University of Siegen
Permanent URI for this communityhttps://dspace.ub.uni-siegen.de/handle/ubsi/7
Browse
99 results
Search Results
- Some of the metrics are blocked by yourconsent settings
Publication Ein datengetriebenes Verfahren zur robusten und zuverlässigen Auslegung des Niederdruck-Kokillen-Gießprozesses(2026)Diese Arbeit demonstriert einen datengetriebenen Ansatz zur effizienten Ermittlung eines robusten und zuverlässigen Arbeitspunkts im Niederdruck-Kokillen-Gießprozess. Durch dynamische Versuchsplanung, Modellierung und Optimierung werden im Rahmen der Prozessauslegung Parameterkombinationen identifiziert, die die Ausschusswahrscheinlichkeit und die Zykluszeit unter Berücksichtigung von Unsicherheiten minimieren.Source Type:Doctoral ThesisSeries:Forschungsberichte des Lehrstuhls für Umformtechnik21 13 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Ein datengetriebenes Verfahren zur robusten und zuverlässigen Auslegung des Niederdruck-Kokillen-Gießprozesses(2026)Ensuring product quality while maintaining high productivity and energy efficiency are key aspects and therefore levers for optimization measures in companies. In many sectors, including light metal foundries, data-driven methods based on machine learning are already being used in the context of Industry 4.0 to continuously improve process efficiency. Many studies are focused on analyzing process data during serial production. A holistic, data-based strategy for determining a robust and reliable process design that considers any process fluctuations that occur has not yet been developed for low pressure die casting. This paper therefore presents a novel, automated method for determining a robust and reliable operating point during the development phase of the low pressure die casting process. The focus is on the process parameters for a given casting and mold design but can easily be extended to include design parameters. Within the framework of design of experiments, the procedure includes the needs-based adjustment of the number of experiments as well as compliance with input and output restrictions, the modelling of both static and dynamic system behavior and, within the multi-criteria optimization, the consideration of both uncertainties in the process parameters and uncertainties due to limited metrological accessibility. A selection mechanism, which supports the user in the rapid analysis of robust and reliable Pareto-optimal solutions, rounds off the procedure. The performance is demonstrated using a real low pressure die casting process at MARTINREA HONSEL GERMANY GMBH. The results show that the approach enables a robust and reliable casting process design with comparatively low energy input, considering the boundary conditions of short cycle times.Source Type:Doctoral ThesisSeries:Forschungsberichte des Lehrstuhls für UmformtechnikVolume number:2111 18 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Real-Time Data-Driven Urban Flood Forecasting(2025-12-12)Flood forecasting is a critical component of mitigating risks to human life, infrastructure, and ecosystems, particularly as climate change increases the frequency and severity of extreme weather events. Especially urban areas are vulnerable due to high sealing rates, dense infrastructure, and concentrated populations. Traditional forecasting approaches rely on physically based hydrodynamic models that provide accurate information on the location, timing, and extent of flood events. While these models represent the state of the art, they are computationally too demanding for real-time forecasting and operational decision support. An alternative is offered by data-driven models, which learn patterns and relationships directly from the data. By approximating the underlying input–output dynamics without solving the governing physical equations explicitly, data-driven approaches achieve substantially reduced computation times, making them suitable for real-time flood forecasting. This dissertation develops and advances data-driven forecast systems tailored for urban environments, addressing three major flood types: (1) sewer surcharge events, (2) fluvial floods, and (3) pluvial floods. The research follows a structured progression, advancing from localized short-term predictions to large-scale dynamic forecasts, and finally to domain-independent transferable, physically informed frameworks. For sewer surcharge events, an Ensemble Forecast System based on nonlinear autoregressive networks with exogenous inputs was developed. The Ensemble Forecast System predicts the onset and duration of surcharge events in manholes while explicitly quantifying structural model- and parameter uncertainty through ensemble modeling. With observed events consistently captured within uncertainty bounds, the system provides a reliable framework for early warnings and can support model-based or real-time control in modern sewer networks. For fluvial floods, a Feature-Informed Forecast System was introduced, framing flood prediction as an image-to-image translation task. By integrating convolutional neural networks (CNNs) with spatial features, such as distance-to-river, this system predicts maximum inundation depths within seconds, generating inundation maps that support real-time risk assessment, traffic management, and emergency response planning. The Feature-Informed Forecast System achieves accuracy comparable to that of hydrodynamic models, thus demonstrating the scalability of data-driven approaches to river flooding in urban areas. Building on this, to capture flood dynamics over time, a Multi-Step Dynamic Forecast System was developed. Employing recursive CNN-based predictions, the Multi-Step Dynamic Forecast System produces temporally resolved inundation maps for up to 24 hours ahead. Compared