Committee:
     PRESIDENT: MIKKOLA, AKI
     SECRETARI: MINGUELLA CANELA, JOAQUIM
     VOCAL: MORENO SASTOQUE, JUAN CAMILO
Thesis abstract: Duchenne Muscular Dystrophy (DMD) and Spinal Muscular Atrophy (SMA) are rare neuromuscular diseases characterized by progressive muscle weakness, leading to loss of ambulation, respiratory and cardiological complications, and ultimately to a premature death. Recent therapeutic advancements have introduced new possibilities for managing these disorders, focusing on preserving upper limb function to maintain patient's autonomy.Nevertheless, the clinical assessment tools rely on observational scales, which lack the sensitivity and objectivity needed to accurately track disease progression and therapeutic outcomes. In this context, the Inertial Measurement Units (IMUs), electronic devices able to quantitatively track the orientation of the body segment where they are attached, represent a valid alternative.For this reason, the main objective of this thesis is to validate the use of an IMU-based system to objectively quantify the upper limb motor function in children with neuromuscular diseases (NMDs). First, a systematic review was conducted to analyze the technical and clinical aspects of using an IMU-based system for clinical application, highlighting key challenges such as sensor-to-segment calibration, not always feasible for patients with motor impairment, and the lack of standardized biomechanical protocols. Additionally, an analysis of the biomechanical metrics generally implemented was performed, emphasizing the most meaningful to evaluate the motor status in people with motor diseases. Using an IMU-based system, it is possible to obtain well-known metrics like spatiotemporal or angular metrics. All of them have been proven to be useful for evaluating upper limb motor function in people with motor impairment. However, there is a lack of agreement regarding the clinical protocol and the biomechanical metrics that can help to better understand the progression of the diseases.To address these open challenges, an IMU-based system was developed, the ArmTracker, equipped with seven sensors to evaluate the kinematics of the upper limb in children with DMD and SMA. The first issue addressed was the sensor-to-segment calibration. A picture-based sensor-to-segment calibration was developed, allowing accurate kinematic analysis even in patients with motor impairment, such as children with NMDs. This method was validated by comparing IMU-derived motion data against the gold-standard in the motion analysis field. Furthermore, the kinematic data obtained with this method were compared against standard calibration methods, commonly implemented in clinical practice. Results, obtained in a group of healthy subjects, shown that the picture-based calibration technique led to accurate kinematic data while being more practical and feasible for clinical applications.The system was further tested in a clinical setting on a group of children with DMD and SMA. Kinematic metrics such as workspace area, curve efficiency, and range of motion were computed and compared with the results of the standard clinical scales. A strong correlation with clinical scores was found, confirming their validity for assessing upper limb motor function. Moreover, the system identified discrepancies with respect to the clinical score, suggesting that IMU-based metrics might provide additional insights beyond traditional scales. These metrics were further analyzed one year after the first assessment to assess their sensitivity in longitudinal assessment. The findings suggest that kinematic metrics, obtained with an IMU-based system, are useful not only to objectively assess upper limb motor function in children with NMDs, but may also support clinicians in conducting longitudinal evaluations.We hope that the work of this dissertation serves as a relevant guideline for future therapists and engineers in the field to enhance the understanding and the evolution of neuromuscular diseases.

