Doctoral School

Reading date: 14/09/2026

• GARÍ GALÍNDEZ, JON: Developing Artificial Intelligence Strategies to Accelerate Scientific Research: Application to Thin-Film Photovoltaic Technologies
  
  Author: GARÍ GALÍNDEZ, JON
  Programme: DOCTORAL DEGREE IN COMPUTATIONAL AND APPLIED PHYSICS
  Department: Department of Physics (FIS)
  Mode: Article-based thesis
  Deposit date: 13/07/2026
  Reading date: 14/09/2026
  Reading time: 10:00
  Reading place: Edifici A, 1a planta, aula A1.07Campus Diagonal-Besòs – UPCAv. d'Eduard Maristany, 1608930 Sant Adrià de Besòs (Barcelona) https://meet.google.com/qmy-zzfx-xnx
  Thesis director: IZQUIERDO ROCA, VICTOR | GUC, MAXIM
  Thesis abstract: The development cycle of new materials-based technologies typically spans 15-25 years due to the constantly increasing complexity of experimental research processes. Accelerating this process is essential to enable large-scale implementation of sustainable technologies that support the green energy transition and secure the electrification of society. Photovoltaics (PV) play a central role in this transition, particularly novel thin-film photovoltaic (TFPV) technologies, which promise to significantly broaden the application of solar modules in everyday life. Therefore, accelerating their development and transfer is key to achieving this goal. In this context, recent advances in artificial intelligence (AI), explainable artificial intelligence (XAI), robotics, and automation have enabled a new paradigm: the data-driven materials science, capable of accelerating materials discovery and optimization. However, despite the increasing use of machine learning (ML) in materials science, the application of XAI approaches to real experimental characterization datasets remains limited, and the translation of ML model outputs into physically meaningful insights is still a major challenge.The present work addresses this challenge by exploring and developing AI-driven strategies for the analysis and interpretation of experimental characterization data, following a proposed interpretability framework aimed at both accelerating research and enhancing explainability of ML models. The work begins by implementing and optimizing different training strategies for restricted Boltzmann machines (RBMs) using benchmark datasets. This study is aligned with the first level of the interpretability framework (Level 1: Using AI) and establishes the basis for the implementation of ML algorithms in experimental characterization datasets. Subsequently, an initial approach to XAI-driven strategies for extracting valuable information from experimental characterization data of kesterite-based TFPV devices is developed (Level 2: Using XAI), demonstrating the potential of interpretable ML models for identifying correlations between the physicochemical properties of materials and the PV performance of devices.Finally, building on this progression, an integrated XAI-driven methodology is implemented to analyze multimodal datasets obtained from holistic characterization of kesterite-based solar cells. It integrates the entire research cycle, from synthesis and characterization to data analysis, interpretation, and extraction of valuable feedback for further technology optimization. Particular emphasis is placed on interpretability, where the combination of intrinsic XAI models with post-hoc XAI techniques enables the identification of the physicochemical properties that govern device performance and reproducibility, linking the relevant spectral zones with the corresponding physical properties. This strategy reaches the third level of the interpretability framework (Level 3: Using XAI+) and provides meaningful feedback to guide subsequent research cycles. This feedback is validated through modified fabrication processes, resulting in improved device performance. Furthermore, it substantially reduces research cycle times, accelerating material optimization from timescales exceeding one year to less than a week. These results contribute to the development of self-driving laboratory (SDL) approaches and the advancement of data-driven materials research, ultimately supporting the green energy transition.
  

Reading date: 18/09/2026

• ERBS, FLORENCE AMANDINE: AI-enhanced Passive Acoustic Monitoring of Tropical Freshwater Environments
  
