy. Collectively, these contributions enhance adaptability, communication efficiency, and fairness, enabling scalable and robust coordination across the Continuum.This thesis employs system modeling, algorithm design, and simulation-driven evaluation in realistic industrial scenarios with dynamic workloads, agent competition, and variable networks. The framework outperforms state-of-the-art baselines in latency, bandwidth, fairness, and SLA compliance. Its modular architecture and enriched descriptors enable constraint-aware deployment at the industrial edge, with the MEO supporting federated and cross-domain coordination. DRL mechanisms improve delay-sensitive scheduling, while multi-agent strategies ensure fairness and resilience. Validated through large-scale simulations and an emulated multi-AGV testbed, the approach achieves higher placement success, lower latency, better resource use, and reduced conflicts, establishing a foundation for research on distributed decision-making, cross-domain orchestration, and AI-driven automation.
  
• TORRES PÉREZ, CLAUDIA: Energy-Aware Service Placement Strategies in Dynamic Edge Environments
  
  Author: TORRES PÉREZ, CLAUDIA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN NETWORK ENGINEERING
  Department: Department of Network Engineering (ENTEL)
  Mode: Normal
  Deposit date: 28/10/2025
  Reading date: 04/12/2025
  Reading time: 12:00
  Reading place: sala C4-028-2 de l'EETAC a Castelldefels.
  Thesis director: CERVELLO PASTOR, CRISTINA | CORONADO CALERO, ESTEFANÍA | SIDDIQUI, MUHAMMAD SHUAIB
  Thesis abstract: The evolution of Beyond 5G (B5G) networks is transforming mobile communications, enabling interconnected environments and the proliferation of latency-sensitive services, such as real-time Internet of Things (IoT) analytics, immersive Extended Reality (XR) experiences, and generative Artificial Intelligence (AI) as a service. At the forefront of this transformation, edge computing is emerging as a pivotal paradigm extending computation resources closer to end users. Within this spectrum, Multi-Access Edge Computing (MEC) plays a central role by providing standardized, cloud-like capabilities at the edge of access networks. However, the inherent dynamism and distributed nature of MEC, especially in extreme-edge environments, adds significant complexity. The dynamism stems from variable service requirements such as fluctuating workloads, coupled with the dynamic infrastructure that includes heterogeneous nodes, variable connectivity, bandwidth requirements, and mobility across the entire Edge-to-Cloud continuum. Consequently, service placement in these dynamic settings faces substantial challenges, demanding adaptive and context-aware strategies to achieve system efficiency. Within this landscape, the growing energy demands of distributed edge nodes emerge as a paramount concern, contributing to significant carbon emissions, undermining global sustainability efforts, but also driving up operational costs, potentially rendering large-scale edge deployments financially unsustainable. Additionally, many edge devices rely on limited power sources, making energy efficiency essential for extending their operational lifetime and ensuring system reliability. As MEC scales geographically, the cumulative energy cost becomes a critical bottleneck for widespread adoption and sustainable growth. Under these volatile and resource-intensive conditions, minimizing energy consumption and optimizing resource utilization becomes critical. Thus, effective placement strategies must address dynamic constraints and ensure long-term sustainability. This thesis proposes intelligent, adaptive, and energy-efficient service placement mechanisms in distributed and extreme-edge environments. Firstly, it introduces an AI-based novel distributed orchestration technique within Distributed Multi-MEC Systems (DMMS). The technique, named Distributed Deep Reinforcement Learning-based Service Placement Availability-Aware Algorithm (DDRL-SP3A), aims to efficiently implement services in a system coordinated by multiple orchestrators, thereby optimizing resource usage by minimizing the number of active nodes. Secondly, the thesis presents an AI-based energy-aware strategy for heterogeneous MEC infrastructures introduced as DDRL-based Energy-Aware Service Placement Algorithm (DDRL-EASPA). The aim is to reduce the number of active nodes under heterogeneous infrastructures and dynamic service demands. Thirdly, this effort introduces an Energy Minimization Service Placement Algorithm (EMSPA), an adaptive, heuristic-based placement method to minimize energy consumption in smart factory extreme-edge environments, characterized by high mobility and severe resource and connectivity constraints. The proposed solutions achieve a near-optimal, efficient performance in low-latency MEC scenarios while meeting service and infrastructure constraints, optimizing resource utilization, and minimizing energy consumption. These solutions are evaluated through simulations of distributed networks with numerous hosting devices, orchestration entities, and service workloads. Additionally, a series of evaluations are conducted in a real-world testbed, demonstrating the differences in service placement performance compared to simulation. Overall, the strategies proposed in this thesis provide a robust and applicable framework for sustainable and high-performing edge computing.
  

DOCTORAL DEGREE IN PHOTONICS

• BESLIJA, FARUK: Hybrid diffuse optical monitoring and imaging: New approaches and applications in muscle and brain
  
  Author: BESLIJA, FARUK
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN PHOTONICS
  Department: Institute of Photonic Sciences (ICFO)
  Mode: Normal
  Deposit date: 28/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: DURDURAN, TURGUT | FERRER URIS, BLAI
  Thesis abstract: The generation of energy in the human body relies on oxygen metabolism, determined by oxygen delivery through blood flow and extraction at the tissue level. Reliable assessment of these parameters is crucial for understanding physiological function and tissue adaptations under various stimuli. Conventional monitoring tools for blood flow and oxygen saturation face trade-offs between cost, portability, and technical limitations (depth, resolution, dynamics), restricting their real-time deep-tissue use.This thesis advances diffuse optics, a non-invasive, safe, scalable approach exploiting light diffusion in scattering media, and introduces methodological and instrumental innovations for monitoring blood flow and oxygenation in adult skeletal muscle and brain—two of the most oxygen-demanding organs.Part I investigated long-term physiological adaptations in forearm muscles of advanced rock climbers versus healthy controls. Rock climbing requires exceptional grip endurance, making it an ideal model for localized neuromuscular and hemodynamic adaptations to chronic training. Two protocols were applied: (1) a resting vascular occlusion test (VOT) combining near-infrared spectroscopy (NIRS, oxygenation) and diffuse correlation spectroscopy (DCS, blood flow), and (2) an intermittent grip endurance test measuring force, NIRS, and electromyography (EMG). Results showed climbers had faster blood flow recovery and higher hemoglobin concentrations after occlusion, indicating enhanced vascular response. During exercise, they maintained force longer and used oxygen more efficiently. However, steady-state measures revealed no significant inter-group differences, suggesting adaptations are demand-driven rather than evident at rest. This study is novel in (1) applying DCS to climbing physiology and (2) integrating mechanical, neuromuscular, and hemodynamic measures in one framework.Part II focused on high-density (HD) cerebral blood flow (CBF) mapping, a key marker of brain metabolism. Current systems are bulky, costly, and clinical-only. We developed a new diffuse optics platform using speckle contrast optical spectroscopy (SCOS) and its tomographic extension (SCOT), leveraging cost-effective CMOS technology to improve signal-to-noise ratio (SNR) and scalability while retaining cortical sensitivity. A fiber-based prototype validated signal quality and flow sensitivity in forearm and forehead tests. Building on this, we designed a full-scale HD-SCOT system, nearing completion, intended for real-time, non-invasive mapping of CBF over large cortical areas (e.g., visual cortex).Final contribution: a proof-of-concept SCOS extension enabling simultaneous blood flow and oxygenation measurement. Using multiple wavelengths, source-detector separations, and exposure times, it offers a simplified alternative to dual NIRS-DCS systems. Preliminary forearm tests confirmed feasibility, suggesting applications in muscle and brain monitoring.In summary, this thesis advances diffuse optical monitoring by developing new instruments and methodologies for deep-tissue hemodynamics. Applications in sport physiology and neuroimaging highlight the potential of multi-modal, high-density optical systems to deepen understanding of oxygen metabolism in naturalistic, real-time contexts, paving the way for broader physiological and clinical applications.
  
