Publikationen

Browse

Now showing 1 - 20 of 5418

Differentiable cost model for neural-network accelerator regarding memory hierarchy
(Universität Ulm, 2024-10-18) Conrad, Joschua; Wilhelmstätter, Simon; Asthana, Rohan; Belagiannis, Vasileios; Ortmanns, Maurits
Dedicated neural-network inference-processors improve latency and power of the computing devices. They use custom memory hierarchies that take into account the flow of operators present in neural networks and convolutional layers. For efficient implementation, such network topologies can greatly benefit from hardware-cost optimization using automated network-architecture search. Thereby, cost functions predict the suitability of a network topology for a given type of inference hardware. A differentiable neural-architecture search that optimizes both weights and topology in a single training requires cost models to be differentiable in the dimensions of weight and activation matrices. State-of-the-art differentiable cost models require time-consuming system-level measurements or simulation results, or do not encounter the hardware structure at all. This work presents a simple yet effective procedure for deriving a differentiable neural-network-accelerator cost-model that is suitable for any type of accelerator. It is based on hardware-independent parameterization and a novel differentiable divide-ceil function, as well as hardware-specific modeling. The resulting differentiable model can be reconfigured to the actual hardware size and memory structure to predict the inference energy for an exact network topology. The modeling and prediction are demonstrated for a state-of-the-art SRAM-based inference-accelerator and for the Eyeriss accelerator, inferring different state-of-the-art neural networks, resulting in excellent agreement with measured hardware.
ΔΣ modulators employing MASH DSM DAC-based dual quantization
(Universität Ulm, 2024-08-30) Abdelaal, Ahmed; Pietzko, Michael; Kauffman, John G.; Jain, Ankesh; Ortmanns, Maurits
Multibit (MB) quantization allows high resolution Delta-Sigma modulators (DSMs) with low oversampling ratio (OSR). Furthermore, it allows higher maximum stable amplitude (MSA), achieves reduced jitter sensitivity, and relaxes the dynamic requirements on the DSM loop-filter (LF). However, MB quantization adds a dominant source of non-linearity due to element mismatch in the MB digital-to-analog converter (DAC) which often dominates the performance. State of the art (SoA) presents many calibration techniques, though digital power and area consumption can be high and calibration time be significant. In this paper we target a calibration-free DAC based on dual quantization and propose to employ a Multistage noise SHaping (MASH) Digital-DSM (DDSM) to avoid architectural compromises between the main DSM LF and the DDSM DAC. The implementation trade-offs are illustrated, and stability constraints in both the main LF and the DDSM are addressed. An exemplary implementation is derived and simulated, and the results shall lay the foundation for future circuit implementations of MASH DDSM to realize MB DSM with intrinsically high linearity.
Application of STED super-resolution optical microscopy in biomedical studies
(Universität Ulm, 2026-02-17) Wondany, Fanny; Michaelis, Jens; Barth, Holger
Super-Resolution Microscopy allows to resolve fluorescently labeled structures at the nanometer scale and has revolutionized fluorescence microscopy. The first diffraction-unlimited optical microscopy technique presented was Stimulated Emission Depletion (STED) super-resolution optical microscopy. In STED microscopy, the resolution limit of a laser scanning confocal microscope is tackled by reducing the effective Point Spread Function (ePSF) by switching fluorescing areas into different states. Hence, previously unattainable scales below 250 nm are now observable and STED microscopy can be applied to any field of modern (life) sciences. In this regard, I utilized STED microscopy to capture super-resolved images and, based on these images, I conducted quantification analyses as well as interpretation. Thereby, I enlightened various biomedical open questions concerning (i) the use of Virus-Imprinted Polymers (VIPs) for selective virus recognition, (ii) the application of Antimicrobial Peptides (AMPs) in the context of tuberculosis treatment and (iii) the mode of action of bacterial AB-type exotoxins. Virus-Imprinted Polymers (VIPs) are promising virus-capturing candidates for manifold areas of application, e.g. laboratory-scale virus purification or decontamination scenarios. Using STED microscopy, ex cellula virus imaging at VIPs allows clear sub-diffraction image captures as no cellular background disturbs the measurement. We were the first to provide visible proof that target virus selectively rebound to VIPs, whereas off-target virus did not. Not only viruses but also other health-relevant species, such as Mycobacterium tuberculosis (Mtb), demand attention. Mtb causes a life-threatening airborne infectious disease, namely tuberculosis. In the treatment of tuberculosis, the use of endogenous Antimicrobial Peptides (AMPs), such as LL-37 or Gran1, constitute a promising additive. But the underlying AMP–pathogen interplay remains unclear. Therefore, STED super-resolution microscopy was utilized to monitor the localization of respective AMPs at Mtb extra– and intracellularly. For sufficient antimicrobial activity, relatively high concentrations (µM range) of AMPs must be applied. To circumvent this, targeted delivery of AMPs across biological barriers can be considered. However, targeted delivery of peptides across biological barriers is still a major challenge. Therefore, the natural ability of bacterial AB-type exotoxins to deliver their toxic subunit across biological barriers into cells provides an excellent starting point for (i) investigating the underlying mode of action and (ii) generating non-toxic modular delivery systems. In this context, STED microscopy was used to observe two selected exotoxins in terms of cell binding and internalization: C3 protein exotoxin from Clostridium botulinum (C3bot) and Diphtheria Toxin (DT). With respect to C3bot, we identified a novel type of target cells. Furthermore, the functionality of a non-toxic C3bot variant was extended to attach desired cargo molecules using click chemistry. For DT, mutational analysis and subsequent sub-diffraction studies using STED super-resolution optical microscopy extended the current knowledge of the functions of individual domains. All together, this cumulative thesis at hand emphasizes the broad biomedical applicability of STED microscopy and its value to get new insights.