  Author: ERBS, FLORENCE AMANDINE
  Programme: DOCTORAL DEGREE IN MARINE SCIENCES
  Department: Department of Civil and Environmental Engineering (DECA)
  Mode: Article-based thesis
  Deposit date: 28/05/2026
  Reading date: 18/09/2026
  Reading time: 14:00
  Reading place: Place: ETSECCPBUPC, Campus NordBuilding C1. Classroom: 002C/Jordi Girona, 1-308034 Barcelona
  Thesis director: ANDRE SANCHEZ, MICHEL
  Thesis abstract: Tropical freshwater ecosystems, particularly in the Amazon basin, are among the world’s most vulnerable environments, facing escalating pressures from human activities such as overexploitation, pollution, and habitat degradation. Effective conservation efforts are constrained by the scarcity of long-term monitoring data, as traditional visual survey methods are often unsuitable for the high turbidity of Amazonian waters and the complex floodplain habitat mosaics where seasonally inundated lakes, channels, and forests make species detection challenging. This dissertation addresses these challenges by developing an innovative and scalable monitoring framework using Artificial Intelligence (AI)-enhanced Passive Acoustic Monitoring (PAM) focused on three ecological indicator species: the pink river dolphin (Inia geoffrensis), the tucuxi (Sotalia fluviatilis), and the Amazonian manatee (Trichechus inunguis).The research is presented as a compendium of two scientific publications that integrates longterm autonomous acoustic monitoring with Deep Learning techniques, specifically Convolutional Neural Networks (CNNs), to automate species detection in complex and noisy freshwater soundscapes. The first study focuses on Amazonian river dolphins, exploiting their near-continuous production of echolocation clicks for presence detection. The developed CNN classifier achieved a high average precision of 0.95 and successfully discriminated between dolphin clicks and impulsive interferences like rain and boat engine noise. This automation enabled the tracking of synchronized seasonal movements into floodplain bay and river channel habitats following the annual flood pulse. Furthermore, the study provided rare insights into the seasonal use of flooded forests revealing regular dolphin presence during high-water periods. Additionally, it quantified the spatio-temporal overlap between dolphins and boat traffic, establishing a baseline for assessing anthropogenic noise exposure in core habitats.The second study applied a similar AI-enhanced approach for detecting the Amazonian manatee, a cryptic species characterized by inconspicuous surfacing behavior and a previously poorly described vocal repertoire in the wild. By training a CNN model on wild vocalizations, the research provided the first detailed characterization of the wild Amazonian manatee’s vocal repertoire, identifying four distinct call types, and leading to new insights into the vocal behavior of the species in the floodplains. The analysis of vocal parameters, including repetition rates and frequency characteristics, revealed that Mamirau´a Lake, where the species was presumed absent, serves as a critical nursery habitat.This dissertation demonstrates that the integration of PAM and machine learning offers a noninvasive, cost-effective, and performant tool for monitoring biodiversity in remote tropical freshwater environments. The transition from manual acoustic analysis to automated processing allows for efficient conversion of high-volume acoustic datasets into ecological insights, facilitating the identification of critical habitats, supporting the understanding of floodplain ecological processes, and advancing the assessment of anthropogenic impacts and ecosystem health. These findings provide a robust foundation for a standardized Amazonian monitoring framework to support evidence-based conservation for these endangered species and their ecosystems.
  

• GOYZUETA MEJÍA, ZARELA MÓNICA: L'orientació de client a partir de l'experiència educativa: la perspectiva d'estudiants de pregrau a les universitats privades del Perú
  
  Author: GOYZUETA MEJÍA, ZARELA MÓNICA
  Programme: DOCTORAL DEGREE IN BUSINESS ADMINISTRATION AND MANAGEMENT
  Department: Department of Management (OE)
  Mode: Normal
  Deposit date: 08/06/2026
  Reading date: 18/09/2026
  Reading time: 12:00
  Reading place: ETSETB- UPC, Aula B3 Teleensenyament Enllaç públic: meet.google.com/ipt-evpc-xzc
  Thesis director: CONSOLACION SEGURA, CAROLINA MARIA | BARREDO IBAÑEZ, DANIEL
  Thesis abstract: Neoliberalism, as a model of global governance, has promoted competition and market efficiency as pillars of economic and social progress. Its influence has permeated various sectors, and higher education has been no exception. In this sphere, this ideology has driven the emergence of marketization, a trend that conceives of the education system under a supply and demand model, where education is understood as a good or service that can be acquired, and whose value is measured primarily by the economic return it generates for the individual and its contribution to the economy as a whole.In this context, the metaphor of the student as customer arises, a concept of business origin that positions the student as the central consumer of an experiential educational service, aimed at providing them with a satisfactory academic foundation for their professional development.This phenomenon has generated extensive debate in literature, with arguments both for and against. However, existing research has not delved deeply into educational experience as a determining variable in shaping the profile of the student as client. Furthermore, most studies have focused on Europe and the United States, leaving Latin America—and particularly Peru—as a largely unexplored context.Within this framework, the purpose of this doctoral thesis is to analyze how client orientation is shaped in undergraduate students at private universities in Peru, considering the dimensions of their educational experience, the influence of demographic variables, and the perspective of academic experts. To this end, a mixed-methods approach was adopted.First, a systematic literature review on the student as customer metaphor (2000–2022) was conducted in the Scopus and Web of Science databases, identifying 83 studies that met the eligibility criteria. Second, the SCOQ (Student Customer Orientation Questionnaire) scale by Koris and Nokelainen (2015) was adapted to the Peruvian context, and 798 surveys were collected from three private universities. The data were analyzed using exploratory factor analysis, which identified the dimensions of the educational experience that define customer orientation in Peruvian students. Subsequently, a multiple regression model was applied to assess the influence of demographic variables. Finally, the results were triangulated with 12 in-depth interviews with academic experts to corroborate and enrich the findings.The research results reveal four main findings. First, the literature is structured around two predominant approaches: a critical perspective, focused on the risks of commodification and the weakening of academic rigor, and a relational perspective, which conceives of the student as a co-producer of the educational service. Second, adapting the SCOQ instrument to the Peruvian context allowed, through exploratory factor analysis, the identification of five factors of the educational experience. In four of these factors, students exhibit a client orientation. In contrast, positive self-demand is the only factor in which this orientation is not observed, highlighting the hybrid and multidimensional nature of the Peruvian student, who combines service demands with academic responsibility. Third, the multiple regression analysis, while having limited explanatory power, showed that some demographic variables have predictive value, with years of study standing out as the most influential factor. Finally, the triangulation with experts showed that teachers adopt a conditional acceptance stance, in which they recognize certain expectations inherent in a service logic on the part of students, but subordinate their validity to the preservation of academic rigor and teacher autonomy.
  