• MORALES CURIEL, LUIS FELIPE: Deep-learning enhanced bioluminescence microscopy
  
  Author: MORALES CURIEL, LUIS FELIPE
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN PHOTONICS
  Department: Institute of Photonic Sciences (ICFO)
  Mode: Normal
  Deposit date: 16/10/2025
  Reading date: 21/11/2025
  Reading time: 10:00
  Reading place: ICFO Auditorium
  Thesis director: KRIEG, MICHAEL
  Thesis abstract: Bioluminescence microscopy presents a powerful alternative to fluorescence imaging by eliminating the need for external illumination, thereby avoiding issues such as phototoxicity, photobleaching, and background autofluorescence. However, the inherently low photon output of luciferase-based reporters significantly restricts the signal-to-noise ratio (SNR), as well as the achievable spatial and temporal resolution—challenges that are especially pronounced in dynamic or volumetric biological imaging. This thesis addresses these limitations by introducing a deep learning-driven imaging pipeline designed to enhance bioluminescence microscopy at both the data acquisition and image reconstruction stages.Our strategy integrates optical system design with advanced neural networks to enable rapid, high-resolution 3D imaging under extremely low-light conditions. We engineered a custom microscope featuring a highly compact optical axis and paired it with a single-photon sensitive camera, significantly boosting the SNR of bioluminescent images. To achieve fast volumetric imaging, we incorporated light field microscopy (LFM) and Fourier light field microscopy (FLFM), enabling single-shot 3D acquisition while improving axial and lateral resolution via Fourier-domain filtering. The primary objective of this work is to demonstrate how deep learning can substantially enhance bioluminescence microscopy, pushing the technique beyond its traditional limits in both 2D and 3D imaging.At the core of our approach is a suite of convolutional neural networks specifically trained on bioluminescent data. Using both synthetic and experimental datasets, we designed and trained models capable of extracting meaningful information from low-SNR raw data, recovering otherwise lost details and offering deeper insight into the biological sample. The models developed in this thesis cover key tasks such as denoising and reconstruction of wide-field, light field, and Fourier light field bioluminescent images. Together, they form a modular, learnable pipeline that significantly elevates the performance of bioluminescence microscopy in terms of both quality and speed.We validate our system using live biological samples, including Caenorhabditis elegans, mouse stem cells, and zebrafish embryos, capturing neuronal activity and intracellular dynamics at subsecond timescales. By placing deep learning at the heart of the imaging process, this work establishes a new paradigm for bioluminescence microscopy, transforming a traditionally low-SNR modality into a robust tool for fast, high-resolution, and label-specific imaging in living organisms.
  

DOCTORAL DEGREE IN POLYMERS AND BIOPOLYMERS

• CASADO GÓMEZ, JAIME: 3D Printable Hybrid Acrylate-Epoxy Vitrimer Resins with Improved Compatibility and Reprocessability
  
  Author: CASADO GÓMEZ, JAIME
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN POLYMERS AND BIOPOLYMERS
  Department: Department of Chemical Engineering (EQ)
  Mode: Article-based thesis
  Deposit date: 21/10/2025
  Reading date: 12/12/2025
  Reading time: 11:00
  Reading place: Escola Tècnica Superior d'Enginyeria Industrial de Barcelona (ETSEIB)Campus Diagonal SudAv. Diagonal, 647 08028Edifici I, Planta 1Aula 28.8
  Thesis director: FERNANDEZ FRANCOS, XAVIER | KONURAY, ALI OSMAN
  Thesis abstract: The Covalent Adaptable Networks (CANs) made of polymeric materials that use dynamic covalent chemistry, allowing bonds to break and reform when stimulated, combine the mechanical properties of thermoset polymers with the ability to be reprocessed and recycled. The integration of 3D printing technology with CANs represents a significant advancement in the field of manufacturing polymer components. This innovative process offers the functional benefits of a thermoset along with the recycling advantages of a thermoplastic, making it a highly sustainable solution.In the following collection of articles, a group of novel dual-curing thermosetting materials have been designed, optimised and improved regarding their compatibility and reprocessability. In the first article, we successfully crafted four genuine resins and explored how their behaviour and properties were influenced by the unique combinations and proportions of their formulation ingredients. The original dual-curing system was performed by means of homogenously mixing an epoxy resin with a di-acrylate monomer rich in β-hydroxy ester and hydroxyls, a dicarboxylic acid and a coupling agent in a fixed proportion. The use of different transesterification catalysts in varying proportions, a methacrylate monomer and a photoinitiator round off the formulation. The combination of these chemicals results in the formation of a hybrid network, which is capable of undergoing transesterifications reactions. The 3D-printed and fully-cured parts from these four innovative resins have proven that their thermo-mechanical properties are in line with the designed specifications. Their repair and recycle capabilities are facilitated by a CAN structure.In the second article, we have optimised the formulations of 3D printable vitrimer resins with the objective of enhancing their processing, mechanical properties, and repairability/reprocessability. An improvement of the formulation was achieved through the determination of the optimal quantities of acrylates and coupling agent. A selection of epoxy resins was also made with the aim of identifying the best performing option. The resins developed in this part of the research have offered a more suitable viscosity for handling in the 3D printer. It has been demonstrated as well that parts printed from these improved resins and subsequently double-cured have shown an enhancement in their thermo-mechanical behaviour.In the third article, we have advanced our research in two key areas. Firstly, we have taken a further step in the facilitation of the Vitrimer formulation elaboration by improving the mixability of the chemical compounds. This improvement involved replacing a powder carboxylic acid with a taylor-made liquid coupling agent.Secondly, an evaluation of the thermo-mechanical behaviour of the fully cured resin was carried out, depending on the sequence of thermal and UV curing stages. The materials developed in this study have demonstrated efficacy in the effective relaxation of internal stresses, attributable to the high dynamic β-hydroxyester groups content. Consequently, processes such as reshaping, repairing, or complete recycling are enabled. Furthermore, modifications made to the resin formulations enabled the production of thermosets with customised mechanical properties. All these properties offer new possibilities for the production of parts using techniques such as 3D printing and thermal post-curing, providing a viable, sustainable and more convenient alternative for the thermosetting materials industry.
  

DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS

• MUTHINENI, KARTHIK: Wireless Infrastructure-Based Indoor Positioning in Controlled Industrial Environments
  
  Author: MUTHINENI, KARTHIK
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 16/10/2025
  Reading date: 20/11/2025
  Reading time: 16:00
  Reading place: Aula Teleensenyament, Edifici B3 - Ricardo Valle Sala 103 Planta 1
  Thesis director: VIDAL MANZANO, JOSE | ARTEMENKO, ALEXANDER | NAJAR MARTON, MONTSERRAT
  Thesis abstract: Wireless communications have become the central nervous system of the Factory of the Future (FoF). According to this, wireless infrastructure in industries serves the dual purposeof providing connectivity and positioning industrial assets such as Automated Guided Vehicles (AGVs). The Non-Line-of-Sight (NLoS) and multipath-dominant environments, such asthose found in industries, hinder wireless signal propagation, leading to inaccuracies in wireless infrastructure-based positioning. While advances have been made in developing approaches to enhance wireless positioning accuracy in complex indoor scenarios, they may not be readily applicable to industrial environments without or with minimal modification. This motivates the study towards developing accurate and precise wireless infrastructure-based positioning approaches tailored explicitly for industrial settings. This PhD thesis aims to address key challenges in wireless positioning for industrial environments and develop algorithms to enhance the accuracy of wireless infrastructure-based positioning by proposing both model-driven and data-driven approaches.This PhD thesis presents simulation analysis and practical experiments to validate the efficiency of the proposed approaches using a range of different wireless infrastructures, including Fifth Generation (5G), Ultra-Wideband (UWB), and Sixth Generation (6G) mobile communication, including Integrated Sensing and Communication (ISAC). This work begins by analyzing the achievable 5G-based positioning accuracy in C-band and Millimeter-Wave (mmWave) bands within a specific model of a dense, cluttered industrial environment, utilizing high-fidelity ray tracing simulations to capture the impact of NLoS and multipath propagation. The findings highlight limitations of 5G positioning and, in general, wireless positioning in complex and cluttered industrial settings. To overcome these limitations, this work advances the State of the Art (SotA) by developing novel enhancement approaches based on sensor and/or data fusion for UWB-based and ISAC-assisted positioning systems. These approaches leverage the power of Deep Neural Networks (DNNs), Long Short-Term Memory (LSTM) networks, and Graph Neural Networks (GNNs) to model spatial and temporal relationships more effectively, yielding substantial improvements in positioning accuracy, robustness, and adaptability to dynamic industrial scenarios. Furthermore, the application of the proposed approaches is illustrated through an Automated Guided Vehicle (AGV) use case. This PhD thesis lays a strong foundation for advancing future research and real-world applications, offering valuable insights that can shape the next generation of industrial wireless positioning system design and deployment.
  