Hybrid fishnet metamaterial based on liquid crystal and P-I-N diodes for reconfigurable transmitarrays at 10 GHz
(Universität Ulm, 2024-12-27) Diepolder, Adrian; Schmidt, Lukas; Brandl, Susanne; Hinz, Philipp; Waldschmidt, Christian; Damm, Christian
This article presents a reconfigurable unit cell for transmitarrays at 10GHz based on the fishnet metamaterial structure. The proposed design achieves a phase shift of up to 180∘ between different states and additional tunability in each phase state by switching the electric response using p-i-n diodes and simultaneously tuning the magnetic response by employing liquid crystal. Thus, it offers the potential to overcome the limitations of switchable unit cells with few discrete phase states in terms of achievable antenna gain and side lobe level. With an insertion loss of 1.8–2.5dB in combination with its high phase tunability it exhibits much higher performance than designs using only liquid crystal. A lumped-element model for prediction of the unit cell behavior is proposed, allowing to reduce the number of full-wave simulations during performance analysis. The biasing network for the liquid crystal is designed for integration in an active-matrix configuration, greatly reducing the complexity of the control circuit. To verify the proposed design, a single unit cell prototype is manufactured and measured inside a waveguide simulator, demonstrating good agreement with full-wave simulations.
Integration of security service functions into network-level access control
(Universität Ulm, 2024-12-25) Bradatsch, Leonard; Kargl, Frank
Service function chaining is an approach to dynamically steer traffic through different service functions like intrusion prevention systems within a local area network. Existing approaches to determining the set of service functions through which specific traffic is steered are relatively coarse-grained. In this article, which focuses on security-related service functions, we present a more fine-grained determination process by integrating security service functions into attribute-based access control and utilizing contextual information attributes, such as access time. By mapping attributes to security service functions, we aim to achieve four key objectives: 1) Minimizing false negative access decisions, 2) minimizing false positive access decisions, 3) enhancing network performance by optimizing the application of security service functions, and 4) ensuring network visibility. The paper includes a detailed list of available security service functions and the security actions each can perform based on a comprehensive literature review. It also explains how attributes can be mapped to security service functions to determine when and which security service function needs to be applied to network traffic. The paper also includes detailed use cases to demonstrate the practical implementation of our approach. In the evaluation of these use cases we achieved an accuracy improvement of up to 16% compared to a standard Zero Trust approach that does not integrate traffic classification into access control. Additionally, we reduced false negatives by as much as 93% and false positives by up to 100%. The network performance was enhanced by decreasing service access times by up to 29% and increasing the number of accesses per second by up to 40% during high concurrency.
Controlling few-body reaction pathways using a Feshbach resonance
(Universität Ulm, 2025-01-09) Haze, Shinsuke; Li, Jing-Lun; Dorer, Dominik; D’Incao, José P.; Julienne, Paul S.; Tiemann, Eberhard; Deiß, Markus; Hecker Denschlag, Johannes
Gaining control over chemical reactions at the quantum level is a central goal of cold and ultracold chemistry. Here we demonstrate a method for coherently steering the reaction flux across different product spin channels for a three-body recombination process in a cloud of trapped cold atoms. We use a magnetically tunable Feshbach resonance to admix, in a controlled way, a specific spin state to the reacting collision complex. This allows us to control the reaction flux into the admixed spin channel, which can be used to alter the reaction products. We also investigate the influence of an Efimov resonance on the reaction dynamics, observing a global enhancement of three-body recombination without favouring particular reaction channels. Our control scheme can be extended to other reaction processes and could be combined with other methods, such as quantum interference of reaction paths, to achieve further tuning capabilities of few-body reactions.