Committee:
     PRESIDENT: OTTAVIANO, ERIKA
     SECRETARI: THOMAS ARROYO, FEDERICO
     VOCAL: ROJAS LIBREROS, NICOLAS ENRIQUE
Thesis abstract: Graceful motion can be loosely defined as the one we observe in natural movements executed by animals and humans, which are characterized by being agile, efficient, and fluid. The generation of graceful robot motions is typically sought through the minimization of cost functions involving not only path length, but also aspects related to smoothness, like the time derivative of acceleration, called jerk, or that of the controls. A widely used approach to compute optimal trajectories is through direct collocation, a technique that converts the continuous-time optimal control problem into a finite-dimensional NLP problem. Collocation proceeds by discretizing the trajectory using so-called collocation points, and imposing the dynamics constraints at such points. The formulation of most collocation methods, however, assumes that the system is governed by a first order ODE, whereas robotic systems are typically described by second or higher order ODEs. As a result, the usual practice is to initially convert those ODEs into first order form via introducing new variables, and adding new equations that link these variables with their integral counterparts. An often overlooked effect of this transformation is that it generates inconsistencies between the trajectory of each variable and that of its time derivative. This is so because a collocation method only imposes the differential relationships at the collocation points, but not in between such points. A closer examination of this effect reveals that the dynamic equations, which should be satisfied in the collocation points, are actually violated in them, despite apparently having been enforced. This thesis introduces new collocation methods designed to overcome these problems. Specifically, we develop improved versions of the most popular piecewise and pseudospectral collocation schemes, including the trapezoidal and Hermite-Simpson methods, as well as the LG, LGR, and LGL methods. The new algorithms are able to treat differential equations of arbitrary order M ≥ 1 without having to convert them into first order form. In all of them, the trajectory obtained for each variable coincides exactly with the time derivative of its corresponding integral variable, and the dynamic constraints are satisfied accurately at the collocation points. These properties allow a drastic reduction of the dynamics error of the obtained trajectories in many cases, meaning that the governing equations are better respected along the continuous time horizon of the problem. Our methods also provide trajectories that are smoother than those of conventional ones, and easily treat variables such as jerk or the time derivative of the controls in the cost function. An hp adaptive refinement algorithm is also proposed to combine the benefits of our piecewise and pseudospectral methods so as to speed up convergence to the solutions.

Committee:
     PRESIDENT: CUETO PRENDES, ELIAS
     SECRETARI: LLOBERAS VALLS, ORIOL
     VOCAL: ROMERO OLLEROS, IGNACIO
Thesis abstract: This thesis explores the possibilities and benefits of computational tools to be used in the analysis and design of shock absorbing metamaterials for industrial applications. The research is organized into two complementary chapters. The first part focuses on the computational modeling of shock propagation in hyperelastic materials, emphasizing its importance in analyzing and designing mechanical energy-absorbing materials. This approach is categorized as shock-capturing, as it aims to capture the effects of shock formation and propagation on the mechanical response of the solid. Numerical simulation of the onset and propagation of discontinuous strain waves (mechanical shocks) across a solid is performed in that specific approach, and it is computationally assessed in different loading situations. The concept of extrinsic dissipation, emerging from the breaking of polyconvexity constraints in hyperelastic materials, is recalled and exemplified via a large strain neo-Hookean polyconvex hyperelastic model which is perturbed in a number of different formats. The obtained mechanical responses, exhibiting propagating mechanical shocks, and the corresponding extrinsic dissipation, are explored together with the inherited computational issues, by means of representative numerical examples. Additionally, the problem of the dynamic impact of a rigid solid on low-density shock-absorbing specimens made of these perturbed hyperelastic materials is analyzed, and the amount and time-evolution of the resulting extrinsic dissipation are evaluated to characterize their performance for shock-absorbing purposes. Furthermore, the exhibited property of shape recovery after impact is highlighted as a distinctive feature in front of alternative shock-absorbing materials. The second part of the thesis explores the potential of computational multiscale methods, in conjunction with massive buckling-lattice structures at the metamaterial core (meso/micro scale), to provide a suitable framework for designing shock-absorbing metamaterials aimed at industrial applications. In this context, a prospective computational setting is considered under the hypothesis that, for a sufficiently complex microlattice topology, some localized regions might buckle within the lattice-structure core and propagate through it, giving rise to different loading-unloading paths, in such a way that hysteretic-like structural behaviors would take place, thus arising dissipative behaviors, even if the base material at the buckling micro-lattice behaves in a hyperelastic (thus intrinsically non-dissipative) manner. Using the standard Hill-Mandel homogenization principle, and assuming that the necessary separation of scales holds, the homogenized body, now living in a classical solid-mechanics setting, displays a homogenized non-convex behavior which exhibits extrinsic dissipation and, thus, could be potentially used (at reduced computational cost) for shock absorbing metamaterials analysis and design purposes. A tentative industrial application, to a sneaker’s insole design, has been then considered as a work’s target for evaluation of the room offered by the explored setting in the context of shock-absorbing metamaterial design. Finally, remarks on the scope and limitations of the work, and its significance for further advances in the field are emphasized.