• RAFIEIAN, BARDIA: Enhancing Word and Document Embeddings for Natural Language Processing Tasks
  
  Author: RAFIEIAN, BARDIA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 02/10/2025
  Reading date: 21/11/2025
  Reading time: 11:00
  Reading place: Sala d'actes de la FIB Manuel Martí Recober
  Thesis director: VÁZQUEZ ALCOCER, PERE PAU
  Thesis abstract: This thesis delves into various aspects of natural language processing, focusing on domain-specific neural machine translation, specialized word embeddings, data augmentation, recommender systems, document embedding techniques, and hierarchical classification with large language models. The work is structured around a couple of key contributions.Chapter 4 introduces a novel data preparation and tokenization method (Hybrid-BTS) for biomedical neural machine translation, alongside a post-specialization technique leveraging Wasserstein GANs to enhance word embeddings with multilingual constraints. Chapter 5 proposes a domain adaptation strategy for biomedical translation involving forward translation, BPE optimization, and term frequency manipulation. It also details the development of two fashion e-commerce recommender systems: one content-based and one collaborative, both integrating practical style rules. Finally, Chapter 6 presents an evaluation of document embedding models (Doc2vec, SciBERT, Longformer, LLaMA-3, GEMMA-2B) for long document classification, and a modular pipeline designed for multi-label hierarchical patent classification using transformer-based models, optimized for efficiency with LoRA and quantization. Collectively, these contributions push the state-of-the-art in both applied and theoretical NLP by providing new methods to boost performance, adaptability, and efficiency in domain-specific and large-scale applications.
  
• TRULLENQUE ORTIZ, MARTÍN: A Cross-Layer Perspective on Radio Resource Management and User Quality of Experience in V2X-Enabled 5G Networks.
  
  Author: TRULLENQUE ORTIZ, MARTÍN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 31/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: SALLENT ROIG, JOSE ORIOL | CAMPS MUR, DANIEL
  Thesis abstract: Vehicular communications are expected to be among the main verticals benefiting from the massive rollout of 5G. The millisecond latencies enabled by 5G, combined with network slicing and the global coverage offered by public cellular networks, leverage the conditions to adopt public cellular networks for vehicle-to-everything (V2X) communications. This vertical will introduce new services that generate additional traffic demands on existing radio infrastructure, posing distinct challenges for mobile network operators. From a service perspective, V2X services require both low latency and high reliability. For example, under a periodic exchange position information between vehicles, a single dropped packet can increase the position error estimate and jeopardize road safety. From a network perspective, vehicular services will demand resources differently from generic Internet users, often creating high user densities along highways, where user density is typically [MT1.1]low and radio deployment is not dense. Moreover, vehicle mobility can lead to long periods during which cells serve very few users, followed by periods of high demand that may compromise resource availability, especially during traffic congestion. To this end, this thesis investigates solutions for cell overloads caused by traffic jams when V2X services run over public cellular networks.The first contribution of this work is an in-depth analysis of vehicular mobility based on a dataset of realistic traces, aimed at understanding how traffic congestion drains radio resources. Studying vehicle flows across cells reveals user concentrations at specific road locations, providing opportunities for congestion control mechanisms. The second contribution examines the repeatability of high-load situations in a real 5G deployment. Measurements over a 5G NSA network show that congestion exhibits consistent daily patterns, duration, and load characteristics, motivating deterministic congestion control strategies.Building on these insights, this thesis proposes and implements two complementary approaches: a network-layer mobility load balancing algorithm that diverts vehicles at cell edges to neighbouring cells during traffic jams, and an application-layer mechanism that manages service degradation to control packet loss under overload. Combining both approaches, this thesis has conceived, designed, implemented and evaluated FOM-5G, a framework capable to derive the best congestion control strategy for each cell overload. The framework leverages the use of historical information data to optimize cell congestion responses through experience. Beyond its design and evaluation, the thesis also proposes an architectural integration of FOM-5G within the O-RAN architecture and network exposure API defined in the 5G core. Finally, the framework is validated through simulations using diverse vehicular traces that capture urban congestion scenarios. Results confirm FOM-5G’s ability to distinguish between different congestion events and select the optimal mitigation strategy, demonstrating its effectiveness in ensuring reliable V2X communications over public cellular networks.
  

DOCTORAL DEGREE IN SUSTAINABILITY

• URIOSTE DAZA, SERGIO ALEJANDRO: Advancing Reform of European Union Plant Variety Registration: Institutional, Empirical, and Policy Insights for Sustainable Agri-Food Governance
  
  Author: URIOSTE DAZA, SERGIO ALEJANDRO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN SUSTAINABILITY
  Department: University Research Institute for Sustainability Science and Technology (IS.UPC)
  Mode: Normal
  Deposit date: 29/09/2025
  Reading date: 12/12/2025
  Reading time: 12:00
  Reading place: Sala de Grados de la ESAB
  Thesis director: TAGHOUTI, IBTISSEM | GIL ROIG, JOSE MARIA
  Thesis abstract: Plant variety registration is a critical regulatory gatekeeper between the breeding of improved varieties and their farm-level adoption. In the European Union, however, this system is being outpaced by technological advances and sustainability challenges. Legislative reform is now underway to address these shortcomings, aiming to improve the system’s efficiency by integrating new technologies and sustainability criteria into variety testing and fostering greater harmonisation across Member States. Although these reforms are broadly welcomed, diverging positions among stakeholder groups and EU institutions remain unresolved.Bridging these differences will require robust evidence to inform the ongoing negotiations. This thesis responds to this demand by providing an evidence-based assessment of the current system’s inefficiencies and by proposing realistic reform pathways to help reconcile core tensions between regulatory drag and productivity, divergent stakeholder interests, and the gap between policy goals and farm-level realities. To achieve these objectives, this research presents an integrated framework that engages all key actors in variety testing and combines econometric analysis, decision-analytic modelling, and qualitative analysis.Using a large panel dataset on crop registration and productivity, a fixed-effects analysis provides evidence of regulatory drag on productivity gains, particularly for crops subject to Value for Cultivation and Use testing. Evidence gathered from stakeholders explores into the factors behind these regulatory delays and identifies pathways to overcome systemic inefficiencies, including the uptake of enabling technologies and the harmonisation of testing processes.Subsequently, an analysis of contested policy alternatives is conducted using a replicable framework that integrates expert judgment with public input through multi-criteria decision methods and complementary weighting techniques. The results reveal a clear consensus on prioritising the adoption of technological advancements to improve the system's efficiency and accuracy. However, the analysis also exposes disagreements over efforts to harmonise the system and include sustainability criteria in testing procedures, revealing significant heterogeneity among stakeholder groups.To further investigate the contested issue of adding sustainability criteria in variety testing, a farm-level study is presented to elicit the preferences of apple growers in Spain. Using a Discrete Choice Experiment, farmers' preferences for sustainability traits in new apple varieties were elicited and examined in relation to risk behaviours. The results show a positive but heterogeneous demand for sustainability-related traits, with willingness-to-pay shaped by farm characteristics rather than by measured risk attitudes.Together, these findings demonstrate how data-driven and stakeholder-informed reforms can reduce institutional friction by establishing common ground for negotiation on key aspects of the regulation. Both stakeholder priorities and farmers’ demands point toward the need to prioritise technological uptake and design mechanisms that facilitate the delivery of climate-resilient and resource-efficient varieties. Effective stakeholder involvement and continuous evidence generation are essential for the regulatory path forward. By integrating evidence across institutional, technological, and behavioural layers, this research advances the goals of the European Green Deal and the Farm to Fork Strategy and provides a transferable framework for future assessments of agricultural policy and innovation.
  

