Doctoral School

Reading date: 29/04/2024

• HÖSCHELE, JONATAN: A strontium quantum-gas microscope
  
  Author: HÖSCHELE, JONATAN
  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: 25/03/2024
  Reading date: 29/04/2024
  Reading time: 10:00
  Reading place: ICFO, Mediterranean Technology Park, Avinguda Carl Friedrich Gauss, 3, 08860 Castelldefels, Barcelona
  Thesis director: TARRUELL PELLEGRIN, LETICIA
  Committee:
       PRESIDENT: SCHRECK, FLORIAN EBERHARD
       SECRETARI: DE RIEDMATTEN, HUGUES
       VOCAL: WEITENBERG, CHRISTOF
  Thesis abstract: The development of quantum-gas microscopes has revolutionized the field of quantum simulation with ultracold atoms. More specifically, their ability of direct observation and manipulation of degenerate quantum gases in optical lattices on a single particle level has brought novel ways of probing and engineering quantum degenerate many-body systems. So far, most of these setups have focused on alkali atoms. Combining quantum-gas microscopy with the properties of alkaline-earth atoms such as strontium gives rise to exciting research directions. In this thesis, we report on the design and construction of a strontium quantum-gas microscope. The findings in this thesis can be divided into three parts.In the first part, we focus on the accumulation of atoms in the science cell and develop a scheme to enhance the atom number in magneto-optical traps of strontium atoms operating on the 461-nm transition. This scheme resonantly populates a short-lived reservoir state, partially shielding the atomic cloud from losses in the cooling cycle. We demonstrate a factor of 2 enhancement in the atom number for the bosonic isotopes Sr-88 and Sr-84, and the fermionic isotope Sr-87, showing the efficient capture of these isotopes in our experiment. Our scheme can be readily implemented in the majority of strontium experiments, given that the shielding transition at 689 nm is commonly used for further cooling. In our case, the shielding scheme facilitates the generation of Bose-Einstein condensates.The second part of the thesis reports on the generation of degenerate quantum gases of Sr-84 with up to 200000 atoms. After summarizing the required cooling steps, we study the formation of Bose-Einstein condensates during evaporative cooling in our experiment. Analyzing the evolution of the horizontal and vertical size of our quantum-degenerate clouds in free fall leads to the characteristic asymmetric expansion, which we compare to theory for our experimental parameters. We also show the generation of smaller Bose-Einstein condensates of less than 20000 atoms with the help of a light-sheet potential. With this highly-anisotropic confinement we can consider our Bose-Einstein condensates two-dimensional for atom numbers of the order of 1000.In the third part we demonstrate site-resolved imaging of a Sr-84 bosonic quantum gas in a Hubbard-regime optical lattice potential. We confine the quantum gas by a two-dimensional optical lattice and the aforementioned light-sheet potential, both operating at strontium's clock-magic wavelength. A high-NA imaging objective enables single-atom and single-site resolved fluorescence imaging by scattering photons on strontium's broad 461-nm transition, while performing efficient attractive Sisyphus cooling of the atoms on a narrower transition at 689 nm. We reconstruct the atomic occupation of the lattice sites from the fluorescence images, obtaining imaging fidelities above 94%. Finally, we realize a Sr-84 superfluid in the Bose-Hubbard regime and observe its characteristic interference pattern after free expansion in the light sheet with single-atom resolution. Our strontium quantum-gas microscope provides a new platform to study dissipative Hubbard models and cooperative effects in atom-light interaction at the microscopic level. Moreover, the ability to capture also the fermionic isotope Sr-87 paves the way to generate degenerate Fermi gases with SU(N) symmetry and study SU(N) quantum magnetism.
  

Reading date: 30/04/2024

• MARTÍ SAUMELL, JOSEP: Agile aerial manipulation: an approach based on full-body dynamics and model predictive control
  