to physically based models, it maintains high accuracy and reliability, thereby bridging the gap between static maximum flood maps and real-time dynamic monitoring. Since these and, in general, conventional data-driven models are often case-site specific and therefore domain-dependent, their applicability in different sites is limited. To overcome this constraint, this dissertation proposes a physically informed spatial-temporal forecast framework, in the context of pluvial floods. This approach integrates CNN-based image-to-image translation with physical constraints by embedding the continuity equation and a kinematic wave approximation into the loss function. This framework improves physical plausibility, enhances predictive accuracy, and enables domain-independent forecasts that generalize to new and unknown areas. Collectively, the contributions of this dissertation demonstrate how data-driven, especially physically informed, models transform flood forecasting into a real-time, scalable, and transferable application. The developed systems advance the state of the art by combining computational efficiency, predictive accuracy, and physical consistency, laying a new foundation for further research in the field of flood forecasting and early warning systems in urban areas.Source Type:Doctoral Thesis10 15 - Some of the metrics are blocked by yourconsent settings
Publication Datengestützte Umformtechnik: Künstliche Intelligenz zwischen Prozesswissen, Regelung und Produktionsrealität(2026)With the 7th Biegeforum, the Chair of Forming Technology at the University of Siegen once again offers researchers the opportunity to present their topics and findings to a specialist audience from industry and academia. This conference volume brings together current research findings and innovative developments in forming technology, highlighting the broad scope of the research field. The Biegeforum thus underscores its role as a key platform for exchange between research and industry — a role it has fulfilled since its first edition in 2011.Source Type:Conference ProceedingsSeries:Forschungsberichte des Lehrstuhls für Umformtechnik4 5 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Human Assistance System Supporting Forming Machine ReconfigurationIncreased product complexity in present manufacturing sector, shorter product lifecycles, and more variable customer demands, require production systems to be flexible and reliable. As a result, machines are required to be reconfigured more frequently and within shorter timeframes. Reconfiguring forming machines is a complex and delicate task that depends strongly on operator experience and practical know-how gained over several years. Appropriate support system is required to transfer the accumulated experience and knowledge in order to enable novice operators to perform such tasks effectively. This work presents an assistance system designed to support novice operators during forming machine reconfiguration tasks. The proposed system comprises an eye-gaze sensor to observe operator actions and visual attention during task execution. Based on this information, the system identifies potential mistakes and provides real-time feedback on a smartwatch to guide the operator through the reconfiguration process. The approach is evaluated in a controlled laboratory environment, where both task performance and user experience are assessed. The results show potential for the use of human assistance systems to support novice operators in real time during forming machine reconfiguration on the shop floor. In addition to improving the overall system and its evaluation in a real production environment under industrial conditions, future work will focus on multimodal sensor fusion to enable more robust intent recognition and contextual understanding of an operator during reconfiguration of forming machines.Source Type:InProceedings10 11 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Enabling scalable and flexible production through digital twin-based robot–machine commissioningThis paper presents a conceptual method for a digital twin–based framework for scalable, flexible robot–machine commissioning in batch production processes. Following the concept of virtual commissioning as defined in VDI 4499, this method proposes a robot-first, event-driven interaction paradigm in which a mobile manipulator initiates transportation, handling, and manipulation tasks while production machines remain in a ready state and respond to the robot's actions. The approach is validated through a proof-of-principle demonstration using a nine-degree-of-freedom mobile manipulator (SkaBot9) with an integrated shelf. The complete process logic and communication architecture are simulated and transferred to the physical system without structural changes. The results demonstrate that the proposed method supports early validation of robot–machine interaction and a potential for commissioning with flexible, mobile robot–based production systems.Source Type:InProceedings3 8 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , A Conceptual Process Architecture for the Digital Integration of Tube Fabrication in Mechanical and Plant EngineeringIn mechanical and plant engineering practice, tube fabrication is still frequently characterized by assembly-driven and experience-based procedures. Geometric specifications are often finalized during assembly based on measurements taken on the physical product, resulting in iterative adjustments, limited process standardization, and a strong dependence on tacit expert knowledge. Although advanced computer-aided design (CAD) and computer-aided manufacturing (CAM) technologies and automated tube bending systems are available, a structured integration of tube fabrication