DOCTORAL DEGREE IN THEORY AND HISTORY OF ARCHITECTURE

• PEDRAGOSA BATLLORI, GEMMA: Santa Coloma d'Andorra: el projecte d'una església a l'Andorra d'abans del S.XI.
  
  Author: PEDRAGOSA BATLLORI, GEMMA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN THEORY AND HISTORY OF ARCHITECTURE
  Department: Department of History and Theory of Architecture and Communication Techniques (THATC)
  Mode: Normal
  Deposit date: 01/09/2025
  Reading date: 21/11/2025
  Reading time: 10:00
  Reading place: ETSAV. Campus de Sant Cugat del Vallès Carrer Pere Serra, 1-15 - 08173 Sant Cugat del Vallès Aula: Seminari 2
  Thesis director: GRANELL TRIAS, ENRIQUE | GINER OLCINA, JOSEP
  Thesis abstract: The church of Santa Coloma d’Andorra belongs to one of the simplest and oldest architectural types of religious architecture: that of a single rectangular nave with a square apse. However, the simplicity of this type should not necessarily be associated with a straightforward or immediate construction or design.The aim of this study is to determine the extent to which the architecture of Santa Coloma follows a complex metrical design, which could only be achieved within a cultural context that, in Santa Coloma — located near two major cultural centres of the time, the Cathedral of La Seu d’Urgell and the Monastery of Sant Serni de Tavèrnoles — is highly plausible.In this work, architecture is used as archaeological material to analyse the key elements of the building’s architectural composition. Historiography has been reviewed, plans have been drawn up, the unit of measurement has been identified, and its dimensions studied in relation to the knowledge of proportion of the period and descriptions of biblical buildings. And it turns out that in order to conceive, design and build an apparently simple church like this, it was necessary to be familiar with the architecture represented in the Bible and with the arithmetical elaborations compiled by Boethius and Cassiodorus in the 6th century.We will therefore see a building which, although rural and seemingly modest, is the result of a layout and proportions based on a specific symbolic language, reflecting and documenting a body of knowledge and a way of applying it.
  

DOCTORAL DEGREE IN TRANSPORT ENGINEERING AND INFRASTRUCTURE

• TEJEDOR FUENTETAJA, JOSÉ: Análisis en la afección de infraestructuras ferroviarias convencionales, con la incorporación de líneas de alta velocidad, durante la construcción y explotación.
  
  Author: TEJEDOR FUENTETAJA, JOSÉ
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN TRANSPORT ENGINEERING AND INFRASTRUCTURE
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 01/10/2025
  Reading date: 02/12/2025
  Reading time: 11:00
  Reading place: UPC Campus Nord, ETSECCPB, C/ Jordi Girona 1-3, edificio C1, Sala 002, Barcelona
  Thesis director: CAMPOS CACHEDA, JOSÉ MAGÍN | GRANDE ANDRADE, ZACARIAS
  Thesis abstract: This paper analyzes the impact on the conventional network caused by the construction of the new high-speed line in Catalonia, once it has entered into operation and its construction work has been completed, except for the Barcelona La Sagrera station, which will be under construction in July 2025.The implementation of high-speed rail has led to the remodeling and construction of new track sections and stations, a situation that is particularly pronounced around the city of Barcelona.To conduct this analysis, we must examine the conditions of the railway infrastructure, superstructure, and technical operating systems of the conventional network in 2000. Subsequently, a determination will be made as to how the construction of the new infrastructure may have affected rail operations on each of the conventional lines, in terms of the number of trains on each of their different sections, and the type of transport services provided on them (long-distance, medium-distance, commuter, and freight). This will determine whether the construction of the new line could have impacted rail operations.The results obtained will be used to determine the current status of the conventional network, in order to identify shortcomings and propose improvements that could lead to improved rail operations, as well as better management, and thus greater efficiency and safety. They will also be used to assist in decision-making regarding future rail infrastructure projects. Once the railway sector was liberalized, initially with freight transport services in 2007 and with the new entry of private companies starting in 2019, and in 2021, in addition to Renfe with its commercial AVE and AVLO products, OUIGO and Iryo also began providing high-speed passenger transport services on some corridors in Spain.However, the European Union's regulations require member states to incorporate competition not only in freight and passenger transport services, but also to extend this competition to medium-distance and suburban passenger transport services, and to be implemented by 2033 at the latest.Therefore, the analysis of the behavior of the railway infrastructure, as well as of the transport services, during the construction of the infrastructure works of the high-speed network in Catalonia, and with the actions that should be carried out previously on the conventional network, in order to offer a better quality of service to medium-distance and suburban passengers, once its liberalization is effective, this analysis can help to see the behavior before future actions on the infrastructure, not only in Catalonia, but at the national level, since liberalization will affect the entire State.
  

DOCTORAL DEGREE IN URBAN AND ARCHITECTURAL MANAGEMENT AND VALUATION

• CARRILLO CERVERA, ALEJANDRO: Efecto del área verde en la salud: parques y enfermedades cardiovasculares en Mérida (2020)
  
  Author: CARRILLO CERVERA, ALEJANDRO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN URBAN AND ARCHITECTURAL MANAGEMENT AND VALUATION
  Department: Department of Architectural Technology (TA)
  Mode: Normal
  Deposit date: 25/09/2025
  Reading date: 27/11/2025
  Reading time: 17:00
  Reading place: ETSAB (Esc. Téc. Sup. Arquitectura de Bcn)-Pl. Baja-Sala GradosEnlace videoconf.: https://meet.google.com/vzy-djqq-sqaInicio conexión 16:30 h
  Thesis director: ROCA CLADERA, JOSE NICASIO | ARELLANO RAMOS, BLANCA ESMARAGDA
  Thesis abstract: This doctoral study analyzes the relationship between urban green spaces and cardiovascular disease (CVD) mortality in the city of Mérida, Yucatán (Mexico), using a pioneering methodology at the urban block scale. The research combined spatial analysis, statistical techniques, and direct observation to assess the impact of green coverage, quality (NDVI), and accessibility to urban parks on public health.In Mérida, the analysis reveals that daily proximity to green areas—especially small or medium-sized parks within 300 meters—has a statistically significant effect (β = –0.001; p < 0.001). These parks exhibit the highest quality of urban greenery and the best plant health. In contrast, larger parks over 5 hectares, where vegetation quality and plant health are poor, show no significant effect (β = 0.000; p = 0.006). This suggests that it is not only the amount of permeable land in parks that matters, but also the quality of urban vegetation and accessibility in the immediate surroundings. By age group, young adults (18–24 years) had better access to parks, while children and older adults faced more barriers.People with disabilities are more vulnerable to developing CVD (β = 0.324; p = 0.010). The findings also indicate an increase in mortality associated with the use of private vehicles as a means of transportation (β = 0.261; p = 0.007). On the other hand, higher spending on food consumed at home has a protective effect (β = –0.165; p < 0.001).Although other factors may also explain changes in mortality rates, urban greenery and the quality of the built environment significantly influence population health. The regression equation was statistically significant F(5, 10036) = 0.001. The R² value was 0.202, indicating that 20.2% of the CVD mortality rate can be explained by the model, which includes variables such as accessibility to urban parks, the percentage of people with disabilities, spending on private vehicle use, and spending on food consumed at home.The mortality rate decreases by –0.001 points for every additional square meter of urban greenery in parks. Excluding medians and sidewalks, the population has access to an average of 7.17 m² of urban green space per inhabitant across all parks within the peripheral ring, and 5.79 m² per inhabitant when considering only parks of 5 hectares or more. The average CVD mortality rate is 29.78% per urban block.
  