Josephson-photonics devices as stable sources and detectors for microwave photons
(Universität Ulm, 2026-02-16) Danner, Lukas; Kubala, Björn; Freyberger, Matthias
At its heart, a Josephson-photonics device combines macroscopic superconducting electronics and microscopic quantum optics. It consists of a dc-voltage biased Josephson junction that is connected in series to one or a few microwave cavities. Cooper pairs can tunnel incoherently across the junction by depositing their energy 2eVdc in the microwave cavities, creating photonic excitations in them. Many parameters, such as the eigenfrequencies ∝ (LC)^(−1/2) or zero-point fluctuations ∝ (L/C)^(1/4) of the microwave cavities, can be experimentally designed; other parameters, such as the strength of the Josephson nonlinearity and the dc voltage are easily modified in situ. In this way, Josephson-photonics devices constitute a versatile source of microwave photons with a wide variety of features extending from the classical to the deep quantum regime. To make them of use for quantum technological applications that rely on a well-defined phase of the microwave light, a crucial drawback must be overcome. The dc voltage does not provide a stable reference phase to Josephson-photonics devices, as it is subject to experimental noise, such that all phase information is lost. In this thesis, we develop the field of Josephson photonics in two directions: As first main result, we resolve the issue of phase instability in Josephson photonics. The phase can be stabilized either by injection locking to a small external signal with a stable and well-defined reference phase or by mutual synchronization to a second device. An alternative approach mitigates the issue for some use cases by optimizing an ansatz to reconstruct the undiffused stationary quantum state from experiments with phase diffusion. Secondly, we show how Josephson-photonics devices can be used for the challenging task of reliably detecting single itinerant microwave photons. The “Inelastic Cooper Pair Tunneling Photon Multiplier” exploits a resonance condition, where the energy provided by the dc voltage enables a multiplication process from one incoming to n outgoing photons. This process can be staged to reach a large number of outgoing photons. Here, we model itinerant Gaussian single-photon pulses of length T with Mølmer’s approach. Further, we define a detection scheme based on heterodyne quadrature measurement of the output signal. Most importantly, we show that realistic devices can achieve a detection efficiency of 84.5% with a small dark-count rate of 10^(−3)/T, promising to outperform competing schemes.
The use of quantum resource theories in the study of nanoscale systems
(Universität Ulm, 2026-02-16) Spaventa, Giovanni; Plenio, Martin B.; Audenaert, Koenraad
The thermodynamical behavior of quantum systems at the nanoscale is often very difficult to capture via standard frameworks of thermodynamics. The latter is, in fact, an equilibrium theory whose predictions are typically only valid on average, and in regimes where systems are composed of a very large number of constituents. In reality, the thermodynamical behavior of nanoscale systems can be much more constrained that what is predicted by macroscopic thermodynamics, due to the unavoidable impact of fluctuations. As a result, the fundamental limitations to a system's dynamics induced by macroscopic thermodynamics might turn out to be considerably less stringent than what can actually be observed at those scales. The theory of open quantum systems is certainly a powerful tool in this context, and can provide solutions which go beyond equilibrium treatments, but solving the dynamics of an open system requires detailed information about its microscopic structure (e.g. system-environment interactions and coupling strengths). In other words, one of the great advantages of macroscopic thermodynamics, i.e. its universality and independence of microscopic details, might be essentially lost if we can only infer thermodynamical results once we specify the full microscopic structure of our system-environment interaction. This highlights the necessity for a framework of thermodynamics that is well suited for studying the behavior of small quantum systems far from equilibrium, but can still be applied to physical systems in a universal way, i.e., without having to specify full microscopic information. Quantum Resource Theories (QRTs), and in particular the formulations of quantum thermodynamics that have emerged from them, constitute a promising tool in this direction. However, these models of thermodynamics have not yet been extensively applied to realistic physical platform, and their consequences on known thermodynamical processes are still unexplored. Furthermore, going beyond a purely thermodynamical context, a case can be made for the use of general QRTs as benchmarking tools for the certification of certain quantum resources. In other words, QRTs could be used to find fundamental bounds to physical quantities or impossibility results (in the form of no-go theorems) that, when compared to experimental data, can be used to falsify specific hypotheses. In this thesis, both these directions will be explored, by considering resource-theoretical questions that are relevant for different physical platforms. These will include the photochemical process of photoisomerization, heat transport in harmonic systems, and tabletop tests of quantum gravity.