  Author: MARTÍ SAUMELL, JOSEP
  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: 26/03/2024
  Reading date: 30/04/2024
  Reading time: 11:00
  Reading place: Sala de Juntes de la Facultat Matemàtiques i Estadística (FME) de la UPC, C/Pau Gargallo, 14, 08028 Barcelona
  Thesis director: SANTAMARIA NAVARRO, ANGEL | SOLÀ ORTEGA, JOAN
  Committee:
       PRESIDENT: MANSARD, NICOLAS
       SECRETARI: MORCEGO SEIX, BERNARDO
       VOCAL: LIPPIELLO, VINCENZO
  Thesis abstract: Aerial manipulators, which commonly take the form of multirotors with attached robotic limbs, primarily employ their limbs for pure manipulation tasks and do not rely on them during aerial locomotion. Besides, their movement tends to be slow. This thesis aims to enhance an aerial manipulator¿s agility by harnessing its limb¿s capabilities to augment its overall motion. This objective involves investigating various modes of utilizing the limb: as a tail for aerial locomotion, as an arm for aerial manipulation, or as a leg for hybrid aerial-contact locomotion. The present thesis contributes to two specific domains: 1. Generation and control of agile motions for aerial manipulators, 2. Design and construction of a specialized aerial manipulator for executing agile motions.Generating agile motions requires predicting the movement of the robot considering its dynamics so that these dynamics can be used to favor the robot¿s motion. Hence, we can achieve complex maneuvers with relative ease. Optimal control is a trajectory-generation technique that meets these requirements, and that is central to this thesis. We encode the robot¿s tasks as cost functions of the optimal control problem (OCP) and use the whole-body dynamics as the constraints of the dynamic system. On the control side, to deploy such trajectories in a real robot, we use model predictive control (MPC) techniques, which is the closed-loop control extension of optimal control. To get the control command, an MPC controller solves the OCP in which we have encoded the agile trajectory, and then the controller applies the first command of the solution control trajectory. Thus, MPC requires solving an OCP at the control rate, i.e., within a few milliseconds. This forces us to use fast, specialized solvers based on the dynamic programming principle, such as differential dynamic programming (DDP). In their original form, these solvers cannot consider the control bounds. These bounds are important to create trajectories compatible with the real robot. To tackle this problem, in this thesis, we propose two DDP-based methods to consider the control bounds: one is based on a squashing function, and the other is based on a projection method. Even with these solvers, we face challenges in meeting the solving rate and are forced to reduce the MPC horizon. Reducing the MPC horizon implies that the MPC can only see a portion of the original OCP, possibly leaving out some of the tasks. This affects the predictive capability of the controller and compromises the accomplishment of the tasks, especially those that require an elaborate and dynamic maneuver. To overcome this difficulty, in this thesis, we propose to update, at each MPC iteration, the terminal cost function in the MPC with a function that encodes the part of the trajectory that remains unseen by the controller.Regarding robot design, deploying agile motions becomes difficult with existing aerial manipulators, which are generally big-size multirotor platforms with non-compliant, high-gear ratio limbs. In this thesis, we present Borinot, an open-source aerial robotic platform designed to research hybrid agile locomotion and manipulation using flight and contacts. This platform features an agile and powerful hexarotor that can be outfitted with torque-actuated limbs of diverse architecture, allowing for whole-body dynamic control. We present experiments with this robot showcasing different agile motions.In addition to the stated contributions, this thesis contributes in other areas required to operate the robot, such as a procedure for identifying the dynamical parameters based on factor-graph estimation or a hardware enhancement that allows for direct thrust control of Borinot¿s rotors.
  

Reading date: 02/05/2024

• MATSUMURA, KAZUAKI: Advancing the state of the art of directive-based programming for GPUs: runtime and compilation
  