DOCTORAL DEGREE IN URBANISM

• MARTÍ ELÍAS, JOAN MARIA: Hidrografies urbanes. Cicle de l’aigua, forma urbana i estructura territorial al la Vall Baixa i al Delta del Llobregat
  
  Author: MARTÍ ELÍAS, JOAN MARIA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN URBANISM
  Department: Department of Urbanism, Territory and Landscape (DUTP)
  Mode: Normal
  Deposit date: 31/10/2025
  Reading date: pending
  Reading time: pending
  Reading place: pending
  Thesis director: CROSAS ARMENGOL, CARLES
  Thesis abstract: The research dives in the relationship between urban morphology and water cycle from an ecosystemic perspective that integrates physical, territorial, and infrastructural components. The objective is to demonstrate that urban morphology is directly related with hydrological dynamics and that, therefore, it can become a key tool for redefining water management strategies in metropolitan contexts.This dissertation is grounded in the intensive exploration of a specific territory (la Vall Baixa and the Delta del Llobregat) through five successive approaches, addressing the territory from the metropolitan scale down to local urban fabrics. This process constructs a multi-scalar reading that relates the physical, infrastructural, and morphological conditions of the city to its hydrological dynamics.The territorial approach is enriched by a historical and conceptual dimension that reconstructs a genealogy of urban thought in an ecosystem perspective, based on figures and projects that conceived the city as a living system, dependent on its exchanges with the territory. The research also revisits nineteenth-century projects that, with a hygienist outlook, anticipated “proto-ecological” visions. In this regard, Garcia Fària’s proposal for the sanitation of the Eixample and the diversion of the river Llobregat constitutes a paradigmatic case that anticipates an integrated territorial interpretation, establishing a valuable precedent for contemporary urban planning practice.On this basis, the final stretch of the river Llobregat is studied, revealing the complexity of a hybrid morphology resulting from the superimposition of natural and anthropic logics. The river is presented as an ecological and productive infrastructure, a space where the persistence of hydraulic and agricultural traces coexists with metropolitan urbanization pressures. From here, the research turns to the underground dimension, analyzing the delta aquifer as a key element for environmental balance. Infiltration dynamics, alterations derived from urbanization, and opportunities for the reconversion of industrial and logistical spaces based on hydrogeological criteria are examined. The three-dimensional reading of the territory makes it possible to link surface and subsoil and to propose infiltration and regulation devices that restore natural functions.In the central part of the dissertation, these criteria are applied at an intermediate analytical scale, proposing the sub-basin as a functional unit of analysis and design, due to its capacity to integrate topographical continuities, road structures, and open spaces. On this basis, a parametric methodology is developed, articulated into six actions (capture, consumption, treatment, infiltration, retention, and reuse) applied to multiple metropolitan fabrics. A multi-scalar analysis makes it possible to identify hydromorphological indicators that guide design decisions according to context and urban structure. The result is an operative atlas that proposes ranges of intervention and adaptive criteria for urban transformation guided by hydrological principles.Finally, the research advances a paradigm shift: water becomes a design vector, not merely a technical constraint. Public space, the road network, and urban voids are recognized as multifunctional water devices, while the underground dimension is considered an active and inseparable layer of the urban system. This three-dimensional and complex vision relates metabolism and form, while articulating both regional scale and urban project.Overall, the dissertation contributes to creating a culture of urban design that integrates water as a vector of intelligibility and ordering, where the term “hidrografia” is not understood in the traditional sense of describing river courses, but rather as a critical instrument for mapping the presence, movement, and structuring capacity of water in all its forms.
  

DOCTORAL DEGREE IN ARTIFICIAL INTELLIGENCE

• FERRANDO MONSONÍS, JAVIER: Interpretability in Natural Language Processing and Machine Translation
  
  Author: FERRANDO MONSONÍS, JAVIER
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN ARTIFICIAL INTELLIGENCE
  Department: Department of Computer Science (CS)
  Mode: Normal
  Deposit date: 13/11/2025
  Deposit END date: 26/11/2025
  Thesis director: RUIZ COSTA-JUSSA, MARTA
  Thesis abstract: This thesis presents a set of methods and analyses designed to improve our understanding of the internal mechanisms of Transformer-based models in natural language processing and machine translation.This work first investigates the role of attention weights in encoder-decoder Transformers, showing that while they do not provide accurate word alignments, they nonetheless help explain model predictions and contribute to a deeper understanding of translation quality.A central contribution of the dissertation is the development of ALTI and its extensions, which offer a new approach to input attribution. These methods challenge prior assumptions about the explanatory power of attention mechanisms and reveal how information propagates between encoder and decoder components. In doing so, they also shed light on sources of hallucinations in translation systems.Further, the thesis introduces techniques to attribute predictions to individual components and positions, enabling contrastive explanations of linguistic behavior. These explanations clarify how language models represent and solve different linguistic phenomena.The dissertation also proposes a methodology for tracking information flow during inference, offering insight into how various components contribute to model predictions. This allows for the identification of domain-specialized components and a better understanding of how representations are transformed across layers.Lastly, the analysis of cross-lingual circuit similarities reveals shared structural patterns in how models handle different languages. These findings point to potential universal mechanisms in language models.Collectively, this thesis advances the interpretability of Transformer models by providing tools and frameworks for probing, attributing, and understanding the behavior of complex NLP systems.
  

DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION

• SHEIKHSAMAD, MOHAMMAD: Learning Methods in Planning and Control for Autonomous Vehicles and Robotic Manipulation
  
  Author: SHEIKHSAMAD, MOHAMMAD
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN AUTOMATIC CONTROL, ROBOTICS AND VISION
  Department: Institute of Industrial and Control Engineering (IOC)
  Mode: Normal
  Deposit date: 11/11/2025
  Deposit END date: 24/11/2025
  Thesis director: SUAREZ FEIJOO, RAUL | ROSELL GRATACOS, JOAN
  Thesis abstract: This thesis deals with the application of machine learning (ML) and deep learning (DL) techniques to enhance planning and control tasks across different fields, particularly on autonomous vehicles and robotic hands. Specifically, it addresses the learning-based development of a robust trajectory-tracking controller for autonomous vehicles, an adaptive path planner for robotic hands enabling dexterous manipulation, and a human-in-the-loop controller for myoelectric robotic hands to ensure precise grasp force regulation.In the field of autonomous vehicles, the thesis develops a Takagi–Sugeno (TS) controller using the Adaptive Neuro-Fuzzy Inference System (ANFIS) as a learning-based approach to infer a control strategy from input–output data of an existing controller. The closed-loop stability of the system is analyzed using Lyapunov theory and Linear Matrix Inequalities (LMIs). The proposed controller eliminates the need for online optimization, significantly reduces computational cost, and enhances real-time performance. Its effectiveness is validated through simulations on a small-scale autonomous vehicle.In the field of robotic dexterous manipulation, the thesis introduces three learning-based path planners using ANFIS and Deep Neural Networks (DNNs) to learn heuristics from an analytical planner and self-tune their parameters based on prior experience. This approach enables robots to manipulate objects of varying shapes, sizes, and material properties. The proposed planners are validated through real-world experiments using an Allegro robotic hand, demonstrating robustness against sensor noise and environmental disturbances.In the field of robotic grasping, the thesis presents a myocontrolled human-in-the-loop (HITL) system for precise grasp strength regulation. The system integrates both DNN-based and fuzzy-based force controllers. The fuzzy-based controller leverages fuzzy logic, with parameter optimization guided by user preferences collected through a graphical user interface (GUI) using Global Learning of Input–Output Strategies from Pairwise Preferences (GLISp). These controllers are compared against heuristic model-based controllers, and the system is validated through real-world experiments using the AR10 robotic hand, showing enhanced adaptability and fine-grained force regulation capabilities.The findings of this research contribute to the advancement of intelligent planning and control systems across multiple application areas, paving the way for more efficient, adaptive, and stable automation in real-world scenarios.
  

