Author: LÓPEZ MENESES, DUVÁN HERNÁN
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: 26/07/2023
Reading date: pending
Reading time: pending
Reading place: pending
Thesis director: GARCÍA DEL MURO SOLANS, JOAN | FRASER, ARABELLA
Committee:
     PRESIDENT: GRIMM, NANCY B.
     SECRETARI: ALVAREZ DEL CASTILLO, JAVIER
     VOCAL: NOGUERA DE ECHEVERRI, ANA PATRICIA
Thesis abstract: The investigation of climate risks affecting the popular neighborhoods in the mountainous outskirts of Bogotá, Colombia, has served as a case study and starting point to critically study the concept of risk and develop analytical tools that facilitate understanding of these dynamics and their governance from both a local and global perspective.For this purpose, the voices of Arraigo, a platform representing citizen resistance to resettlement due to risk, have been amplified. Meanwhile, the study has explored the complexity of risk through theoretical perspectives encompassing environmental thought, environmental theology, risk sociology, and an understanding of reality provided by the ontology of difference.An initial hypothesis suggests that in Bogotá, institutional mechanisms for managing climate risk are overwhelmed by the complexity of Colombian society impacting its capacity for adaptation. Therefore, investigating analytical tools to understand risk conflicts and improve institutional management is seen as an opportunity to "adapt" the relationship between Colombian society and its territorial complexity.Based on documentary research and significant testimonies, the institutional and citizen narratives were contrasted to support the research efforts. Additionally, by integrating different theoretical backgrounds, a notion of "risk" is framed as a measure of entropy, introduced in the interaction between human systems and nature, through operations of «distancing» and «organized irresponsibility» characteristizing modern epistemological and administrative practices.In response to risk, the study supports a project aligned with an organizational trend nested within «consciousness». Overall, the research presents a conceptual framework of descriptive and analytical risk tools and their management horizons, unveiling multiple polarities or tensions, such as those between «organization» and «risk», «consciousness» and «distance», or «responsibility» and «predation».Finally, three distinct risk operations or dynamics are defined: risk production, attribution, and distribution. These are presented as areas of action through which risk can be understood and effectively managed. In this way, the study contributes to a vision of risk governance and adaptive evolution of conflict scenarios linked to complexity.

Author: BLASCO COLL, ANDREU
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: 12/07/2023
Reading date: 27/09/2023
Reading time: 12:00
Reading place: Aula de Seminaris D5-007, Dpt. Teoria del Senyal i Comunicacions - ETSETB
Thesis director: RODRIGUEZ FONOLLOSA, JAVIER
Committee:
     PRESIDENT: CRESPO BOFILL, PEDRO
     SECRETARI: COCCO, GIUSEPPE
     VOCAL: SENTÍS HERRERA, GAEL
Thesis abstract: Quantum computers use the laws of quantum mechanics to do complex operations that classical computers can't solve. Quantum computers and other physical quantum systems require quantum information to be completely isolated from the environment in order to be perfectly functional. Quantum systems that are not completely isolated suffer from decoherence noise and may suffer from errors. Implementing these systems is a huge challenge and development of suboptimal systems with a low error rate seems to be the best way forward to building these systems.In quantum communications, designing codes that are able to reduce or protect against errors is a necessity. This work focuses on perfect and quasi-perfect quantum codes, which are a family of codes that exists in the classical setting and are optimum in terms of minimizing the error probability for a given number of channel uses. In the first part of the work we generalize the definition of quasi-perfect codes to include both classical codes and quantum codes. We also provide an example of a quantum quasi-perfect code for 2-qubits classical-quantum channels that make use of quantum entanglement and that can be extrapolated to an N-dimensional classical-quantum channel. The second part of this work focuses on quasi-perfect codes in optical communications, where coherent states are used to convey information through an optical channel, known as the Bosonic channel. The Bosonic channel has infinite dimension, so instead we consider a finite-dimensional approximation of the Bosonic channel with a negligible approximation error for a sufficiently large channel dimension. We show that phase-modulated coherent states constitute a codebook that is quasi-perfect for the approximated channel, and thus are close to optimal for the Bosonic channel.The last part of the work focuses on stabilizer error correction codes, which are practical codes that use redundancy to protect against errors. Error correction is not specifically used to transmit classical information, but to protect quantum information instead. However, it is possible that we require to obtain classical information from a quantum state after performing error correction. For these particular cases we may be able to prove that an error correction code is quasi-perfect and thus optimum.