• KANJ BONGARD, SEBASTIEN: Contribution to the Systematic Study of Threat Actor Tools and Techniques in Real-World Cyber Incidents
  
  Author: KANJ BONGARD, SEBASTIEN
  Programme: DOCTORAL DEGREE IN NETWORK ENGINEERING
  Department: Department of Network Engineering (ENTEL)
  Mode: Normal
  Deposit date: 22/06/2026
  Reading date: 18/09/2026
  Reading time: 10:30
  Reading place: Aula C3-304a (Seminari de Telemàtica), Campus Nord
  Thesis director: PEGUEROLES VALLES, JOSEP RAFEL
  Thesis abstract: Cyber incident response and digital forensics (DFIR) research increas- ingly faces a tension between academic rigor and the operational re- alities of real-world investigations. While academic studies often rely on controlled datasets and synthetic scenarios, practitioners must oper- ate under time pressure, incomplete evidence, and strict confidentiality constraints. This doctoral research addresses that gap through an In- dustrial PhD program conducted in collaboration with Incide Digital Data S.L., which enabled access to real-world incident data and practi- tioner workflows while maintaining a rigorous scientific methodology. The thesis investigates how threat-actor tools, techniques, and behaviours can be systematically analyzed and transformed into reusable, vali- dated, and operationally meaningful artifacts. Three complementary research directions are explored. First, an integrated forensic methodol- ogy is proposed for the analysis of abused legitimate tools and mobile activity, combining tool-centric investigations with structured iOS ac- tivity characterization to support accurate reconstruction, automation, and legally robust analysis in real incident-response contexts. Second, the research develops a Business Email Compromise (BEC)-specific map- ping of Tactics, Techniques, and Procedures using the MITRE ATT&CK framework, addressing limitations of narrative-driven BEC reporting. The proposed matrix is validated against real-world cases and demon- strates improved behavioural comparison, clustering, and defensive alignment. Third, the thesis designs and empirically evaluates a YARA- based detection library targeting malware anti-analysis techniques, with a focus on anti-virtual machine and anti-sandbox behaviour, highlight- ing both the strengths and limits of static detection in operational con- texts. Across these contributions, the thesis demonstrates that behavioural and tool-based indicators provide durable defensive value and that hy- brid analytical pipelines are necessary to address modern evasive threats. 6 By grounding methodological innovation in industrial practice, this work advances DFIR towards more systematic, reproducible, and operationally relevant research, supporting faster investigations and more consistent defensive decision-making.
  

• TOBAJAS ASENSIO, LUIS MIGUEL: ANÁLISIS DE LA INFORMACIÓN INSTITUCIONAL Y EL TRATAMIENTO DE LA PRENSA DE LOS INCIDENTES NUCLEARES ESPAÑOLES (1989-2009)
  
  Author: TOBAJAS ASENSIO, LUIS MIGUEL
  Programme: DOCTORAL DEGREE IN NUCLEAR AND IONISING RADIATION ENGINEERING
  Department: Department of Physics (FIS)
  Mode: Normal
  Deposit date: 17/06/2026
  Reading date: 18/09/2026
  Reading time: 11:30
  Reading place: Aula I 28.8 de l'Escola Tècnica Superior d'Enginyeria Industrial de Barcelona, ETSEIB
  Thesis director: REVENTOS PUIGJANER, FRANCESC-JOSEP | PONT SORRIBES, CARLES
  Thesis abstract: The general objective of this research is to analyze the safety of the nuclear power plants in Spain during the period between 1989 and 2009, through a comprehensive analysis of institutional information and press coverage of nuclear incidents.The methodology employed includes a mixed qualitative and quantitative model. Scientific production on nuclear incidents has been analysed through technical information databases, academic databases (Wos and Scopus), and newspaper archives.The newspapers ABC, La Vanguardia, and El País were selected based on the criterion of being among the most widely followed newspapers in Spain during the period studied, considering the press as the best means of monitoring incidents at the end of the 1980s.In 1990, following the Chernobyl accident, the INES Scale was introduced, created by the International Atomic Energy Agency (IAEA) and the Nuclear Energy Agency, with the aim of facilitating international communication to the public and the media. This study analyses incidents classified as level two or higher according to IAEA criteria. The results of this work consider the use of the INES Scale in institutional information and in the media.The conclusions allow for a deeper understanding of the treatment of information provided to the public and the media coverage of nuclear incidents. The analysis of lessons learned from the four incidents studied and their follow-up makes it possible to conclude that there have been advances in nuclear safety and in institutional and public information, with the consolidation of the INES Scale in nuclear communication in Spain. However, it is determined that the completion of corrective actions does not reach public opinion despite its relevance.The research provides proposals for improvement with the aim of promoting better communication in the future, including the participation of experts. Communication is one of the most relevant issues in nuclear crises and must be clear, timely, accurate, transparent, and proactive.