Histogram-Based Analysis of UAV-SAR Data for Agricultural Vegetation Classification
(Universität Ulm, 2025-12-30) Bormuth, Frederik; Riekenbrauck, Ron; Kanz, Julian; Schmidt, Daniel; Fischer, Robert F.H.; Krieger, Gerhard; Waldschmidt, Christian; Damm, Christian
Climate change poses new challenges and efficiency requirements for the agricultural sector. This necessitates the coordination of fertilization and irrigation efforts by monitoring crops. Precision agriculture aims to achieve this using modern technology including uncrewed aerial vehicles (UAVs). This paper shows a histogram-based approach to distinguish between crops by evaluating data of a high-performance UAV-based dual band radar at 1-4 GHz and 6-9 GHz. Three agricultural fields, were selected for this study, one each for maize, barley, and wheat. High-resolution synthetic aperture radar (SAR) images are generated for each crop in order to differentiate them based on amplitude distributions, as well as elevation dependencies.
The impact of hippocampal CA2 excitatory/inhibitory imbalance on social interaction and network activity
(Universität Ulm, 2026-02-13) Franz, Alessa Adriana; Kätzel, Dennis; Böckers, Tobias; Hengerer, Bastian
Clinical and preclinical studies associated disturbances in the neuronal integrity of hippocampal CA2 with pathomechanisms of schizophrenia. This potential link between CA2 and schizophrenia substantiates the need to gain a deeper understanding of CA2’s role in social interactions in social groups for extended periods and in brain-wide network function. Utilising a chronic chemogenetic approach, the excitatory output of CA2 was reduced by expressing an inhibitory Designer Receptor Exclusively Activated by Designer Drug (DREADD) in CA2 pyramidal neurons. The effects of this manipulation were explored using the social arena and functional Ultrasound imaging (fUS), thereby moving beyond currently employed assessment strategies in CA2 research. Chronic DREADD-mediated reduction of excitatory output from CA2 resulted in brain-wide hypoactivity and lasting functional dysconnectivity within frontal brain circuits in mice, exhibiting impaired social interaction with familiar mice - possibly due to a social memory impairment – that lasts for hours. Mice furthermore exhibited a more subtle “sniff and hop” interaction, characterised by more frequent but shorter direct interactions with familiar mice during the dark phase. Importantly, chronic CA2 inhibition impaired the social long-term memory of mice, as indicated by the higher interaction with familiar mice (compared to controls) upon long-term separation and exposure to novel mice. Furthermore, acute and chronic CA2 inhibition revealed a crucial role of CA2 in the functionality of global neuronal networks. CA2 controls the neuronal activity of numerous brain regions, particularly the hippocampus (HPC), anterior cingulate cortex (ACC), thalamus and pallidum. Moreover, CA2 regulates the long-range HPC-medial prefrontal cortex (mPFC) connectivity as well as the communication between frontal areas. In addition, CA2’s influence on the frontothalamic circuit was uncovered. Brain regions influenced by CA2 overlap with core components of the rodent social memory circuit, as well as the rodent homologues of the human social brain and human default mode network, suggesting key role of CA2 in major brain networks. Inhibiting the neuronal activity of CA2 reproduced some pathological features, observed in patients with schizophrenia: impaired social interaction and social memory, disruptions in the HPC-mPFC axis and frontal hypoconnectivity. Thus, behavioural and network data highlight the potential link of hypoactivity in CA2 pyramidal neurons and pathomechanism in schizophrenia. In summary, this work provided novel insights into CA2’s impact on the frontal brain circuit, brain-wide network communication, as well as social interaction and social long-term memory of mice in social groups. Concluding, this work substantiates the need to further exploit hippocampal CA2 in health and disease.