Author: FERNÁNDEZ ESTEBERENA, PABLO RICARDO
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: 18/07/2023
Reading date: 27/09/2023
Reading time: 11:00
Reading place: ICFO, Mediterranean Technology Park, Avinguda Carl Friedrich Gauss, 3, 08860 Castelldefels, Barcelona
Thesis director: DURDURAN, TURGUT
Committee:
     PRESIDENT: LIGUORI, NICOLETTA
     SECRETARI: REVERTER CALATAYUD, JORGE LUIS
     VOCAL: BUSCH, THERESA
Thesis abstract: Near-infrared diffuse optical spectroscopy (DOS) techniques are capable of non-invasive measurement of microvascular hemodynamic parameters of deep tissues (>1 cm penetration). One important focus of the field has been applications in oncology, where the characterization of the tumor vasculature could play both a diagnostic and therapeutic role. This is mainly because malignant tissue growth requires the co-option and generation of blood vessels to be supported. Microvessel density, microvascular blood flow and hypoxia have been linked to the progression of the disease, the likelihood of metastasis and patient survival, and can thus guide diagnosis, treatment plans and prognosis.In this thesis, I have applied DOS methodologies in two projects exploiting hemodynamic information for cancer management. The first one consists in the demonstration of a toolbox for the optimization of plasmonic photothermal therapy (PPTT) in mice, while the second focuses on the contribution to the clinical diagnosis of thyroid cancer in nodule patients.PPTT uses plasmonic nanoparticles (NPs) that are injected into the body and act as localized sources of heat upon external illumination to induce tumor cell death. Various aspects of the therapy are modulated by the optical and hemodynamic parameters of tumor tissue and can thus be studied with DOS monitoring. Moreover, it enables simultaneous quantification of NP concentration. Such information can improve the understanding and the outcome of the treatment and accelerate its slow progress to the clinics.To prove this, we conducted experiments using DOS monitoring along PPTT to model the therapy steps and explain the variability among individuals. This was done on patient-derived orthotopic renal cell carcinoma models, injected with gold nanorods and treated with fixed conditions. The hybrid device combined continuous-wave broadband diffuse reflectance (DRS) and diffuse correlation (DCS) spectroscopies. The data was related to the NP accumulation, the temperatures reached during treatment, tumor growth and animal survival. Moreover, we analyzed the underlying mechanisms with simulations and demonstrated the extrapolation of therapy conditions for individual mice. With this, we managed to determine the relevant interactions and prognostic factors to guide the personalization of the therapy in a way that could be readily applied to other PPTT protocols.The second study was part of the LUCA project, which aimed at implementing an ultrasound-guided hybrid DOS device that could be integrated into the clinical workflow of thyroid cancer screening and provide relevant hemodynamic information to improve diagnostic capabilities. The incidence of thyroid cancer has been on the rise for decades and the low specificity of standard screening methods implies tens of thousands of unnecessary thyroid extraction surgeries are carried out each year just in Europe. Therefore, any improvement in the discrimination between benign and malignant nodules can have a relevant impact on the large scale.A clinical campaign was carried out with the LUCA device, a state-of-the-art device combining time-resolved spectroscopy (TRS) and DCS around a regular ultrasound transducer. The properties of thyroid, nodules and neck muscles were measured in nodule patients and healthy volunteers. Data from sixty-six subjects allowed to characterize these tissues, study the effect of demographic and anatomical variables and assess their diagnostic capabilities. In this way, we gathered a large reference data set relevant to various medical applications beyond oncology and identified the most promising indicators of malignancy.With these studies I have shown how the information obtained with DOS has important roles in these particular applications and what are some of the mechanisms behind them. These results can guide the future steps of research endeavors and have implications for the advancement of solutions to these clinical problems.

Author: GONÇALVES FERREIRA, CATARINA
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: 26/06/2023
Reading date: 27/09/2023
Reading time: 16:00
Reading place: ICFO, Mediterranean Technology Park, Avinguda Carl Friedrich Gauss, 3, 08860 Castelldefels, Barcelona
Thesis director: MARTORELL PENA, JORDI
Committee:
     PRESIDENT: VAN HULST, NIEK
     SECRETARI: KOZYREFF, GREGORY
     VOCAL: ANDREU ARBELLA, TERESA
Thesis abstract: To mitigate the present environmental crisis, caused by the excessive use of fossil fuels and associated release of carbon dioxide into the atmosphere, it is necessary to significantly reduce worldwide energy consumption, to rely more strongly on clean and renewable sources of energy, but also to maximize energy efficiency in currently existent technologies that make use of energy. To reach such maximal energy efficiency, it is necessary to optimize light propagation, harvesting, and utilization in the different existent optoelectronic technologies. Given that a considerable portion of the global energy consumption is dedicated to illumination or devices incorporating illumination sources in them, a clear path to maximize energy-efficiency would imply minimizing the light losses in such kind of systems. In addition, for maximal energy conversion efficiency it is essential to optimize light absorption in systems that perform an unassisted sunlight transformation into other forms of energy, such as electrical or chemical.To reach the double goal of optimizing light utilization and transformation, in this thesis we consider the study of optical ergodic configurations, where light rays are randomized after a few bounces at the interfaces, losing any correlation with the external incident state and giving rise to an isotropic radiation inside the material. In Chapter 2 of the thesis, we demonstrate that an ergodic geometry can be used to obtain homogeneously distributed polarized light emission. In the same ergodic system, we also demonstrate that the light with the unwanted polarization can be trapped and transformed back into electricity by using a couple of perovskite solar cells. Such features are potentially useful to increase energy efficiency in optoelectronic devices incorporating illumination sources in them, as is the case of liquid crystal displays. A similar ergodic light propagation is also considered in Chapter 3 to determine what the maximal light trapping and effective light absorption is in a BiVO4-based photoanode of a photoelectrochemical cell used for light transformation into hydrogen. The limits in the efficiency of such energy transformation are seen to be strongly linked to the weakly light-absorbing sub-bandgap states. A three-dimensional nano-structuration of the photoanode in the photoelectrochemical cell is explored as a path to eventually reach ergodicity for light propagation in the photoanode. In the final chapter of the thesis, we consider a tandem construction of two complementary light absorption elements, such as a BiVO4 photoanode and an organic solar cell, to obtain an unassisted conversion of sunlight into hydrogen in photoelectrochemical cells. Optical multilayers designed by implementing an inverse problem-solving approach are found to be an essential ingredient to properly balance light absorption among such two light-absorbing elements in the tandem, leading to an optimal solar-to-hydrogen conversion.