into a consistent digital process architecture is often lacking. This paper develops a conceptual process architecture that structures the digital integration of tube fabrication in project-driven mechanical and plant engineering environments. Based on an analysis of typical industrial process structures, the proposed architecture relocates geometric specification, manufacturability validation, and product–process coordination to early digital stages. Defined transition mechanisms between engineering, manufacturing, and assembly domains are introduced to prevent installation from functioning as a geometry-defining validation environment. The scientific contribution lies in the domain-specific operationalization of integration principles for tube fabrication through a governed process architecture that relocates geometric finalization to engineering, formalizes transitions between engineering and manufacturing, and institutionalizes manufacturing knowledge within digital design processes. The proposed process architecture comprises model-based tube system specification, manufacturing-oriented design validation, deterministic derivation and controlled release of fabrication data, digitally driven prefabrication, and structured geometric verification prior to assembly. In addition, technical and organizational boundary conditions, including data continuity, interface governance, and knowledge integration, are systematically addressed. The resulting framework provides a conceptual basis for reducing assembly-driven iteration, increasing reproducibility, and supporting the formalization of manufacturing knowledge in mechanical and plant engineering.Source Type:InProceedings3 12 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , TubeBEND: A Real-World Dataset for Geometry Prediction in Rotary Draw BendingThis paper presents TubeBEND, a real-world dataset comprising 318 rotary tube bending processes, which were collected and sorted by experts from various fields to evaluate machine learning and signal analysis methods. The dataset addresses the industrial challenge of predicting the geometry of a first-stage bend, which can be beneficial for designing machine clamping molds for the second-stage bend in two-stage rotary draw bending. Some geometry criteria, such as the tube's final bent angle (or springback) and its cross-sectional deformation, are being recorded in this dataset. This dataset enables the development of machine learning models for geometry prediction and improved machine setup, supporting operators in reducing springback and deformation. Moreover, by recording some process parameters, such as tool movements and forces or torques applied to them, we deliver detailed information about their impacts on the final tube geometry. The focus of our work is to discover solutions that can replace traditional methods, such as trial-and-error or simulation-based predictions, by including experimental process variables in ML algorithms. Our dataset is publicly available as a benchmark to improve data-driven methods in this field at: https://github.com/zeyneddinoz/tubebend and https://zenodo.org/records/18458823.Source Type:InProceedings2 13 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Entscheidungsfindung zur Rekonfiguration von Umformwerkzeugen mit Künstlicher Intelligenz(2026); Unsicherheiten und Störungen an Umformwerkzeugen wirken sich negativ auf die Produktion aus. Um eine robuste und flexible Produktion sicherzustellen, werden resiliente und rekonfigurierbare Fertigungssysteme benötigt. Die vorliegende Arbeit präsentiert ein bereits getestetes Vorgehen mit Fuzzy-Logik und Case-Based Reasoning (CBR) für Umformwerkzeuge, in diesem Fall ein mehrstufiges Werkzeug einer Transferpresse. Diese zugrunde liegenden Methode erkennt Fehlersituationen und schlägt dem Bediener Maßnahmen zur Rekonfiguration vor, um angemessen auf die jeweilige Situation reagieren zu können. Die Nutzung von Maschinen- und Sensordaten in Verbindung mit Expertenwissen, das beim Aufbau von Zugehörigkeitsfunktionen, Fuzzy-Regeln und Case-Based Reasoning eingesetzt wird, ermöglicht die frühzeitige Erkennung, Identifikation und Beseitigung von Fehlern im Produktionsprozess. Die Leistungsfähigkeit des Dienstes wird anhand eines Benchmarks quantifiziert.Source Type:InProceedings7 12 - Some of the metrics are blocked by yourconsent settings
Publication Access status: Open Access , Numerische Analyse des Rotationszugbiegens runder und quadratischer RohreDas Rotationszugbiegen (RZB) ist ein etabliertes Verfahren zur Herstellung anspruchsvoller Rohrbögen mit engen Radien und geringen Wanddicken. In dieser Arbeit wird das Biegeverhalten von Rohren mit rundem und quadratischem Querschnitt mithilfe von Simulationsmodellen auf Basis der Finite-Elemente-Methode (FEM) untersucht. Ziel ist es, Unterschiede im Biegeverhalten beider Querschnittsformen unter nahezu identischen Biegebedingungen zu analysieren, wobei ein einheitlicher Biege- und Wanddickenfaktor als Grundlage dient. Im Fokus der Untersuchung stehen die Spannungsverteilung und die Verformungsmuster während des Biegeprozesses, mit besonderem Schwerpunkt auf dem Einspannbereich. Die Ergebnisse zeigen, dass Quadratrohre, insbesondere dünnwandige, anfälliger für Faltenbildung und das Einfallen der Flanken sind. Ihre ebenen Flächen reagieren unter den entlang des Umfangs wirkenden Spannungen anders als die der Rundrohre und stehen beim Spannen nur in den Radien in Kontakt mit den Werkzeugflächen. Folglich sind die für Rundrohre entwickelten Arbeitsdiagramme nicht direkt auf Quadratrohre übertragbar. Die Resultate verdeutlichen, dass für das Rotationszugbiegen von Quadratrohren spezifische Arbeitsdiagramme und angepasste Biegeprozesse erforderlich sind.Source Type:InProceedings6 11