A novel energy storage system with cascaded H-bridge converter
(Universität Ulm, 2026-02-12) Li, Boyang; Xie, Jian; Dieckerhoff, Sibylle
This dissertation explores the design and optimization of a Cascade H-Bridge Converter (CHBC) for energy storage systems (ESS), addressing the efficiency limitations of traditional inverters. Traditional inverters exhibit high switching losses due to their high switching frequencies and working voltages, resulting in efficiencies generally below 98%. The proposed CHBC reduces these losses by lowering the switching frequency and enhancing overall system efficiency. The hardware design includes 100 slave modules, a master control system, and a communication bus. Each slave module, powered by an Nickel Manganese Cobalt Oxide (NMC) battery, collects battery voltage and temperature data via an MCU, which also controls the H-bridge states. Power efficiency is achieved by deriving the 12V power supply for the H-bridge drivers from cascaded charge pumps boosting from 3.3V. The master control system samples grid voltage and current, performs power calculations, and coordinates the slave modules to generate the desired voltage waveform. Communication between master and slaves is facilitated by an SPI protocol with digital isolators ensuring proper isolation and an open-drain gate circuit enabling MISO line sharing. A robust design using a two-layer Flexible Printed Circuit Board (FPCB) and additional digital isolators and Schmidt trigger buffers achieved a communication bus supporting an 8 MHz clock frequency with strong real-time communication and interference resistance. Software evaluations of modulation methods for CHBC with varying numbers of slave modules were conducted to minimize harmonic distortion. Nearest Level Modulation (NLM) produced lower harmonic distortion when a large number of slave modules was used, while Level Shift Pulse Width Modulation (LSPWM) maintained low harmonic distortion across different numbers of slave modules. The CHBC-based ESS employs a single-phase power controller with a Phase-Locked Loop (PLL) based on a Second- Order Generalized Integrator, verified by a simulation model. To address production inconsistencies in battery capacities, the dissertation discusses active and passive balancing methods and introduces rotation balancing for CHBC. Three methods for State-of- Charge (SOC) comparison were proposed: current integration, external battery voltage comparison, and internal voltage estimation. Each method has its advantages and limitations, with internal voltage estimation providing the highest accuracy under varying conditions. The modified CHBC prototype reduced system complexity and cost, with each slave module controlling five batteries, resulting in 24 slave modules. Designed for the residential energy storage market, the system uses safer Lithium Ferrophosphate (LFP) batteries. Harmonic currents were mitigated by adding a third harmonic current controller, and compliance with German VDE 4105 standards was demonstrated through steady-state and transient tests. Theoretical and direct testing confirmed maximum efficiencies of 98.91 % and 98.6 %, respectively. Thermal balancing methods were proposed to address significant temperature differences among the batteries during full power operation. Reactive power-based thermal balancing and battery-sorting-based thermal balancing were evaluated, showing effectiveness in reducing heat generation depending on the phase difference between voltage and current. The CHBC system was extended for three-phase power grid applications, with a master controller managing 72 slave modules in a delta configuration. Positive sequence, negative sequence, and zero sequence controllers were designed and simulated. Compensation coefficients were derived to correct power output errors, and zero sequence current control was used to achieve SOC balancing among three phases.
Peer-Begleitung in der psychiatrischen Versorgung aus der Perspektive von Professionellen: Qualitative Ergebnisse zur Implementierung in fünf Ländern
(Universität Ulm, 2026-02-10) Haun, Maria; Krumm, Silvia; Gündel, Harald
Hintergrund: Im Rahmen von Peer-Begleitung erhält eine Person mit einer psychischen Erkrankung Unterstützung von einer Person, die selbst Erfahrung mit schweren, psychischen Krisen und/oder psychiatrischen Diagnosen gemacht und damit einen konstruktiven Umgang gefunden hat. Die Bedeutung von Peer-Begleitung als Bestandteil der Versorgung von Menschen mit psychischer Erkrankung hat weltweit zugenommen. Der Großteil der Forschung zu Peer-Begleitung stammt jedoch aus Ländern mit einem hohen Durchschnittseinkommen, in denen mehr Ressourcen für die psychiatrische Versorgung zur Verfügung stehen als in Ländern mit niedrigem Einkommen. Die Übertragbarkeit der Ergebnisse zur Peer-Begleitung auf Versorgungskontexte in Ländern mit mittlerem und niedrigem Einkommen ist nicht ausreichend belegt. Methode: UPSIDES („Using Peer Support In Developing Empowering Mental Health Services“) ist eine multizentrische Studie, in der an sechs Studienstandorten in Asien (Israel, Indien), Afrika (Uganda, Tansania) und Europa (Deutschland, 2 Standorte) Peer-Begleitung für Erwachsene mit psychischer Erkrankung implementiert wird. Die vorliegende Doktorarbeit ist Teil der qualitativen Evaluation der UPSIDES Intervention und präsentiert die Ergebnisse aus drei Publikationen, in denen Erwartungen an und Erfahrungen mit UPSIDES Peer-Begleitung