We propose a method to locate the current in a printed circuit board caused by an electrostatic discharge. This approach uses data acquired before the discharge event by using a vector network analyzer. The localization is performed by using single-shot time-domain data and the time-reversal algorithm. The time-reversal backpropagation uses experimental data, eliminating the need for numerical models of the device under test. The method allows us to distinguish two traces placed 8 mm apart.

In a previous paper [Opt. Lett. 49, 1820 (2024)], a prime symbol was missing in the argument of the current density.

Files related to publications under FNS grant 200020_204072: (1) Zviagin, A.; Kopysov, V.; Kozlovskii, V.; Chernyshev, D.; Makarov, A.; Boyarkin, O. V. Identification of Isomeric Metabolites Using Cold Ion Spectroscopy Add-on to Orbitrap-Based Mass Spectrometer, Anal Chem 2025, 97, 14097-14102. (2) Zviagin, A.; Boyarkin, O. V. Ion Spectroscopy Reveals Structural Difference for Proteins Microhydrated by Retention and Condensation of Water, J Phys Chem A 2024, 128, 2317-2322. (3) Kopysov, V.; Yamaletdinov, R.; Boyarkin, O. V. Oligomers of diphenylalanine examined using cold ion spectroscopy and neural network-based conformational search, Phys Chem Chem Phys 2024, 26, 27964-27971. (4) Kopysov, V.; Yamaletdinov, R.; Boyarkin, O. V. Quantification of enantiomers and blind identification of erythro-sphingosine non-racemates by cold ion spectroscopy, Analyst 2024, 149, 4600-4604. (5) Switzerland, 2024. (6) Zviagin, A.; Yamaletdinov, R.; Nagornova, N.; Domer, M.; Boyarkin, O. V. Revealing the Structure of Tryptophan in Microhydrated Complexes by Cold Ion Spectroscopy, J Phys Chem Lett 2023, 14, 6037-6042. (7) Zviagin, A.; Cimas, A.; Gaigeot, M. P.; Boyarkin, O. V. Salt Bridge Structure of Microhydrated Arginine Kinetically Trapped in the Gas Phase by Evaporative Cooling, J Phys Chem A 2023, 127, 4832-4837. (8) Zviagin, A.; Kopysov, V.; Nagornova, N. S.; Boyarkin, O. V. Tracking local and global structural changes in a protein by cold ion spectroscopy, Phys Chem Chem Phys 2022, 24, 8158-8165. (9) Zviagin, A.; Kopysov, V.; Boyarkin, O. V. Gentle nano-electrospray ion source for reliable and efficient generation of microsolvated ions, Rev Sci Instrum 2022, 93, 114104. (10) Saparbaev, E.; Zviagin, A.; Boyarkin, O. V. Identification of Isomeric Biomolecules by Infrared Spectroscopy of Solvent-Tagged Ions, Anal. Chem. 2022, 94, 9514-9518.

We introduce a framework for describing the real-time dynamics of quantum impurity models out of equilibrium which is based on the influence matrix approach. By replacing the dynamical map of a large fermionic quantum environment with an effective semigroup influence matrix (SGIM) which acts on a reduced auxiliary space, we overcome the limitations of previous proposals, achieving high accuracy at long evolution times. This SGIM corresponds to a uniform matrix-product state representation of the influence matrix and can be obtained by an efficient algorithm presented in this Letter. We benchmark this approach by computing the spectral function of the single impurity Anderson model with high resolution. Further, the spectrum of the effective dynamical map allows us to obtain relaxation rates of the impurity toward equilibrium following a quantum quench. Finally, for a quantum impurity model with on-site two-fermion loss, we compute the spectral function and confirm the emergence of Kondo physics at large loss rates.

Modern engineered systems are increasingly complex, comprising interconnected components within hierarchical structures that interact dynamically under operational and environmental influences. As these systems degrade, performance and reliability decline, with failures potentially causing severe economic and safety impacts. Robust algorithms are therefore essential for monitoring wear and performance in such environments. Machine learning has driven advances in data-driven for Prognostics and Health Management (PHM) approaches, which detect anomalies, diagnose degradation and faults, and predict their evolution from condition monitoring data. However, these methods often struggle to capture nonlinear, evolving relationships and lack robustness under novel conditions. Many focus narrowly on temporal dynamics or static interdependencies, overlooking evolving inter-sensor relationships indicative of incipient faults. Without modeling exogenous influences, they suffer high false alarm rates, leading to alarm fatigue. Accurate degradation monitoring requires understanding how operational history drives progression and how accumulated damage affects performance. Most methods model only operational dynamics, inferring latent dynamics at a single timescale and missing the coupling of slow degradation with fast dynamics. This results in noisy degradation estimates, limiting subsequent prognostic accuracy. Furthermore, operational loads, which significantly affect degradation, are often impractical to measure directly. Virtual sensing can estimate these loads from condition monitoring data, but the diversity of sensors, with varied signal characteristics and sampling frequencies, complicates integration. Existing methods typically handle only a single sensor modality, struggle to fuse heterogeneous data, and generalize poorly to extreme or underrepresented conditions. To tackle these challenges, this thesis proposes a unified framework with three interconnected modules that capture interdependencies and dynamic behaviors across system levels. Designed for hierarchical, interconnected systems, the modules work independently or together for comprehensive monitoring and decision-making. By leveraging temporal and relational learning, the framework effectively models dynamic interactions, integrates heterogeneous data, and captures multiscale behaviors, enabling robust fault detection, reliable load estimation in underrepresented conditions, and accurate degradation inference. Specifically, the framework includes: (1) a module that dynamically captures evolving inter-sensor relationships using attention-based inference in temporal graph neural networks and incorporates operational context into node dynamics, enabling early and robust fault detection; (2) a module that fuses multirate data and models intra- and inter-modality interactions through a heterogeneous temporal graph with modality-specific, condition-aware encoders for reliable virtual load sensing; and (3) a module that disentangles long-range, slow-fast dynamics using a hierarchical differential model with monotonicity constraints, providing accurate degradation inference and supporting proactive control. The framework demonstrates effectiveness and robustness through extensive evaluations on both simulated and real-world systems for condition monitoring and degradation inference. Primarily developed for industrial systems, it has also shown potential in infrastructure health monitoring