  Author: MATSUMURA, KAZUAKI
  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: (DAC)
  Mode: Normal
  Deposit date: 20/03/2024
  Reading date: 02/05/2024
  Reading time: 11:00
  Reading place: Room E101 - Building C6 - Campus Nord - Barcelona - FIB
  Thesis director: PEÑA MONFERRER, ANTONIO JOSE
  Committee:
       PRESIDENT: CASTELLÓ GIMENO, ADRIÁN
       SECRETARI: AYGUADÉ PARRA, EDUARD
       VOCAL: EL HAJJ, IZZAT
  Thesis abstract: The rapid development in computing technology has paved the way for directive-based programming models towards a principal role in maintaining software portability of performance-critical applications. Efforts on such models involve a least engineering cost for enabling computational acceleration on multiple architectures, while programmers are only required to add meta information upon sequential code. Optimizations for obtaining the best possible efficiency, however, are often challenging. The insertions of directives by the programmer can lead to side-effects that limit the available compiler optimization possible, which could result in performance degradation. This is exacerbated when targeting asynchronous execution or multi-GPU systems, as pragmas do not automatically adapt to such mechanisms, and require expensive and time consuming code adjustment by programmers. Moreover, directive-based programming models such as OpenACC and OpenMP often prevent programmers from making additional optimizations to take advantage of the advanced architectural features of GPUs because the actual generated computation is hidden from the application developer.This dissertation explores new possibilities for optimizing directive-based code from both runtime and compilation perspectives. First, we introduce a runtime framework for OpenACC to facilitate dynamic analysis and compilation. Especially, our framework realizes automatic asynchronous execution and multi-GPU use based on the status of kernel execution and data availability while taking advantage of an on-the-fly mechanism for compilation and program optimization. We add a versatile code-translation method for multi-device utilization by which manually-optimized applications can be distributed automatically while keeping original code structure and parallelism. Second, we implement a novel flexible optimization technique that operates by inserting a code emulator phase to the tail-end of the compilation pipeline. Our tool emulates the generated code using symbolic analysis by substituting dynamic information and thus allowing for further low-level code optimizations to be applied. We implement our tool to support both CUDA and OpenACC directives as the frontend of the compilation pipeline, thus enabling low-level GPUoptimizations for OpenACC that were not previously possible. Third, we propose the use of a modern optimization technique, equality saturation, to optimize sequential code utilized in directive-based programming for GPUs. Our approach realizes less computation, less memory access, and high memory throughput simultaneously. Our fully-automated framework constructs single-assignment forms from inputs to be entirely rewritten while keeping dependencies and extracts optimal cases. Overall, we cover runtime techniques and optimization methods based on dynamic information, low-level operations, and user-level opportunities.We evaluate our proposals on the state-of-the-art GPUs and provide detailed analysis for each technique. For multi-GPU use, we show in some cases nearly linear scaling on the part of kernel execution with the NVIDIA V100 GPUs. While adaptively using multi-GPUs, the resulting performance improvements amortize the latency of GPU-to-GPU communications. Regarding low-level optimization, we demonstrate the capabilities of our tool by automating warp-level shuffle instructions that are difficult to use by even advanced GPU programmers. While evaluating our tool with a benchmark suite and complex application code, we provide a detailed study to assess the benefits of shuffle instructions across four generations of GPU architectures. Lastly, with sequential code optimization, we demonstrate a significant performance improvement on several compilers through practical benchmarks.Then, we highlight the advantages of computational reordering and emphasize the significance of memory-access order for modern GPUs.
  

Reading date: 03/05/2024

• AVTZI, STYLIANI: Hybrid diffuse optics methods to assess the emergence of dementia in older adults
  
  Author: AVTZI, STYLIANI
  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: 25/01/2024
  Reading date: 03/05/2024
  Reading time: 10:00
  Reading place: ICFO, Mediterranean Technology Park, Avinguda Carl Friedrich Gauss, 3, 08860 Castelldefels, Barcelona
  Thesis director: DURDURAN, TURGUT
  Committee:
       PRESIDENT: BUSQUETS FACIABÉN, ALBERT
       SECRETARI: ARTIGAS GARCIA, DAVID
       VOCAL: ARMENDOLA, CATERINA
  Thesis abstract: Hybrid diffuse optical devices offer a non-invasive and continuous and cost-effective method for monitoring cerebral blood flow and metabolism on the bedside use and realistic simulation applications. The incorporation of diffuse correlation spectroscopy (DCS) and near-infrared spectroscopy (NIRS) in these devices extends their versatility. This PhD project focused on utilizing diffuse optics to assess brain activity during functional and stress tests in older populations.Ageing is the primary risk factor for various brain conditions such as stroke, cognitive disorders, and mobility issues. As the population becomes increasingly older, these age-related pathologies are becoming a significant social and economic burden. The underlying assumption is that microvascular damage and changes in brain blood flow regulation contribute significantly to an increased risk of cerebrovascular diseases, cognitive and mobility disorders. This underscores the importance of creating a widely accessible monitoring system and associated protocols able to detect these changes early on, ultimately leading to personalised interventions. Two multi-disciplinary studies were performed during my doctorate studies to identify alterations in the haemodynamic parameters of older adults in response to existing pathologies.Microvascular cerebral blood flow (CBF) in a cohort of younger and older adults (>65 y.o.) with and without mild cognitive impairment (MCI) in overall good health was monitored during functional and stress tests. It was observed that CBF of older adults with MCI could not recover to baseline conditions compared to younger participants indicating possible autoregulation and vasoreactivity problems similar to those previously observed in chronic sleep apnea and chronic carotid stenosis patients. CBF measurements during functional cognitive tasks revealed gender differences. For a given test MCI participants presented a statistically higher response than normocognitive (NC) subjects. The combination of these results favour the "inefficiency hypothesis" that suggests that older adults activate the brain networks as NC individuals to cope with behavioural demands but with increased activity. A new hybrid diffuse optics device was developed combining a custom-made fast-DCS with a commercial NIRS device along with external devices for physiological signal recordings in the second study. The project aimed to measure changes in cerebral haemodynamics in older adults with Motoric Risk Syndrome (MCR) during functional cognitive and motor tasks protocols to evaluate the pre-post impact at 3 and 6 months of physical exercise alone or combined with transcranial direct current stimulation (tDC). Results revealed higher CBF but not oxy-haemoglobin (HbO2) responses in dual tasks (DT) compared to single (ST). There were no differences between groups at baseline and 3 months but statistically different responses in CBF were observed at 6 months for both intervention groups compared to the control group but not in HbO2 response, indicating that intervention affects CBF response possibly due to improvements of vascular health, highlighting the importance of physical activity and transcranial stimulation on the maintenance of vascular health. A big part of my research focused on the development of new algorithms for de-contaminating the measured data from extracerebral signal to develop an optimal model to minimise the effect for both studies. In summary this study proves the capability of hybrid optics to capture the evoked haemodynamic responses in the pre-frontal cortex and offers insights into the use of techniques to assess cognitive function in older adults, specifically those with MCI and MCR. The findings highlight the complex relationship between blood flow responses and cognitive activities suggesting that compensatory mechanisms may play a role in individuals facing cognitive challenges. Future research in these areas holds promise, for enhanc
  