aus Sichtweise zweier Gruppen (psychiatrische Fachkräfte und Key Informants) untersucht werden. Hierfür wurden leitfadengestützte Fokusgruppen mit psychiatrischen Fachkräften (n=60) und Key Informants, also Personen, die über Fachwissen zur lokalen Implementierung der UPSIDES Intervention verfügen (n=54), durchgeführt. Die Datenerhebung fand zu zwei Zeitpunkten (vor und mind. eineinhalb Jahre nach dem Start der UPSIDES Intervention) statt. Die Auswertung der Daten erfolgte durch eine qualitative Inhaltsanalyse, mit Schwerpunkt auf induktiver Kategorienbildung in Publikation I und II und einer deduktiven Kategorienbildung durch die Anwendung eines Implementierungsmodells (Consolidation Framework for Implementation Research) in Publikation III. Ergebnisse: Standortübergreifend zeigten sowohl psychiatrische Fachkräfte als auch Key Informants vor und nach der Implementierung der UPSIDES Intervention eine positive Einstellung gegenüber Peer-Begleitung. Peer-Begleiter:innen wurden als Vorbilder für Klient:innen angesehen, die eigene Erfahrungen teilen, Hoffnung vermitteln und Selbstvertrauen bei Klient:innen stärken. Teilnehmende aus Standorten mit geringeren Ressourcen (Dar es Salaam und Butabika) berichteten zusätzliche Vorteile von Peer-Begleitung, wie die Arbeitsentlastung von Fachkräften durch Peer-Begleitung oder verbesserten Zugang zu psychosozialen Diensten für Betroffene und deren Angehörige. Die Perspektiven von Fachkräften und Key Informants unterschieden sich zwischen den Standorten auch in Hinblick auf die Erwartungen in der Zusammenarbeit mit Peer-Begleiter:innen oder die gewünschten Trainingsinhalte für Peer-Begleiter:innen. Die bisherige Erfahrung mit Peer-Begleitung am Standort wurde als Faktor herausgearbeitet, der für die Implementierung wie auch die Zusammenarbeit von Fachkräften und Peer-Begleiter:innen eine wesentliche Rolle spielt, d.h. die Zusammenarbeit stützen oder erschweren kann. Unabhängig von der Vorerfahrung schienen an Standorten mit geringeren Ressourcen organisationale Vorteile, wie die Entlastung von Fachkräften durch Peer-Begleitung, die Implementierung zu beschleunigen. Schlussfolgerung: Die Ergebnisse dieser Doktorarbeit verdeutlichen das Potential von Peer-Begleitung aus der Perspektive von Fachkräften und Key Informants für die Verbesserung der psychosozialen Versorgung von Menschen mit psychischen Erkrankungen in verschiedenen Ländern. Gleichzeitig zeigt die vorliegende Arbeit, dass kontextuelle Faktoren (Vorerfahrung mit Peer-Begleitung, das Vorhandensein von Ressourcen) sowie deren spezifisches Zusammenspiel am Implementierungsstandort maßgeblich die Einstellung der Akteure gegenüber Peer-Begleiter:innen sowie die lokale Umsetzung von Peer-Begleitung beeinflussen. Zukünftige Forschung zu Peer-Begleitung sollte insbesondere die Perspektive von Akteuren aus Ländern mit mittlerem und niedrigem Einkommensniveau einbeziehen, um die Möglichkeiten und Grenzen von Peer-Begleitung als Intervention in verschiedenen Kontexten zu untersuchen.
Polycytotoxic T lymphocytes in human tuberculosis
(Universität Ulm, 2026-02-10) Zumwinkel, Marc; Stenger, Steffen; Sparrer, Konstantin; Mwandumba, Henry
Tuberculosis (TB) remains a major global health challenge, with millions of new infections and deaths annually. This thesis highlights the roles of innate and adaptive immune responses in TB, focusing on how different immune cells contribute to controlling the infection. The primary aim was to understand how polycytotoxic T lymphocytes (P-CTL) contribute to the restriction of Mycobacterium tuberculosis (Mtb) growth in primary human macrophages. P-CTL are defined by their simultaneous expression of granzyme B, perforin, and granulysin. All three effector molecules act together to mediate antimicrobial activity. Here, we describe the identification, purification, and functional characterization of P-CTL and their interaction with Mtb-infected macrophages. P-CTL circulate in the peripheral blood of healthy donors with a low and variable frequency (1 % - 5 %). To characterize their immune functions, we purified two subsets of unconventional T cells with an enhanced frequency of P-CTL based on the expression of prototypical NK cell receptors NKG2A and NKG2C. We found that both receptors are sufficiently enrich for P-CTL (up to 50 %). NKG2C+ T cells featured a greater cytotoxicity and constitutive degranulation, while NKG2A+ T cells released higher levels of IFN-γ and IL-10. Strikingly, both subsets significantly inhibit the growth of intracellular Mtb, which correlated with the frequency of P-CTL. Initial encounter (24 hrs.) with an oxygen-restricted environment did not significantly affect the frequency of P-CTL. Moreover, we found that P-CTL are induced by stimulation with soluble mycobacterial antigens, further substantiating their preventive role in the control of infectious diseases. Still, it is necessary to enhance the frequency of P-CTL to harness them for therapeutic immune interventions. We utilized the well-known cytokine IL-2, but co-stimulation could not significantly increase the frequency of P-CTL in Mtb-specific T cells. This thesis discusses the potential role of NKG2A and NKG2C T cells in early TB infection and suggests that targeting these receptors could be a novel strategy for TB immunotherapy. Due to their prominent antimicrobial activity, targeting P-CTL should be considered for novel vaccination strategies to combat severe infectious diseases.
