Offenburg University of Applied Sciences
Hochschulschriftenserver der Hochschule OffenburgNot a member yet
6641 research outputs found
Sort by
Ion chromatography coupled with optical emission spectrometry (IC-ICP-OES) methodology for the analysis of inositol phosphates in food and feed
Get PDFThis study presents the development of ion chromatography coupled with inductively coupled plasma optical emission spectrometry (IC-ICP-OES) for the simultaneous determination and quantification of inositol phosphates (InsPx). Using a CarboPac PA100 column with a nitric acid-water gradient, 28 InsPx isomers (InsP6 to InsP2) were separated within 33 min. The method eliminates baseline drift and post-column derivatization thereby simplifying detection and quantification. It achieves low detection limits of 63 μg/L P across a range of 63–3200 μg/L P. Various extraction and sample preparation methods for food and feed matrices were tested, including acidic and alkaline agents, C18 SPE and spin concentrators. The analysis shows intra-day and intra-laboratory reproducibility with deviations smaller than 1 % for standard solutions and under 4 % for feed samples (80 % recovery rate of phytate). This methodology is applicable to explore enzymatic degradation pathways and the analysis of InsPx in complex food and animal feed matrices
Reliable Evaluation of Attribution Maps in CNNs: A Perturbation-Based Approach
Get PDFIn this paper, we present an approach for evaluating attribution maps, which play a central role in interpreting the predictions of convolutional neural networks (CNNs). We show that the widely used insertion/deletion metrics are susceptible to distribution shifts that affect the reliability of the ranking. Our method proposes to replace pixel modifications with adversarial perturbations, which provides a more robust evaluation framework. By using smoothness and monotonicity measures, we illustrate the effectiveness of our approach in correcting distribution shifts. In addition, we conduct the most comprehensive quantitative and qualitative assessment of attribution maps to date. Introducing baseline attribution maps as sanity checks, we find that our metric is the only contender to pass all checks. Using Kendall’s τ rank correlation coefficient, we show the increased consistency of our metric across 15 dataset-architecture combinations. Of the 16 attribution maps tested, our results clearly show SmoothGrad to be the best map currently available. This research makes an important contribution to the development of attribution maps by providing a reliable and consistent evaluation framework. To ensure reproducibility, we will provide the code along with our results
Umsetzung und initiale Validierung eines Systems zur Perturbationserzeugung bei menschlicher Lokomotion auf Laufbändern
No full textIm Rahmen dieser Arbeit wird das Bump’em System des Stanford Biomechatronics Laboratory aufgebaut und anhand von Funktionstests initial validiert. Mit diesem Vorgehen soll die Eignung des Open-Source-Systems zur Perturbationserzeugung im Laborkontext ermittelt werden. Perturbationen sind Einflüsse, die die Bewegung eines Objektes stören. Systeme, die diese Einflüsse auf Menschen erzeugen, können genutzt werden, um die Mechanismen von Stürzen beim menschlichen Gang zu untersuchen. So können Maßnahmen zur Sturzprävention entwickelt werden.
Das Bump’em System wurde entwickelt, um mit einem vielseitigen und zuverlässigen perturbationserzeugenden System ein breites Feld an Forschungen zu ermöglichen. Es ist aus modular aus Seilzügen, die durch Motoren angetrieben werden, aufgebaut. Ein optionaler Kraftsensor ermöglicht die Steuerung über einen präzisen Regelungskreis und wird in dieser Arbeit verwendet. Der modulare Aufbau des Systems ermöglicht Anpassungen an die jeweilige Forschungsfrage.
Zum Aufbau wird von den Entwicklern eine Anleitung bereitgestellt. An dieser müssen Anpassungen gemacht werden, die neben dem Aufbau in dieser Arbeit erläutert werden. Beim Aufbau fallen einige Schwierigkeiten auf, die entsprechend in der Arbeit erläutert werden.
Die Validierung erfolgt durch verschiedene Funktionstests. Es wird plötzlich eine Kraft auf einen Probanden aufgebracht, um die Reaktionszeit des Systems zu testen. Außerdem wird die Genauigkeit anhand von Kraftprofilen getestet. Zur Auswertung der Tests wird die vom Kraftsensor gemessene Kraft mit der eingegebenen Kraft in Form von Diagrammen verglichen.
Es ist auffällig, dass die Werte der Kraft zunächst deutlich über die der Soll-Kraft ansteigen sich aber nach einiger Zeit an diese angleichen. Die Reaktionszeit des Systems beim ersten Anstieg der Kraft ist länger als erwartet. Als Hauptgründe dafür werden die Einstellung der Parameter des verwendeten Regelkreises und das Durchhängen der Seile vermutet.
