KITopen
Not a member yet
63711 research outputs found
Sort by
Informationsmodelle und Integrationskonzepte für die Produktion elektrischer Traktionsmotoren. Eine Antwort auf volatile Märkte und Technologien
Get PDFDie wandlungsfähige Produktion von elektrischen Traktionsantrieben lässt sich mithilfe modularer Produkt- und Produktionsbaukästen realisieren. Hierfür sind Informationsmodelle und Integrationskonzepte zu entwickeln. Dieser Beitrag thematisiert die Entwicklung und Implementierung der IT-Infrastruktur für ein agiles Produktionssystem von elektrischen Traktionsantrieben im Rahmen des Forschungsvorhabens AgiloDrive2. Um notwendige Traceability-Anforderungen sowie Transparenz und Flexibilität zu erreichen, wurde ein Integrationskonzept zur Verknüpfung unterschiedlicher in der Produktion vorliegender IT-Systeme erarbeitet und für Evaluierungszwecke umgesetzt
Selecting Data Assets in Data Marketplaces – Leveraging Machine Learning and Explainable AI for Value Quantification
Get PDFIn the era of digital transformation, data is a critical asset driving innovation and competitive advantage for businesses. Data marketplaces have emerged as a key solution for data sharing, yet they face significant challenges, including competitive concerns, matchmaking between data providers and consumers, and a lack of appropriate market mechanisms. This study introduces a data asset value quantification and selection mechanism (DQSM) as an innovative feature for data marketplaces to address these concerns. The DQSM uses Machine Learning and Explainable AI methods to assess the value of data assets, aiding consumers in making informed purchasing decisions. This mechanism addresses the inherent complexities of data asset valuation and selection, thereby increasing marketplace efficiency. Using a design science research approach, the study identifies design principles for the development of the DQSM as a feature of data marketplaces, which are validated through technical experiments with industry and public datasets, as well as interviews with experts in this field. The findings highlight the potential of the DQSM to optimize the discovery and implementation of viable data sharing use cases and to incentivize the adoption of data marketplaces, thereby contributing to more viable and sustainable data ecosystems
Fifty years of power systems optimization
Get PDFThis review paper examines the evolution of power systems optimization over the past fifty years by considering two distinct periods: from 1970 to 1990 and from 1990 to the present. The initial period is typically defined by a centralized power system framework that prevailed around the world. The latter observes a transition to a decentralized structure in a market environment, marked by an increasing integration of renewables and advanced technologies, in addition to maintaining a centralized power system structure for some countries. Since we review the broad topic of power systems optimization, this analysis focuses on the main power system problems — investment, operation planning, operations, control, and forecasting — to define the main research streams. We provide a thorough exploration of the operations research methods applied to specific problem types within each period. Thus, this review not only underscores the pivotal role of operations research in addressing the challenges posed by changing landscapes and advanced technologies but also unveils the transformative journey of power systems optimization along with future research directions
Identifying peace pathways: how can forest conservation contribute to environmental peacebuilding?
Get PDF
Compartment-specific effect of sulfamethoxazole at low μg/L concentrations on microbial nitrogen assimilation in a river system
Get PDFSulfamethoxazole (SMX) is one of the most frequently detected antibiotics in rivers, with concentrations occasionally exceeding the predicted no-effect concentration (PNEC). The impact of such concentrations on the microbial activity of riverine microbial communities remains poorly studied. Here, we investigated the effect of SMX concentrations at the upper end of reported PNEC values (12.5 µg/L) on microbial communities in flume systems with either near-pristine or wastewater-impacted river water. Using a combination of microbiological and chemical methods, we found that SMX was persistent in both near-pristine and wastewater-impacted river water over a time course of 63 days, and had no significant impact on the planktonic bacterial community composition. However, there was an increase in microbial activity after SMX addition. Tracking 15N incorporation in both sample types using Nanoscale Secondary Ion Mass Spectrometry (NanoSIMS) and an Elemental Analyser - Isotope Ratio Mass Spectrometer (EA-IRMS) revealed that SMX concentrations in the test range (10, 100 and 1000 µg/L) enhanced nitrogen assimilation from ammonium up to 64 %. The highest increase was found almost always at 10 µg/L SMX. The response was stronger in samples from the near-pristine site compared to the wastewater-impacted site, and in planktonic biomass compared to biofilms. Overall, our findings reveal a transient increase in microbial nitrogen assimilation with environmentally relevant concentrations of SMX in a habitat-specific manner, but not of SMX degradation, which could be of significance for nutrient dynamics and primary productivity in impacted rivers
Automated manufacturing process for sustainable prototyping of nuclear magnetic resonance transceivers
Get PDFAdditive manufacturing has enabled rapid prototyping of components with minimum investment in specific fabrication infrastructure. These tools allow for a fast iteration from design to functional prototypes within days or even hours. Such prototyping technologies exist in many fields, including three-dimensional mechanical components and printed electric circuit boards (PCBs) for electrical connectivity, to mention two. In the case of nuclear magnetic resonance (NMR) spectroscopy, one needs the combination of both fields; we need to fabricate three-dimensional electrically conductive tracks as coils that are wrapped around a sample container. Fabricating such structures is difficult (e.g., six-axis micro-milling) or simply not possible with conventional methods. In this paper, we modified an additive manufacturing method that is based on the extrusion of conductive ink to fast-prototype solenoidal coil designs for NMR. These NMR coils need to be as close to the sample as possible and, by their shape, have specific inductive values. The performance of the designs was first investigated using electromagnetic field simulations and circuit simulations. The coil found to have optimal parameters for NMR was fabricated by extrusion printing, and its performance was tested in a 1.05 T imaging magnet. The objective is to demonstrate reproducible rapid prototyping of complicated designs with high precision that, as a side effect, hardly produces material waste during production
Leak detection using thermal imagery: Deep learning versus traditional computer vision state-of-the-art
Get PDFAs a cornerstone of climate-neutral heat supply in urban areas, district heating systems require monitoring to detect and mitigate leaks in their subterranean pipelines. Recent research has focused on an approach involving thermography, where leaks are detected as hot-spots in remote sensing imagery. To this end, various traditional computer vision algorithms have been implemented to automate anomaly detection.
This paper pursues a new approach that has so far received little attention in the context of leak detection in district heating pipelines: deep learning, specifically supervised semantic segmentation. By creating a generalisable, multi-stage training procedure to tackle the prevalent limited dataset problem, various architectures are tailored to this anomaly detection task, of which the SegFormer-B2 with Tversky loss is found to perform best. Via comprehensive quantitative, qualitative, explainable AI, and holistic evaluation, the model is assessed and compared to state-of-the-art traditional algorithmic alternatives. It is found to excel, outperforming previous intersection over union scores by almost 10 %pt and maintaining a high precision with little detriment to recall and detection rate