Decision making in management accounting
(Universität Ulm, 2026-02-10) Hupfer, Garlef; Seiter, Mischa; Trapp, Rouven
This cumulative doctoral thesis examines how cognitive processes influence decision-making in management accounting. Three studies examine the use of algorithmic information in performance evaluation, the effects of management controls on behaviour in inter-organizational collaborations, and the role of task complexity in decision-making. Using experiments, survey data, and a meta-analysis, the dissertation shows that black-box algorithms are not systematically underweighted when presented as reliable, that impaired cooperation promotes authoritarian behaviour reinforced by results controls, and that higher task complexity leads to more deliberate and extreme decisions. Overall, the findings provide new insights into how information characteristics, control systems, and task complexity jointly influence managers' judgements and behaviour.
On the succinctness of languages for static variability
(Universität Ulm, 2026-02-09) Moosherr, Benjamin; Bittner, Paul Maximilian; Tichy, Matthias; Thüm, Thomas
Software product lines, such as the Linux kernel, are software systems that can be configured to numerous variants and adding features to software product lines often results in exponential growth of the number variants. There are dozens of variability languages to encode the variants of a software product line but their ability to solve this exponential growth problem is not yet explored formally. To this end, we define and formalize \emph{succinctness} as a relation comparing the size of variability language expressions. Our key idea, based on big-\(\mathcal{O}\) notation, is to compare the asymptotic growth of expression sizes between variability languages. We explain succinctness and its implications intuitively, formalize succinctness in the proof assistant Agda, and explore succinctness relationships between a range of existing variability languages which have been used in previous research. With succinctness, we offer a tool to compare variability languages aiming to improve our understanding of variability languages and aiding variability language users, such as researchers and tool authors, into choosing the best fitting variability language for their use case.
Covariance-modulated optimal transport and gradient flows
(Universität Ulm, 2024-12-03) Burger, Martin; Erbar, Matthias; Hoffmann, Franca; Matthes, Daniel; Schlichting, André
Abstract We study a variant of the dynamical optimal transport problem in which the energy to be minimised is modulated by the covariance matrix of the distribution. Such transport metrics arise naturally in mean-field limits of certain ensemble Kalman methods for solving inverse problems. We show that the transport problem splits into two coupled minimization problems: one for the evolution of mean and covariance of the interpolating curve and one for its shape. The latter consists in minimising the usual Wasserstein length under the constraint of maintaining fixed mean and covariance along the interpolation. We analyse the geometry induced by this modulated transport distance on the space of probabilities as well as the dynamics of the associated gradient flows. Those show better convergence properties in comparison to the classical Wasserstein metric in terms of exponential convergence rates independent of the Gaussian target. On the level of the gradient flows a similar splitting into the evolution of moments and shapes of the distribution can be observed.
Shelter selection in females of two scorpion species depends on shelter size and scent
(Universität Ulm, 2024-11-02) Hladik, Janina; Bailer, Yorick; Wolf, Harald; Stemme, Torben
Abstract Shelter selection is an important task in an animal’s life. Concerning scorpions, little is known on the evaluation of potential shelters and the importance of chemosensation. To address these issues, we conducted a two-choice shelter test in rectangular open field arenas to identify properties rendering shelters attractive for female scorpions of the species E. italicus and M. eupeus . Shelters varied in size (large, small) and scent (none, conspecific: male or female, attractive: prey, aversive: rosemary oil). Contact with the shelters was video-recorded under red light for 13 h, including the whole night phase. Results revealed a preference for larger shelters, with conspecific scent having minor or no influence. Striking differences occurred with regard to prey and rosemary oil scents. Prey scent was more attractive to M. eupeus , while rosemary oil did not act as a repellent. E. italicus was not very attracted by prey scent, but was repelled by rosemary oil. These findings might reflect the different habitats, semi-arid vs. Mediterranean climates: prey and rosemary are scarce in the semi-arid climate (habitat of M. eupeus ), whereas they are abundant in the Mediterranean climate (habitat of E. italicus ). We carried out impairment experiments to identify the main sensory organs responsible for the above observations. These are the pectines and pedipalps which function as mechano- and chemosensors. Scorpions could not detect size nor scent properly when either their pectines or pedipalps were impaired.
Exact and parameterized algorithms for the independent cutset problem
(Universität Ulm, 2024-10-18) Rauch, Johannes; Rautenbach, Dieter; Souza, Uéverton S.