Zum Ende der Arbeit ist das Projekt grundsätzlich aufgebaut und funktionsfähig, bedarf jedoch stellenweise einer Optimierung. Daraus folgt die Erkenntnis, dass das System prinzipiell für die Perturbationserzeugung im Laborkontext geeignet ist, aber mit hohem Zeit- und Rechercheaufwand verbunden ist.In this work, the Bump'em system of the Stanford Biomechatronics Laboratory is set up and initially validated in functional tests. This procedure is intended to determine the suitability of the open source system for generating perturbations in a laboratory context. Perturbations are influences that disturb the movement of an object. Systems that generate these influences on people can be used to investigate the mechanisms of falls in human gait. In this way, measures can be developed to prevent falls.
The Bump'em system was developed to enable a wide range of research with a versatile and reliable perturbation-generating system. It is constructed from modular pulleys driven by motors. An optional force sensor enables a precise control loop and is used in this work. The modular design of the system allows it to be adapted to the respective research question.
The developers provide instructions for the setup. Adjustments have to be made to these instructions. They are explained in this paper along the setup. It is noticeable that some difficulties appear during the setup process, which are explained accordingly in the paper.
Validation is carried out using various functional tests. A force is suddenly applied to a test person to test the reaction time of the system. The accuracy is also tested using force profiles. To evaluate, diagrams that compare the force measured by the force sensor to the target force are generated.
It is noticeable that the force values initially rise significantly above the target force, but then level off after some time. The reaction time of the system at the first increase in force is longer than expected. The main reasons for this are assumed to be the settings of the parameters of the control loop used and the sagging of the cables.
At the end of the work, the project is set up and functional but requires optimization. This leads to the conclusion that the system is suitable for perturbation generation in a laboratory context, but is associated with high time and research costs
Die Rolle von akustischen Elementen bei der Markenbildung
No full textHeutzutage sind Marken mehr als nur Logos oder Slogans. Sie sprechen die Menschen auf verschiedenen sensorischen Ebenen an. Akustische Signale und Musik sind in der Lage, Emotionen hervorzurufen, Wiedererkennung zu gewährleisten und die Marke erlebbar zu machen. Diese Publikation zielt darauf ab, zu analysieren, welchen Effekt Sound Branding auf die Markenbildung bzw. das Markenimage hat. Dazu wird zunächst ein Überblick über zentrale Modelle der Markenidentität gegeben. Im Anschluss daran wird untersucht, welche Rolle Sound Branding innerhalb der Markenstrategie spielt und welche akustischen Elemente einschließlich Parameter eingesetzt werden können. Die Analyse zeigt, dass Sound Branding mehr als eine kreative Marketingidee ist und Effektivität für die Markendifferenzierung leisten kann. Mithilfe von auditiven Reizen wird die Markenbotschaft emotional aufgeladen und auf lange Sicht im Gedächtnis des Konsumenten verankert. Da sich derzeit der Wettbewerb verschärft und Marken auf immer mehr Plattformen präsent sein müssen, gewinnt Sound Branding zunehmend an Bedeutung.In the contemporary business environment, brands have evolved to encompass much more than mere logos or slogans, becoming multifaceted entities that resonate with consumers on various sensory levels. The utilization of acoustic signals and music has emerged as a potent tool for evoking emotions, ensuring brand recognition, and fostering a tangible brand experience. This publication seeks to delve into the impact of sound branding on brand development and image management. To this end, an initial overview of seminal models of brand identity is provided. The subsequent examination analyses the role of sound branding within brand strategy, exploring the acoustic elements and parameters that can be employed. The analysis demonstrates that sound branding is not merely a creative marketing concept, but a strategic tool for brand differentiation. The utilization of auditory stimuli enhances the emotional resonance of the brand message, ensuring its long-term memorability. In the context of intensifying competition and the growing need for brands to be present across multiple platforms, sound branding is becoming increasingly pivotal
Green IT
Get PDFDieser Leitfaden bietet eine kompakte Übersicht zu strategischen, organisatorischen und technischen Aspekten, um Green IT nachhaltig an Hochschulen und öffentlichen Einrichtungen zu verankern. Er ist auf die spezifischen Anforderungen dieser Institutionen abgestimmt und versteht sich als praxisbezogene Orientierungshilfe. Die Inhalte und Ergebnisse basieren auf den Erkenntnissen und Arbeiten des Forschungsprojekts "Green Academic IT Potential (GAIT)"
Demonstration and Evaluation of Model Predictive Control (MPC) for a Real-World Heat Pump System in a Commercial Low-Energy Building for Cost Reduction and Enhanced Grid Support
