Quantum metrology and quantum thermodynamics are two relatively young fields that, in the spirit of the second quantum revolution, apply quantum theory to address fundamental questions and to develop technological applications. While the research in these fields is predominantly conducted with discrete, finite-dimensional quantum systems in mind, this thesis makes the case for continuous variable mechanical degrees of freedom. In the first of three projects, we present two realizations of the Otto cycle with a planar rotor as the working medium. As a mechanical system, the planar rotor has a well-defined classical analogue that allows the identification of genuine quantum effects by comparing classical and quantum machine. Here, we mainly focus on the parameter regimes, where the machine admits a certain operation mode, i.e. an engine, a refrigerator, or a heater. In the first realization, we find a systematic disadvantage of the quantum machine. The opposite is true for the second realization: It can be shown that the classical machine can, in general, not be run in a useful operation mode. The quantum machine, on the other hand, admits an engine operation mode for sufficiently cold temperatures of the cold bath. The second project is devoted to the dynamics of a quantum system subjected to a thermal gas. In contrast to repeated interaction models, we consider any gas particle as a motional degree of freedom. We Employ quantum mechanical scattering theory to derive a low-density limit master equation, including gases with internal structure. Then, the thermodynamic consistency of the master equation is shown. A comparison with repeated interaction models makes evident that the inclusion of motional degrees of freedom of the gas plays a curial role for the consistency. Finally, we consider a nonequilibrium scenario, where the internal and motional degrees of freedom of the gas are thermal with respect to different temperatures. We show that, under the influence of this gas, the ergotropy of the system can increase. In the last project, we apply quantum metrology on the measurement of magnetic moments in an electron microscope. We consider two types of sample, one that is static and does not change under the influence of the electron and another one that is described by a quantum system and experiences quantum backaction. For both samples, we derive the scattering operator from first principles. Then, two metrological tasks are considered: First, the sensing of the strength of the magnetic moment. We derive the Fisher information in several bases and find that momentum measurements are already optimal in the case without backaction. With backaction included, a measurement of angular momentum is optimal. The second task, we consider, is the optimal discrimination of the scattered and unscattered motional electron state. We derive the trace distance and find the experimentally achievable basis with the highest classical trace distance, which is still significantly worse than the theoretical optimum.

In recent years, eXplainable Artificial Intelligence (XAI) has received huge attention in the area of explaining the decision-making processes of machine learning models. The aim is to increase the acceptance, trust, and transparency of AI models by providing explanations about the models' decisions. But most of the prior works on XAI are focused to support AI practitioners and developers in understanding and debugging. In this paper, we propose a user-centered explainable energy demand prediction and forecasting system that aims to provide explanations to end-users in the smart home. In doing so, we present an overview of the explainable system and propose a method combining Deep Learning Important FeaTures (DeepLIFT) and Shapley Additive Explanations (SHAP) to explain the prediction of an LSTM-based energy forecasting model.

Analytical chemistry plays a crucial role in ensuring the safety, quality, and authenticity of products and processes, e.g. in the food and beverage industry, pharmaceutical production, but also in forensics, diagnostics, and environmental sciences. One of its most powerful tools is mass spectrometry (MS) which allows scientists to selectively detect substances even at lowest amounts. Conventional MS techniques usually require extensive and complex sample preparation and additional separation procedures which make them time-consuming, resource-intensive and therefore costly. A novel approach known as ambient desorption/ionization MS (ADI-MS), offers a faster and more sustainable alternative by allowing direct analysis of samples in their original state. This dissertation presents the development, optimization, and application of a specific ADI-MS technique termed surface-assisted flowing atmospheric-pressure afterglow MS (SA FAPA MS). This method is based on a plasma-based ionization source in combination with functionalized sample carrier surfaces and enables rapid, preparation-free detection of small molecules in matrix-containing samples, e.g. in beverages, pharmaceutical products or even body fluids. One core goal was to improve the quantitative capabilities of SA FAPA MS as ADI MS in general faces challenges regarding signal stability and overall reproducibility. A key finding was that the choice of the sampling surface dramatically influences performance. Among the surfaces tested, cyano- and dimethyl-functionalized thin-layer chromatography (TLC) plates performed better than conventional supports such as glass or metal mesh. These surfaces enabled higher ion yields, better signal stability, and broader dynamic ranges, ultimately leading to detection limits down to the femtomole (fmol) range. The method was successfully applied to accurately quantify model substances in beverages and electronic cigarette liquids even when matrix effects were expected. Precision and accuracy were further improved by isotopically labeled standards. Further, the semi-quantitative detection of benzocaine in human saliva demonstrated the potential for biomonitoring and diagnostic applications. For the applications above, SA FAPA MS proves to be a fast, sensitive, and sustainable alternative to conventional MS methods. Minimal sample pre-treatment requirements, versatility, and high throughput make it very attractive for use in quality control, toxicology, diagnostics, and environmental monitoring where accurate results are required as quickly as possible.

Das Siegener Löhrtor-Schwimmbad, in den Fünfzigerjahren im Bauhausstil errichtet und seither der Ort, an dem die meisten Siegener das Schwimmen gelernt haben, genügt nicht mehr den Anforderungen an ein modernes Hallenbad: Es ist nur eine Frage der Zeit, wann es geschlossen und vielleicht sogar abgerissen werden soll. In dieser politisch ungeklärten und emotional bewegenden Situation hat eine Gruppe von Studierenden der Siegener Universität unter der Leitung der Professoren Ulrich Exner (Architektur) und Martin Herchenröder (Musik) das Bad architektonisch, klanglich, historisch und funktional untersucht und aus den Analyseergebnissen unter dem Titel "Spiegelungen" eine Performance entwickelt, die Architektur, Installation, Licht, Video, Musik und das allgegenwärtige Wasser in einen neuen, ungewohnten Zusammenklang bringt und in der jedes Element die anderen widerspiegelt - das spannungsvolle architektonische Ineinander von Innen und Außen, Lichtreflexe und Schallreflexionen auf Kacheln, Glas und Wasser, der Raum und sein Kontext in Video und Musik, und in allem die Geschichte und die Geschichten rund um diesen besonderen Ort, Erinnerungen und Träume, Erlebtes und Gedachtes, Gefundenes und Erfundenes. Die Performance ist hier mit Entwürfen, Skizzen, Fotos, Texten und mit einer vollständigen Video-Live-Aufzeichnung dokumentiert. (Link zum Video unter "Externe Ressource")