The Independent Cutset problem asks whether there is a set of vertices in a given graph that is both independent and a cutset. This problem is NP-complete even when the input graph is planar and has maximum degree five. We first present a O∗ (1.4423n)- time algorithm to compute a minimum independent cutset (if any). Since the property of having an independent cutset is MSO1 -expressible, our main results are concerned with structural parameterizations for the problem considering parameters incomparable with clique-width. We present FPT-time algorithms under the following parameters: the dual of the maximum degree, the dual of the solution size, the size of a dominating set (where a dominating set is given as an additional input), the size of an odd cycle transversal, the distance to chordal graphs, and the distance to P5 -free graphs. We close by introducing the notion of α-domination, which generalizes key ideas of this article.
Mechanisms of change in digital interventions for depression: A systematic review and meta-analysis of six mediator domains
(Universität Ulm, 2024-09-14) Angerer, Florian; Mennel, Vera; Grund, Simon; Mayer, Axel; Büscher, Rebekka; Sander, Lasse B.; Cuijpers, Pim; Terhorst, Yannik; Baumeister, Harald; Domhardt, Matthias
Background: While the efficacy of digital interventions for the treatment of depression is well established, comprehensive knowledge on how therapeutic changes come about is still limited. This systematic review aimed to provide an overview of research on change mechanisms in digital interventions for depression and meta-analytically evaluate indirect effects of potential mediators. Methods: The databases CENTRAL, Embase, MEDLINE, and PsycINFO were systematically searched for randomized controlled trials investigating mediators of digital interventions for adults with depression. Two reviewers independently screened studies for inclusion, assessed study quality and categorized potential mediators. Indirect effects were synthesized with a two-stage structural equation modeling approach (TSSEM). Results: Overall, 25 trials (8110 participants) investigating 84 potential mediators were identified, of which attentional (8 %), self-related (6 %), biophysiological (6 %), affective (5 %), socio-cultural (2 %) and motivational (1 %) variables were the scope of this study. TSSEM revealed significant mediation effects for combined self-related variables (ab = −0.098; 95 %-CI: [−0.150, −0.051]), combined biophysiological variables (ab = −0.073; 95 %-CI: [−0.119, −0.025]) and mindfulness (ab = −0.042; 95 %-CI: [−0.080, −0.015]). Meta-analytical evaluations of the other three domains were not feasible. Limitations: Methodological shortcomings of the included studies, the considerable heterogeneity and the small number of investigated variables within domains limit the generalizability of the results. Conclusion: The findings further the understanding of potential change mechanisms in digital interventions for depression and highlight recommendations for future process research, such as the consideration of temporal precedence and experimental manipulation of potential mediators, as well as the application of network approaches.
Cross‐referenceable microscopic and micro‐spectroscopic analysis of fiber reinforced thermoplastics
(Universität Ulm, 2024-10-08) Yoshida, Yuki; Furushima, Yoshitomo; Okino, Shunnosuke; Mizaikoff, Boris
Abstract This study aims at the assessment of damage on crystallinity of fiber reinforced thermoplastic (FRTP) composites resulting from mechanical polishing to demonstrate a novel methodology enabling comprehensive crystalline structural characterization. Investigating the crystalline structure is relevant for understanding the relationship with the mechanical properties. In a previous study, FRTPs have successfully been processed to the thickness of 5 μm or below via an innovative polishing strategy, which for the first time visualized the crystalline morphologies by polarized optical microscopy. Analyzing these sections via micro‐Raman and micro‐infrared spectroscopies facilitate the quantification of these structure. However, no study reports on the correct interpretation of FRTP data appropriately considering the damage resulting from mechanical polishing to date. Herein, we report fundamental knowledge on this aspect via detailed Raman micro‐spectroscopic investigations to demonstrate our novel strategy. Four types of FRTPs with major matrix resins (GF/PBT, CF/PA6, CF/PEEK and GF/PP) were analyzed and the damage on both polished cross‐sections, prepared by traditional process, and the thin sections was confirmed to be approx. 1%–2% crystallinity, which solidified the credibility of the obtained analytical data. Then, the newly established methodology of ‘cross‐referenceable microscopic and micro‐spectroscopic analyses for FRTPs’ was applied for comprehensive crystalline structural characterization. Highlights Cross‐sectioning and thin‐sectioning of FRTPs. Evidencing of damage on the crystallinity resulting from mechanical polishing. Equations to convert Raman spectral data into crystallinity. Strategy for considering the anisotropy of Raman measurements. Practical application of novel microscopic and micro‐spectroscopic strategies.

Browse

Recent Submissions