Get PDFHeat pumps play a crucial role in decarbonizing buildings, yet conventional control strategies limit their grid-supportive potential. Model Predictive Control (MPC) offers a promising alternative to optimize energy costs and grid performance, but real-world implementations remain scarce. This study demonstrates the feasibility of MPC in a low-energy, non-residential building by integrating a controller based on electricity market prices. The system, deployed on a Raspberry Pi and integrated into the building automation system, utilizes weather forecasts and a grey-box model for load prediction. A key challenge is the lack of standardized interfaces for heat pump controls, requiring custom solutions. A 7-day performance analysis compares MPC with conventional control, focusing on economic efficiency and grid support. MPC shifts heat pump operation to periods of lower electricity prices, increasing storage temperatures and reducing the average COP from 7.6 to 6.0. Despite this, energy costs decrease by 40%, lowering the electricity procurement price from 0.36 EUR to 0.12 EUR/kWh, while the Grid Support Coefficient improves by 13%. These results confirm that MPC can enhance heat pump operation with simple component models, provided the system allows flexibility and demand is predictable
Optimizing Computational Efficiency in TPMS Structural Design through Surrogate Modeling
No full textThis paper introduces a surrogate modeling approach for simulating Triply Periodic Minimal Surfaces (TPMS), which can be used in various engineering applications due to their unique geometric, mechanical properties and light weight properties as a second hierarchy level below the component design level. The detailed finite element analysis of theses structure in components, require inten- sive computational resources due to the complex geometries of TPMS. To address these challenges, we propose a surrogate model that simplifies the geometric and mechanical representation of TPMS, enabling more effcient simulations without compromising accuracy. This model was validated against experimental data from bending tests performed on TPMS structures fabricated using additive manufac- turing techniques. The surrogate model demonstrated excellent agreement with experimental results, offering a promising tool for the rapid and effcient design of TPMS-based components. The model’s utility is underscored by its application in predicting the mechanical behavior of TPMS structures under various loading conditions, confirming its potential for broader adoption in engineering design processes
A Concept for Form-Adaptive and Force-Controlled Gripping with Magnetorheological Material
Get PDFThis paper presents a shape-adaptive gripper concept based on a 3D-printed structure made from TM5MED, filled with magnetorheological fluid (MRF). The gripper utilizes electromagnets to control the viscoelastic properties of the MRF, allowing for rapid shape adaptation. High-performance strain sensors made from flexible carbon nanotube (CNT) films measure pressure and deformation at the contact surface and along the sides, allowing for nuanced force control. This design concept is particularly suited for robotic applications that require flexible and safe object handling
KI im Marketing
No full textKünstliche Intelligenz (KI) beeinflusst zahlreiche Bereiche des Marketing und bietet umfangreiche Möglichkeiten, das Marketing effektiver zu gestalten und einen großen Mehrwert für Unternehmen zu bieten. Im Rahmen des Beitrages werden die vielfältigen Einsatzmöglichkeiten von KI-Anwendungen entlang des Marketingprozesses, also der Situationsanalyse, Strategie, Implementierung und Kontrolle vorgestellt. Neben den vielen Vorteilen, die KI-Anwendungen im Marketing haben können, sollen gleichzeitig auch immer die Herausforderungen im Zusammenhang mit dem Einsatz von KI-Anwendungen aufgezeigt werden. Da viele Unternehmen bei einem solchen vielfach noch Schwierigkeiten haben, wird neben den Anwendungsmöglichkeiten noch eine KI-Marketing-Strategie vorgestellt, die Hinweise zum Einsatz von KI im Marketing gibt
Development, Construction, and Characterization of a Lab-Scale Airlift Reactor for Methanotrophic Bioconversion of Landfill Gas
No full textGreenhouse gases are becoming an increasingly pressing problem over time. This Master's thesis focuses on the design, development, and experimental characterization of a lab-scale airlift bioreactor for the cultivation of methanotrophic microorganisms. The motivation behind this work lies in the growing need to biologically remediate landfill gas a potent greenhouse gas mixture or, ideally, to convert it into valuable bioproducts such as polyhydroxyalkanoates (PHAs). To address this, a custom designed internal loop airlift reactor was developed, constructed, and successfully put into operation. The focus in this work was placed on investigating the gas liquid mass transfer coefficient (kLa) as well as changing time in different solutions (water, nitrate mineral salt medium, and xanthan). The results show that the reactor achieves kLa values of up to 0.63 min⁻¹, making it comparable or even superior to systems reported in the literature. First comprehension in energy efficiency might show many advantages as well. Overall, the system provides a solid foundation for future scale-up and applications in biological gas utilization