DOCTORAL DEGREE IN COMPUTATIONAL AND APPLIED PHYSICS

• MIRZAY SHAHIM, MAAHIN: Catalytic Properties of Amorphous Alloys
  
  Author: MIRZAY SHAHIM, MAAHIN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTATIONAL AND APPLIED PHYSICS
  Department: Department of Physics (FIS)
  Mode: Normal
  Deposit date: 17/11/2025
  Deposit END date: 28/11/2025
  Thesis director: PINEDA SOLER, ELOY | SOLER TURU, LLUIS
  Thesis abstract: This thesis explores the catalytic potential of metallic glasses (MGs) and their combination with cerium oxide (CeO₂) for low-temperature carbon monoxide (CO) oxidation and CO preferential oxidation (COPrOx) reactions. Metallic glasses, due to their non-crystalline structure and tunable composition, offer a promising platform for catalytic applications when appropriately engineered. The study focuses on three primary MG systems: Ce65Al35, Pd77Si16.5Cu6.5, and Cu48Zr48Al4, examining their structural characteristics, and catalytic behavior. The results showed that the Ce65Al35 metallic glass has limited catalytic activity, even after various activation treatments such as ball milling, calcination, or combination with CeO2. However, doping the binary Ce-Al system with Pd (Ce61Al35Pd4) markedly improved performance, achieving 100% CO conversion at 300°C when ball milled. Interestingly, mixing this ternary MG with CeO₂ did not provide further enhancement, indicating that Pd’s role is dominant and not synergistic with ceria. The Pd77Si16.5Cu6.5 MG emerged as the most effective standalone catalyst, delivering full CO conversion at only 240°C. Which could be attributed to Pd and its optimized distribution in the amorphous matrix. Control experiments with binary alloys (Pd77Si23 and Cu6Si94) highlighted the importance of both composition and structural processing, particularly the necessity of melt spinning and ball milling to generate active, fine-particle structures.Another major contribution of this work is the development and detailed characterization of Cu-based MG/CeO2 composites, especially Cu48Zr48Al4.These systems showed strong activity and stability in both CO and COPrOx reactions, with performance enhanced through ball milling. Structural and operando analyses (XPS, EXAFS, NEXAFS, and XRD) confirmed that the catalysts undergo surface rearrangement during reaction, stabilizing catalytically active Cu(I) atoms. A spontaneous aging phenomenon and a similar change under hydrogen pre-reduction pointed to the dynamic evolution of active sites during real operation conditions. This study demonstrates that mechanochemical synthesis and careful structural design of MG/CeO₂ composites enable the development of efficient, low-cost, and stable oxidation catalysts. These findings offer new strategies for creating highly active materials for pollution control and hydrogen purification technologies, opening the path to use amorphous metals for heterogeneous catalysis.
  

DOCTORAL DEGREE IN COMPUTER ARCHITECTURE

• BANCHELLI GRACIA, FABIO FRANCISCO: Evaluation and methods to increase efficiency of HPC systems with different maturity levels
  
  Author: BANCHELLI GRACIA, FABIO FRANCISCO
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 12/11/2025
  Deposit END date: 25/11/2025
  Thesis director: MANTOVANI, FILIPPO | GARCIA GASULLA, MARTA
  Thesis abstract: High-Performance Computing (HPC) has entered an era of increasing architectural diversity and complexity, with systems ranging from experimental prototypes to large-scale production machines. This evolution presents a fundamental challenge: how to consistently evaluate performance, scalability, and efficiency across platforms with varying levels of technological maturity. Traditional benchmarking methods, while effective for fully deployed systems, often fall short when applied to early-stage prototypes where software stacks are incomplete or hardware is still under development.This thesis proposes and develops a comprehensive evaluation methodology capable of addressing these challenges. The approach gives a multi-layered perspective on performance, and it is structured around three complementary levels: micro-benchmarks, standard HPC benchmarks, and full scientific applications. Technology Readiness Levels (TRLs) are introduced as a guiding concept, allowing the methodology to be adapted according to the maturity of the system under study. At high TRL, the methodology enables comparative assessments of production supercomputers, while at low TRL, it helps identify bottlenecks and optimization opportunities early in the design cycle.The thesis contributes both conceptual and practical tools. It formalizes performance and efficiency models (including Roofline, Top-Down, and efficiency metrics) and demonstrates their use across multiple architectures. It further extends tracing and monitoring capabilities for emerging processors, introduces methods to access and interpret hardware counters on novel architectures such as \riscv, and evaluates the integration of experimental hardware through Software Development Vehicles (SDVs) and FPGA-based emulation. These tools are validated through case studies on production systems, such as the MareNostrum 5 supercomputer and other HPC clusters deployed at the Barcelona Supercomputing Center (BSC), as well as on prototypes from European projects, such as EPAC.Results show that the proposed methodology provides actionable insights at all maturity levels: from guiding hardware-software co-design in early-stage processors to enabling reproducible performance comparisons across pre-exascale systems. Beyond benchmarking, it provides valuable feedback for hardware architects, system software developers, and application scientists alike. By bridging the gap between low-TRL prototypes and production-ready HPC systems, this work contributes to building a consistent framework for evaluating and improving the efficiency of future European and global supercomputers.
  
• BARRERA HERRERA, JAVIER ENRIQUE: Improving Time Predictability and Code Coverage of Embedded GPUs for Real-Time Systems
  