This cumulative thesis addresses the environmental burden associated with single-use and disposable sensors, particularly in point-of-care and smart packaging applications. It presents the development of a sustainable wireless sensor system that integrates experimental sensor development with environmental assessment. The result is a comprehensive approach combining sustainable material deployment, sensor system design, and Life Cycle Assessment (LCA) - addressing sustainability from component to system level. Particular focus is placed on the systematic investigation of the sensing behavior of chitosan films under controlled exposure to water vapor and VOCs. The results show that chitosan ex-hibits a dominant response to humidity compared to its response to VOCs. This work presents the first fully additively manufactured chitosan-based humidity sensor, demonstrates direct laser carbonization of chitosan films, and identifies proton hopping as the dominant, oxygen-independent conduction mechanism. While VOC selectivity was shown to be limited, the findings enable a critical reflection on chitosan's real-world limitations in printed electronics, including encapsulation and substrate compatibility. Building on this, a printed hybrid near-field communication (NFC)-based sensor tag was developed for wireless monitoring of acetone vapor and humidity. The system comprises a chitosan-based resistive sensor, a printed antenna, and a commercial NFC chip, all integrated onto a single bio-based substrate. The system features a room-temperature flip-chip bonding process and demonstrates a feasible manufacturing pathway for hybrid sensor tags on thermally sensitive substrates. The compact, energy-autonomous design has practical potential for point-of-care diagnostics and smart packaging applications. To critically assess the environmental impacts of such systems, a full LCA with the focus on carbon footprinting was conducted for representative printed and hybrid sensor configurations. Contrary to expectations, despite being the dominant mass fraction, the substrate showed relatively low environmental impact, whereas the sensor chip and silver nanoparticle-based inks were identified as major environmental hotspots. Furthermore, challenges related to end-of-life treatment and the lack of recycling infrastructure for hybrid systems were highlighted. The study emphasizes the importance of early-stage sustainability considerations in printed sensor design and provides guidance for informed material and process selection. By combining experimental, system-level, and environmental perspectives, this work contributes to the advancement of sustainable printed electronics and identifies key challenges and opportunities for future development.

This thesis proposes and analyzes randomized algorithms for solving large-scale numerical linear algebra problems. A major theme of the thesis is to extend existing randomized low-rank approximation techniques to develop fast and reliable algorithms for parameter-dependent and time-dependent problems.

In Chapter 2, we present subspace embedding properties of random matrices with Khatri-Rao product structures. These properties are fundamental for analyzing randomized numerical linear algebra algorithms. We also propose two eigenvalue solvers that leverage the structure of Khatri-Rao product random matrices, enabling us to solve eigenvalue problems efficiently and with low memory usage.

In Chapter 4, we extend randomized low-rank approximation algorithms to compress parameter-dependent matrices. We propose parameter-dependent versions of the randomized SVD and the generalized Nyström method. In particular, our algorithms do not resample the random matrix for each parameter, making them computationally attractive. Based on the theoretical guarantees presented in Chapter 3, we show that our parameter-dependent algorithms provide sub-optimal approximation guarantees. We also present numerical findings to illustrate these results.

In Chapter 5, we combine randomized low-rank approximation with time-integration schemes to obtain a low-rank approximation of the solutions of matrix differential equations. In contrast to existing methods, which typically rely on tangent space projections, our method constructs a low-rank approximation from random sketches of the discretized dynamics, making it simpler and more robust when the dynamics deviate from the tangent space. We provide error bounds and numerical results demonstrating that our method is efficient and accurate with high probability.

In Chapter 6, we analyze the use of the discrete empirical interpolation method to accelerate the evaluation of the tangent space projector in dynamical low-rank approximation when approximating solutions of matrix differential equations. Using the theory of differential inclusions, we provide guarantees on the approximation error with respect to the true solution in the continuous-time setting. We also propose a practical numerical scheme for time integration, supported by both theoretical analysis and numerical experiments showing its effectiveness.