• NESTOROV, ANNA MARIA: Optimizing serverless architectures for data-intensive analytics workloads
  
  Author: NESTOROV, ANNA 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 COMPUTER ARCHITECTURE
  Department: (DAC)
  Mode: Normal
  Deposit date: 25/03/2024
  Reading date: 03/05/2024
  Reading time: 10:00
  Reading place: Sala C6-E101 - Facultat d'Informàtica de Barcelona (FIB)
  Thesis director: CARRERA PÉREZ, DAVID | BERRAL GARCÍA, JOSEP LLUÍS
  Committee:
       PRESIDENT: SANCHEZ ARTIGAS, MARC
       SECRETARI: GUITART FERNANDEZ, JORDI
       VOCAL: TAHERKORDI, AMIRHOSEIN
  Thesis abstract: Recently, serverless computing has garnered attention from academia and industry due to its on-demand resource provisioning, allowing users to focus solely on their core business logic by breaking down tasks into small stateless functions.Serverless offers benefits like a 'pay-per-use' cost model, greater flexibility, and transparent elastic resource auto-scaling.Researchers in academia and industry are increasingly exploring serverless computing's potential in complex, data-intensive analytics. These tasks, resource-heavy and highly parallel, involve significant inter-function communications. However, this shift presents challenges, requiring alignment with the specific needs and constraints of such applications. This research area is currently considered one of the most compelling areas of study. This thesis shows that it is possible to efficiently execute modern data-intensive analytics workloads, traditionally deployed in managed cloud clusters, within serverless computing environments using direct data inter-function communication and optimized performance-cost efficient resource allocation policies. To demonstrate this thesis, we first build a performance model for serverless workloads, considering data sharing, volume, and communication technologies. The model, with a relative error of 5.52%, evaluates the performance of a representative workload in serverless, analyzing task granularity and concurrency, data locality, resource allocation, and scheduling policies. Our results indicate that the performance of data-intensive analytics workloads in serverless can be up to 4.32x faster depending on how these are deployed. Furthermore, this characterization highlights inefficiencies in centralized object storage and stresses the primary importance of efficient resource use.We then introduce Floki, a data forwarding system that tackles the centralized object storage bottleneck. It enables direct communication between producer-consumer function pairs using fixed-size communication methods. Floki establishes point-to-point data channels for intra- and inter-node data transmission, allowing transparent data transfer and reducing network data copying. This workflow-oriented approach boosts performance and minimize resource requirements without restricting function placement.Our experimental evaluation, performed on the principal communication patterns in distributed systems, shows that Floki reduces the end-to-end time up to 74.95x, decreasing the most extensive data sharing time from 12.55 to 4.33 minutes, saving almost two-thirds of time. Additionally, Floki achieves up to 50,738x of resource-saving, equivalent to a memory allocation of approximately 1.9MB instead of an object storage allocation of 96GB.Finally, we investigates how to achieve efficient resource utilization in modern serverless environments and proposes Dexter, a novel resource allocation manager, leveraging serverless computing elasticity. Dexter continuously monitors application execution, dynamically allocating resources at a fine-grained level combining predictive and reactive strategies to ensure performance-cost efficiency (optimizing total runtime cost). Unlike black-box Machine Learning (ML) models, Dexter reaches a sufficiently good solution, prioritizing simplicity, generality, and ease of understanding. The proposed experimental evaluation demonstrates that our solution achieves a significant cost reduction of up to 4.65x, while improving resource efficiency up to 3.50x, when compared with the default serverless Spark resource allocation that dynamically requests exponentially more executors to accommodate pending tasks. Dexter also enables substantial resource savings, demanding up to 5.71x fewer resources. Dexter is a robust solution to new, unseen workloads, achieving up to 2.72x higher performance-cost efficiency thanks to its conservative resource scaling approach.