  Author: BARRERA HERRERA, JAVIER ENRIQUE
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 07/11/2025
  Deposit END date: 20/11/2025
  Thesis director: CAZORLA ALMEIDA, FRANCISCO JAVIER | KOSMIDIS, LEONIDAS
  Thesis abstract: This dissertation addresses challenges that the adoption of GPUs in Critical Embedded Systems (CES) faces, namely, Time Predictability and Code Coverage. Different domains that deploy CES are constantly adding Artificial Intelligence (AI)-based features, such as autonomous driving, that demand high performance levels. Multi-Processors Sytem-on-Chip (MPSoCs) are widely used to provide said performance levels, as they are equipped with accelerators, among which, Graphics Processing Units (GPUs) are a common choice. However, CES must undergo a rigorous Verification and Validation (V&V) process, in which a certain level of Execution Time Determinism (ETD) must be guaranteed. The use of several tasks to increase the overall utilization introduces contention in shared resources, which induces time variability. To provide the ETD guarantees, the time variability must be either mitigated or tracked and controlled. Another challenge for the adoption of GPUs in CES, is that the V&V process demands evidence of the thoroughness of the testing phase, for which Code Coverage is used as a test quality indicator. However, Code Coverage, as traditionally understood for sequential CES does not cover all possible scenarios in which a GPU thread might execute.For low-criticality and mixed-criticality CES, we contend that we can allow tasks to share the Last Level Cache (LLC) if hardware support for contention tracking is provided. Providing a clear understanding on how tasks contend with each other enables CES developers to balance performance and time predictability. For high-criticality CES, it is a common practice to implement LLC partitioning as it allows tasks to access LLC without suffering from inter-kernel contention, however, tasks may experience a performance loss due to lack of resources. In this Thesis, we propose Demotion Counters, a novel technique that tightly tracks how much each task has been demoted towards eviction in the LLC, thus, effectively quantifying their impact in CES. Additionally, we also assess the use of NVIDIA’s Multi-Instance GPU (MIG) feature as means to improve ETD in high-criticality CES.Code Coverage is used as a test quality indicator to provide evidence of the thoroughness of the testing, as required by the V&V process. However, if applied as traditionally understood, it will ignore the threading dimension of GPUs. Threads have private regions of memory, as well as shared regions at different granularities. This means that errors that are innocuous to one thread are potentially harmful for another, hence, it does not cover all possible cases under which GPU threads might execute. In this Thesis, we propose the use of Per-Thread Statement Coverage (PTSC), which tracks the Code Coverage at thread granularity. In order to mitigate the overheads caused by PTSC, several variants that apply different orthogonal optimizations are also proposed. Finally, we also evaluate the potential benefits of using hardware support for PTSC, mitigating the memory consumption of PTSC, as well as the execution time impact at deployment.Summarizing, this Thesis advances the state of the art in the adoption of GPUs in CES. The proposal of hardware contention tracking support and assessment of NVIDIA’s MIG, as means to improve ETD, effectively tackles the Time Predictability challenge in shared LLC. The proposal of software PTSC allows providing CES designers with the whole picture of the execution in commercially available GPUs. The use of hardware support for PTSC mitigates the overheads of software PTSC in deployment, while the different compression techniques reduce the volume of information during testing phase without losing data. Therefore, this Thesis provides means to face the Time Predictability and Code Coverage challenges of GPUs in CES.
  
• SABRI ABREBEKOH, MOHAMMAD: Improving Efficiency of ReRAM-Based Accelerators for Cognitive Computing Workloads
  
  Author: SABRI ABREBEKOH, MOHAMMAD
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN COMPUTER ARCHITECTURE
  Department: Department of Computer Architecture (DAC)
  Mode: Normal
  Deposit date: 07/11/2025
  Deposit END date: 20/11/2025
  Thesis director: GONZÁLEZ COLÁS, ANTONIO MARIA | RIERA VILLANUEVA, MARC
  Thesis abstract: Deep Neural Networks (DNNs) have achieved remarkable success across a wide range of applications. The main operation in DNNs is the dot product between quantized input activations and weights. Previous works have proposed memory-centric architectures based on the Processing-in-Memory (PuM) paradigm. ReRAM technology is especially appealing for PuM-based DNN accelerators because of its high density for weight storage, low leakage energy, low read latency, and high-performance capabilities to perform DNN dot products massively in parallel within ReRAM crossbars. However, there are three main bottlenecks in ReRAM-based accelerators.First, the energy-hungry Analog-to-Digital Converter (ADC) required for in-ReRAM analog computations, which undermines the efficiency and performance benefits of PuM. To improve energy efficiency, we present ReDy, a hardware accelerator that implements a novel ReRAM-centric dynamic quantization scheme, leveraging bit-serial streaming and processing of activations. The energy consumption of ReRAM-based DNN accelerators is directly proportional to the numerical precision of input activations in each layer. ReDy exploits the fact that activations in convolutional layers are often grouped according to filter sizes and crossbar dimensions. It quantizes each group of activations on-the-fly with different precision levels, based on a heuristic that considers the statistical distribution of each group. Overall, ReDy significantly reduces ReRAM crossbar activity and the number of A/D conversions compared to static 8-bit uniform quantization. Evaluated on a set of modern CNNs, ReDy achieves on average 13% energy savings over an ISAAC-like accelerator, with negligible area overhead.Second, the costly writing process of ReRAM cells has led to accelerators designed with enough crossbar capacity to store entire DNN models. Given the continuous growth of DNN model sizes, this approach is infeasible for some networks and inefficient due to huge hardware requirements. These accelerators lack flexibility and face an adaptability challenge. To address this, we introduce ARAS, a cost-effective ReRAM-based accelerator that uses a smart scheduler to adapt various DNNs to resource-limited hardware. ARAS also overlaps computation of one layer with weight writing of others to mitigate high ReRAM write latency. Furthermore, ARAS introduces optimizations to reduce the energy overhead of ReRAM writes, including re-encoding weights to increase similarity across layers and reduce energy when overwriting cells. Overall, ARAS significantly reduces ReRAM write activity. Evaluated on multiple DNN models, ARAS delivers up to 2.2× speedup and 45% energy savings compared to a baseline PuM accelerator without optimizations, and up to 1.5× speedup and 62% energy savings compared to a TPU-like accelerator.Third, ReRAM cells suffer from limited endurance due to wear-out caused by repeated updates during inference, reducing the lifespan of ReRAM-based accelerators. Overcoming this endurance limitation is essential for making such accelerators viable in long-term, high-performance DNN inference. To address this, we propose Hamun, an approximate computing method designed to extend the lifespan of ReRAM-based accelerators through multiple optimizations. Hamun introduces a mechanism to detect and retire faulty cells caused by wear-out, preventing them from degrading accuracy. It also applies wear-leveling techniques across different abstraction levels and introduces a batch execution scheme to maximize cell utilization across inferences. Additionally, Hamun leverages the fault-tolerance of DNNs with a new approximation method that delays cell retirement, reducing the performance penalty and further extending lifespan. Evaluated on a set of DNNs, Hamun improves lifespan by 13.2× over a state-of-the-art baseline, with its main contributions coming from fault handling (4.6×) and batch execution (2.6×).
  

DOCTORAL DEGREE IN CONSTRUCTION ENGINEERING

• POSADA CÁRCAMO, HÉCTOR JOSÉ: Digital Twins for Concrete Building Construction Processes
  
  Author: POSADA CÁRCAMO, HÉCTOR JOSÉ
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN CONSTRUCTION ENGINEERING
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 12/11/2025
  Deposit END date: 25/11/2025
  Thesis director: CHACÓN FLORES, ROLANDO ANTONIO
  Thesis abstract: This dissertation primarily investigates the use of Digital Twin (DT) technology to enhance construction management in concrete buildings, offering an end-to-end analysis of related workflows, tools, and frameworks. It contributes to the academic field by addressing identified research challenges and supporting the advancement of DT technology both theoretically and practically.The dissertation explores the foundational concepts underpinning DTs, providing a review of their conceptual development and evolution in recent years within the construction industry. It identifies and elaborates on key technological enablers crucial to this research: i) OpenBIM (IFC standard), ii) Computational Design, iii) Internet of Things (IoT), iv) knowledge graph databases, and v) human-twin interfaces (DT platforms). Particular attention is given to the current state and challenges of integrating structural analysis into DT frameworks. A real-world case study anchors the research: the construction of a concrete office building. Through this empirical approach, six twinning information pipelines were developed, aiming to establish data flow from construction site measurements to actionable insights. These pipelines were crucial for identifying four research challenges addressed in this research: A) How can multi-layered information related to concrete construction be generated, prepared, and streamlined while ensuring accuracy and interoperability for DTs? B) What roles and workflows should stakeholders adopt to enable coordinated DT implementation? C) How can structural analysis be effectively integrated into DT systems in a scalable and interoperable manner? D) What kind of system architecture can unify diverse data layers and information pipelines to support right-time, data-driven decision-making?Research challenge A is addressed through a software development: MatchFEM. Conceived as a plugin within a computational design tool, MatchFEM streamlines the often fragmented processes of 4D IFC-BIM modeling, IoT data integration, and structural analysis by unifying them within a single parametric environment. The plugin follows a visual programming paradigm, thus simplifying the generation and preparation of DT data.The second research challenge encompasses a general mind map and two complementary workflows. They delineate the essential job roles involved in the creation and operation of DTs during building construction. Moreover, a conceptual framework for the emerging role of the DT Manager is proposed, highlighting their importance in coordinating and overseeing all DT-related activities.Furthermore, to bridge the gap between structural analysis and DT ecosystems, research challenge C, two novel data models are introduced: O-SAM (Open Structural Analysis Models), a JSON schema for encoding and transferring structural simulation data via web-based platforms, and SSO (Structural Simulation Ontology), an ontology designed to represent O-SAM data as a graph, enabling its integration within knowledge graph-based DTs.To knit all these proposals, a comprehensive DT system connected to real-time structural simulations during concrete construction is presented. This system addresses research challenge D and consolidates the knowledge, tools, and frameworks developed throughout the dissertation. The implementation of knowledge graphs as a central linking framework is emphasized, alongside the development of a human-twin interface that delivers Performance Indicators to Construction Managers. The system is validated through a prototype implementation, which incorporates an exemplary construction management workflow: the Maturity Method for concrete slabs.The dissertation concludes by reflecting on findings and contributions to the field, discussing limitations encountered, and outlining avenues for future research, envisioning DTs as data-driven assistants that enhance productivity and sustainability in the construction sector.
  