Aging leads to progressive changes in brain metabolism, structure, and function, resulting in declines in motor performance. These changes include reduced cortical inhibition, impaired brain structure, and altered functional connectivity. Consequently, older adults often experience diminished motor ability, postural instability, and increased fall risk, which can negatively affect their quality of life. However, emerging evidence shows that motor learning, such as balance and strength training, can promote neuroplasticity and help mitigate age-related decline. The neural mechanisms supporting motor ability and the brain's adaptations to motor learning remain incompletely understood, particularly regarding the interplay of neurochemical, structural, and functional changes.

This thesis explores the brain mechanisms underlying motor function and the neuroplastic changes associated with motor learning in older adults. The findings are based on data from a longitudinal study in which older adults were assigned to balance training, strength training, or control groups for three months. Participants underwent multimodal neuroimaging before and after the intervention. This design allowed us to assess large-scale brain networks involved in balance and examine training-induced metabolic, structural, and functional changes.

To understand the neural basis of balance, we used connectome-based predictive modeling with baseline data to identify structural and functional brain networks associated with balance performance. Our results show that both structural and functional connectomes predict balance ability. Key networks include motor-subcortical, fronto-parietal, and visual circuits, emphasizing the importance of coordinated activity across these systems in supporting postural control in older adults.

We then examined how balance training drives neural plasticity. Following three months of training, participants showed improved balance, enhanced functional connectivity within sensorimotor networks, and increased cortical inhibition, evidenced by elevated GABA levels in the sensorimotor cortex. These findings suggest that balance training can counteract age-related reductions in connectivity and inhibitory function, underscoring the role of the GABAergic system in neuroplastic adaptation and motor behavior.

While functional changes are important for adaptive motor control, structural changes are also essential. Our third study investigated structural plasticity following balance and strength training. We found that strength training promotes white matter reorganization in key pathways and that both training types counteract age-related structural degeneration in critical brain regions. These results highlight the capacity of motor training to induce beneficial neural plasticity at multiple levels.

In conclusion, this thesis provides evidence that motor learning can drive neuroplastic changes across metabolic, structural, and functional domains in the aging brain. By integrating results from various neuroimaging techniques, this work offers a comprehensive view of the neural basis of motor function and the brain's capacity for adaptation through targeted training in older adults.

As photovoltaics (PV) scale rapidly to meet climate and energy goals, ensuring long-term reliability and performance is increasingly important. With global installed PV capacity surpassing the terawatt scale, and expected to continue to grow exponentially in the coming years, even modest improvements in system lifetime, monitoring, and degradation understanding can yield major environmental and economic gains. At this scale, where millions of systems must operate reliably for decades, standardised diagnostics and fleet-wide assessment frameworks are essential. This thesis addresses the challenge of quantifying PV degradation in real-world conditions by developing and validating a set of data-driven methods that improve both long-term performance loss assessments and short-term fault detection. Though demonstrated mainly on Swiss systems, the approaches are broadly applicable across climates, technologies, and deployment contexts, offering tools for more accurate benchmarking, maintenance, and optimisation as PV continues its expansion in the terawatt age.

The first part focuses on improving long-term performance assessments while reducing uncertainty. A key contribution is the novel multi-annual Year-on-Year (multi-YoY) method, which increases comparison depth and reduces statistical noise in performance loss rate (PLR) estimates. Applied to synthetic and real-world datasets, it lowers statistical uncertainty by over 90% compared to conventional methods, enabling clearer benchmarking across systems of varying age, quality, and environmental exposure.

The second part addresses short-term variability and its impact on long-term diagnostics. It introduces a fault detection and diagnosis algorithm (FDDA) that identifies reversible loss conditions such as snow, shading, and inverter downtime, based on deviations from expected output. Applied to over 300 PV strings, the FDDA classifies six fault types at 15-minute resolution and enables the definition of intrinsic PLR (i-PLR), which excludes temporary faults and isolates true degradation. In one case, fault filtering reduced the apparent degradation by 80%, linked to a threefold increase in partial shading within a few years due to a growing tree. Across the fleet of PV strings, i-PLR was significantly lower in fault-prone systems, and differences correlated strongly with the fault time factor, a new metric quantifying fault occurrence. Overall, the FDDA reveals how transient faults can distort PLR trends.

The final part translates the research into an integrated monitoring platform for PV operators, developed and deployed in collaboration with 3S Swiss Solar Solutions. The platform combines fault detection and long-term analytics into a scalable, modular tool that supports automated diagnostics and predictive maintenance. Real-world case studies include inverter failure detection, soiling loss tracking and recovery leading to 25% yield gains, and shading impact quantification.

Overall, this thesis bridges the gap between short-term fault detection and long-term degradation analysis. The tools and methods proposed enable more accurate health assessments, improve maintenance and warranty planning, and support the design of more reliable PV systems. These contributions help advance reliable, data-driven PV operation at terawatt scale, supporting global energy and climate goals.