Author: SANZ GRÀCIA, ELVIRA 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 AUTOMATIC CONTROL, ROBOTICS AND VISION
Department: Department of Automatic Control (ESAII)
Mode: Normal
Deposit date: 28/07/2023
Reading date: pending
Reading time: pending
Reading place: pending
Thesis director: PUIG CAYUELA, VICENÇ | BLESA IZQUIERDO, JOAQUIN
Committee:
     PRESIDENT: STETTER, RALF
     SECRETARI: TORNIL SIN, SEBASTIAN
     VOCAL: ANSARI, FAZEL
Thesis abstract: Transformation of industrial processes to embrace Industry 4.0 revolution needs to use data: valuable data from different data sources, labelled data, semantic layers, metadata to understand better, universes, and data bases integration (relational and non-relational data bases, stored in physical servers or in the cloud, real-time plus historical data, etc.).With data analyses and Artificial Intelligence algorithms can be applied, resulting data brings new insights and can be used as inputs to new analyses and new algorithms. Multiplicity of data analyses and algorithms on the same data but changing the perspectives and the focus generate valuable insights that improve the knowledge in the industrial process. This thesis defines an ecosystem where all the data is integrated in the same Data Lake from the heterogeneous Industrial Cyber-Physical Systems in the industrial process.The car coating process is introduced for illustration purposes as it is found in it several disconnected contributions that, at the end, they are correlated and allow us to build an Ecosystem to integrate all the ¿Things¿ of the ¿Industrial Internet of Things (IIoT)¿ with data, advanced analytics and artificial intelligence.Paint shop process is a complex industrial process in terms of logistics, fabrication times, tact times, re-works in-line and out of line due to quality defects in the surface. How decisions in the Paint Shop affected (from a production point of view) to the rest of processes involved in car manufacturing (Body shop and Assembly lines) because of changes in cycle times, sequences, process incidences, etc. are to complex and need deeper analyses.This Thesis proposes an ecosystem that integrate eleven data sources involved in an industrial process in order to apply advanced analytics and artificial intelligence models. BiDrac Ecosystem is defined and implemented in the Paint Shop process of an automotive industrial plant. The Thesis is focused on maintenance area so first main objective is to integrate the Industrial Cyber-physical systems that were in two data sources not connected, from two control systems in the control room of the Shop. In the literature, studies are focused in parts of the ecosystem proposed in this thesis but there is a leak of studies based on real data from a real industrial process considering the data of the entire process, all Industrial Cyber-Physical systems, equipment and machines, and data from the different areas involved.The solution adopted includes the proposed data flow, that is designed as follows, data collected from the Technical Ubications and Equipments (installation) arrive to different data sources not integrated, then are pre-processed with an extract, transform and load (ETL) tool, to detect any abnormality or fault and each diagnosis and prognosis in data, and are stored into an edge computing Data Lake (BiDrac) with a new metadata layer target (label) and integrate them to priorize those data that are relevant for advanced monitoring and artificial intelligence modelization. The different phases of the BiDrac methodology are: (1) Ecosystem design and technical specification and requirements; (2) Research work to achieve project objectives and (3) The development of use cases analytics applying advanced analytics and artificial intelligence in order to validate and value of the proposed approach. The methodology is applied from general to specific. In this case, the specific industrial process of coating bodies, but this methodology can be applied to all industrial process over different industry sectors.The use of Artificial Intelligence techniques inside the BiDrac Ecosystem with all the data available to improve the knowledge that areas have in the process and to discover hidden knowledge with deeper analysis using the Digital Production Platform of the Volkswagen Group in the Cloud is demonstrated.