Committee:
     PRESIDENT: MARTINCIC, EMILE
     SECRETARI: RODRÍGUEZ VILLARREAL, ÁNGELES IVÓN
     VOCAL: RAMON GARCIA, ELOI
Thesis abstract: This thesis began with the goal of revolutionizing continuous monitoring through the use of sweat. Over the course of this research, the field of sweat monitoring has evolved significantly, being consolidated as the most convenient non-invasive alternative to blood analysis. This thesis compiled the knowledge acquired during these years across different areas related to the design of a microfluidic sweat platform.Initially, a perspiration model was developed as an in vitro experimental setup to enable realistic and cost-effective testing of microfluidic sweat platforms. Sweat gland dimensions and structure were replicated using on-plane soft lithography, solving the fabrication challenges associated. A functional prototype was successfully operated with appropriate skin surface characteristics, demonstrating its use with adhered microfluidic devices. A microfluidic platform integrating custom electrochemical sensors was developed for the measurement of sweat volume, sweat conductivity and sweat lactate. Continuous monitoring has been proven for extended periods of time (exceeding 1 hour) with robustness and reproducibility. This system was incorporated into a complete wearable device featuring remote communication and embedded computation, becoming a novel tool for sports monitoring. The device was validated in on-body trials to assess the accuracy of the output parameters to the final user, for both sweat-based variables (sweat loss, sodium and sodium loss) and blood bioequivalence predictions (lactate).Additionally, a novel microfluidic design for the chrono-sampling of sweat was explored in vitro. It combines a passive valve and pump to renovate the volume in a sensing chamber in a cyclic manner. The mismatch between the absorption rate of the passive pump and the sweat generation at the inlet was found to disconnect the fluid, restoring the passive valve function once surpassed. The system's characteristics can be adjusted to different requirements by modifying the valve widths or passive pump properties. However, the fabrication of this design requires high resolution, which can limit throughput and complicating the current large human testing involved in a microfluidic sweat platform. Therefore, new materials and fabrication techniques at the laboratory were explored to reduce the design to test time and enhance production capacity. In summary, the initial vision of a novel microfluidic sweat wearable platform capable of providing non-invasive, real-time measurements of a variety of biomarkers has been accomplished. Moreover, several promising research lines have emerged, including advancements in perspiration models, microfluidic design, fabrication scalability and algorithm development. Finally, this research has led to the development of a commercially available product, demonstrating that continuous monitoring through sweat is now a market-ready technology.

Committee:
     PRESIDENT: TRAVIESO RODRIGUEZ, JOSE ANTONIO
     SECRETARI: GÓMEZ GRAS, GIOVANNI
     VOCAL: MARKOPOULOS, ANGELOS
Thesis abstract: Abrasive machining processes are usually applied after turning or milling processes in order to improve dimensional accuracy and to reduce surface roughness of the machined parts. The sound signal contains information about machining processes, and it is possible to use it to monitor and control them, by analysis of received data directly from sound signal during ongoing machining. In this thesis, honing and tangential grinding tests were performed, and the acoustic emission signal was measured and examined. Experimental tests were carried out in order to verify the correlation of the selected machining parameters with the obtained results, regarding surface roughness parameters and shape errors of machined workpiece. In the tests, the acoustic emission signal was recorded, in order to check the possibility to implement to supervising of conducting of abrasive machining in workshop practice. To collect acoustic signals, a unidirectional condenser microphone was employed, and Matlab and Audacity software was used.In honing processes, different experiments were carried out. Rotation speed of the workpiece and feed speed values of the honing head were varied, together with the grain size and density of the honing stones, as well as the pressure value of the whetstones to the honed surface. Selected responses were surface roughness, cylindricity error and material removal rate. As for tangential grinding, three parameters were varied: depth of cut, feed rate and transverse infeed. Responses of the grinding process were surface roughness and material removal rate.In both cases, honing and grinding tests, linear regression models were searched for roughness and for material removal rate, and multiobjective optimization was carried out by means of the Derringer-Suich desirability function.It is possible to supervise the abrasive machining process, both grinding and honing, while the machining is in progress and without interrupting it. The proposed method consists of carrying out machining tests, measuring several responses such as roughness or material removal rate and supervising the received acoustic signals. Afterward, it will be possible to carry out comparative analysis of the acoustic signal on an ongoing basis registered.It was confirmed that it is possible to predict the result parameters of the conducted abrasive machining (including surface roughness profile parameters) by analyzing the recorded acoustic signal.