DOCTORAL DEGREE IN EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS

• WANG, FUMIN: A Simplified Approach for the Seismic Analysis of Compliant Soil Slopes
  
  Author: WANG, FUMIN
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN EARTHQUAKE ENGINEERING AND STRUCTURAL DYNAMICS
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Normal
  Deposit date: 07/11/2025
  Deposit END date: 20/11/2025
  Thesis director:
  Thesis abstract: Over the past few decades, post-earthquake investigations have revealed extensive damage to both natural and engineered slopes, including landslides, slope collapses, and other forms of instability. These observations highlight the urgent need to improve our understanding of slope behavior under seismic loading and to refine analytical approaches for stability assessment. This thesis develops simplified yet robust methods for evaluating the seismic stability and dynamic performance of slopes, balancing physical insight with analytical efficiency.Chapter 2 reviews existing analytical frameworks —including the traditional Newmark rigid block, decoupled, and coupled analyses— used to estimate slope stability and dynamic responses under earthquake excitation. It compares their performance, identifies limitations, and outlines research needs.Chapter 3 proposes an extremely simplified and computationally efficient formulation for estimating the Factor of Safety (FS) of co-seismic compliant slopes, explicitly accounting for soil slope, ground motion direction, and groundwater effects. The method clarifies the differences between coupled and Newmark-type analyses and introduces a modified yield acceleration for compliant slopes. Validation through experimental, numerical, and field data confirms its simplicity and accuracy.Chapter 4 presents, as well, a coupled linear analysis based on a simplified two-block flexible model subjected to harmonic seismic excitation. Analytical and numerical solutions are derived and verified, and sensitivity analyses reveal the influence of ground motion direction, amplitude of input motion, damping ratio, and slope properties. Simplified upper bounds for maximum acceleration and displacement for the first cycle of rigid and compliant slopes are proposed, demonstrating the model’s predictive capability despite its simplicity.Chapter 5 examines the performance of classical discrete lumped-mass models against continuous wave-equation-based formulations under sinusoidal seismic loading without sliding. Both approaches yield consistent results for relative displacement and shear stress. It also shows that both models predict similar failure patterns, with cohesion playing a key role in shifting the critical sliding surface from the surface to deep depth. Incorporating Maxwell-type damping into the continuous model proves valid as an alternative to the traditional Kelvin-Voigt model in linear analysis.As a summary, this thesis extends and generalizes the simplified traditional methods for seismic analysis of soil slopes, including the stiffness of the sliding material. The proposed methodologies, validated through published experiments and real cases, provide an efficient and reliable framework for seismic slope stability assessment and offer valuable guidance for future research and engineering practice.
  

DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS

• IRAWAN, AMIR MUSTOFA: Explainable Artificial Intelligence Applied to Geoscience and Remote Sensing: Development and Application to Wild Fire Forecasting Related to Climate Change
  
  Author: IRAWAN, AMIR MUSTOFA
  Thesis file: (contact the Doctoral School to confirm you have a valid doctoral degree and to get the link to the thesis)
  Programme: DOCTORAL DEGREE IN SIGNAL THEORY AND COMMUNICATIONS
  Department: Department of Signal Theory and Communications (TSC)
  Mode: Normal
  Deposit date: 17/11/2025
  Deposit END date: 28/11/2025
  Thesis director: VALL-LLOSSERA FERRAN, MERCEDES MAGDALENA | LOPEZ MARTINEZ, CARLOS
  Thesis abstract: this thesis presents a progressive exploration of wildfire prediction by integrating process-based understanding with machine learning and causal inference frameworks. Chapter 3 focuses on variable importance and sensitivity by applying perturbation-based interventions, altering key drivers such as vapour pressure deficit (VPD), soil moisture (SM), and jet stream metrics by up to ±25% to simulate intensified environmental conditions and assess their impact on burned area. In contrast, Chapter 4 employs formal causal inference through do-calculus, enabling targeted counterfactual analysis within a structural causal model (SCM). Unlike the continuous perturbation-based interventions in Chapter 3, the intervention scenarios here are implemented by bootstrapping input variables and setting them to the 25th, 50th, 75th, and 100th percentiles. This allows the model to simulate the impact of each variable across a range of conditions, from typical to extreme (worst-case), and to quantify both direct and indirect effects on burned area, particularly for key drivers such as ∆Z500 and v300. Chapter 5 extends the causal reasoning to a global scale by using PCMCI-derived graphs as structural priors within a deep learning framework. It introduces regime-specific directed acyclic graphs (DAGs) generated through spatial clustering using the DBSCAN algorithm, enabling the identification of region-specific land–atmosphere interactions. These causal graphs are then embedded into Graph Attention Networks (GATs), allowing the model to learn weighted connections informed by causal structure, thereby enhancing both predictive performance and physical interpretability. Finally, Chapter 6 synthesizes these advances by embedding causal graphs within a GAT to simulate complex, multiscale interventions. It incorporates explicit counterfactual scenarios simulating intensified El Niño (via doubled negative SOI) and jet stream ridging (via increased positive ∆Z500, v300, and jet core), revealing spatially distinct fire responses. The use of different intervention strategies across chapters reflects the evolving methodological focus, from assessing input sensitivity (Chapter 3), to inferring causal mechanisms (Chapter 4), validating causal structures across regions (Chapter 5), and finally quantifying scenario-based outcomes (Chapter 6). Building on this foundation, Chapter 6 introduces a causal GAT capable of predicting global burned area by integrating physically grounded causal graphs derived from PCMCI. This approach enables the model to follow meaningful land–atmosphere interactions, improving interpretability and aligning predictions with known physical processes. The results show that the causal GAT outperforms models using fully connected graphs. Excessive or non-informative edges in fully connected structures can lead to over-smoothing, a common issue in Graph Neural Networks, where repeated message passing across redundant links blurs key distinctions among node representations. This can obscure critical predictive features and degrade overall model accuracy. By pruning spurious or weakly informative connections, the causal GAT preserves sharper, more meaningful node embeddings and avoids the performance loss typically associated with over-parameterized graph structures. Collectively, these advances underscore that correlation-based models fail to capture the complex, non-linear interactions among ignition sources, vegetation dynamics, and climate feedbacks. They advocate for a shift toward process-based and machine learning models that can better represent the multifaceted mechanisms governing wildfire regimes in a warming world.