Promotionszentrum
Refine
Document Type
- conference proceeding (article) (25)
- Article (14)
- conference proceeding (presentation) (5)
- Preprint (5)
- Report (2)
- Other (1)
Publication reviewed
- peer review (31)
- nicht begutachtet (12)
- begutachtet (9)
Keywords
- SOTIF (5)
- Automated Driving System (3)
- Cybersecurity Management System (3)
- Functional Safety (3)
- HADF (3)
- Highly Automated Driving (3)
- Large Language Model (3)
- Artificial Intelligence (2)
- Condition Monitoring (2)
- Deep Learning (2)
Institute
- Promotionszentrum (52)
- Fakultät Informatik (26)
- IFM - Institut für Fahrerassistenz und vernetzte Mobilität (23)
- Fakultät Maschinenbau (22)
- EPT – Institut für Effiziente Produktionstechnik (12)
- IDF – Institut für Datenoptimierte Fertigung (8)
- IEAT - Institut für Energie- und Antriebstechnik (4)
- IMS - Institut für Maschinelles Sehen (4)
- Fakultät Elektrotechnik (2)
- IKR – Institut für angewandte Künstliche Intelligenz und Robotik (2)
Efficient production planning, machine optimization, and bottleneck analysis are critical for reducing cycle times in manufacturing processes. This paper introduces an automated process mining (PM) approach designed to streamline these tasks by generating and optimizing finite state machines (FSMs) from event logs derived from signal changes in industrial plants. Unlike conventional methods, this approach requires no prior knowledge of the configuration or behavior of the target programmable logic controllers (PLCs). The synthesized FSMs facilitate the extraction of process-oriented insights, enabling the identification of underperforming manufacturing processes and detailed analysis of their durations. The proposed method is validated through a case study on a real industrial plant, demonstrating its efficacy in uncovering process inefficiencies and supporting decision-making. This work provides a novel, generalizable framework for process analysis in manufacturing environments, contributing to the broader field of automated process optimization.
Machine Learning (ML)-based LiDAR 3D object detectors in automated driving produce false detections, missed detections, and localisation errors under adverse weather and reduced visibility. Detection errors arising without hardware or software faults constitute performance insufficiencies under ISO 21448, Safety of the Intended Functionality (SOTIF), and the standard requires identification of the triggering conditions responsible. The prescribed analysis methods assume a design specification, but ML-based LiDAR object detectors have no design specification because the mapping from point clouds to bounding boxes is learned from training data. This paper proposes an uncertainty evaluation methodology that uses disagreement among deep ensemble members to separate correct from incorrect detections. Ensemble disagreement and performance insufficiencies arise from insufficient training data coverage of the operating condition. The methodology evaluates whether three uncertainty indicators derived from ensemble disagreement (mean confidence, confidence variance, and geometric disagreement) separate correct from incorrect detections. The evaluation produces outputs mapped to ISO 21448 analysis activities: discrimination metrics, triggering condition rankings by false positive share, frames flagged for investigation, and acceptance gates reporting coverage and false acceptance rate. A case study using simulated ensemble predictions across 22 weather configurations shows that geometric disagreement achieves the strongest separation, with acceptance gates that retain only true detections at reduced coverage. The observed separation arises because false detections produce spatially inconsistent bounding boxes across ensemble members where no physical object constrains the predicted position, while true detections remain spatially consistent.
Uncertainty in LiDAR sensor-based object detection arises from environmental variability and sensor performance limitations. Representing these uncertainties is essential for ensuring the Safety of the Intended Functionality (SOTIF), which focuses on preventing hazards in automated driving scenarios. This paper presents a systematic approach to identifying, classifying, and representing uncertainties in LiDAR-based object detection within a SOTIF-related scenario. Dempster-Shafer Theory (DST) is employed to construct a Frame of Discernment (FoD) to represent detection outcomes. Conditional Basic Probability Assignments (BPAs) are applied based on dependencies among identified uncertainty sources. Yager's Rule of Combination is used to resolve conflicting evidence from multiple sources, providing a structured framework to evaluate uncertainties' effects on detection accuracy. The study applies variance-based sensitivity analysis (VBSA) to quantify and prioritize uncertainties, detailing their specific impact on detection performance.
Exploring the Potential of Dialogue-Based Robots to Motivate Social Connection in Older Adults
(2026)
This study investigates whether a dialogue-based robot, employing motivational interviewing techniques, can enhance the intrinsic motivation of older adults to engage with their local social networks. A user study was conducted in which a Furhat robot interacted with participants, first presenting information about upcoming local social events and subsequently using motivational interviewing to encourage reflection on their personal motivation to attend. The study included 42 older adults (aged between 57 and 90 years old, mean age = 73.9 years). Participants completed the Situational Intrinsic Motivation Scale (SIMS) before and after the interaction with the robot to assess changes in intrinsic motivation, extrinsic motivation, identified regulation, and external regulation. Additionally, the Negative Attitudes Toward Robots Scale (NARS) was administered, and semi-structured interviews were conducted post-interaction. Results indicated no statistically significant changes in SIMS scores, though a trend toward significance was observed for identified regulation (p = 0.076). Analysis of NARS scores and qualitative interview data revealed predominantly positive attitudes toward the robot, with many participants expressing openness to future use of dialogue-based robots for social motivation. These findings suggest promising avenues for further research on the potential of robotic systems to support social engagement among older adults.
Gegenwärtige Entwicklungen der Digitalisierung und Datenökonomie, insbesondere multilateraler Plattformen zum Datenaustausch, bieten das Potenzial für eine beschleunigte Umsetzung von Kreislaufwirtschaftspraktiken in der produzierenden Industrie. Der Beitrag untersucht systematisch und anhand originärer Forschung, inwieweit die Digitalisierung als Katalysator der Kreislaufwirtschaft im Beschaffungswesen solcher Unternehmen dienen könnte. Dafür wurden acht Experten aus fünf weltweit führenden Herstellern und Zulieferern der Automobil- und Luftfahrtbranche interviewt. Es werden praxisnahe Hypothesen für die nachhaltige Gestaltung von Lieferketten entwickelt und zwei spezifische Use Cases für Kreislaufwirtschaftspraktiken vorgeschlagen, die dem Ressourceneinsatz proaktiv entgegenwirken können.
Aging workforces in western industrialized countries are leading to a growing loss of experiential knowledge, particularly on the shop floor where expertise is largely tacit and informally acquired. This HICSS presentation reports on a study that examines current challenges in industrial knowledge management and identifies promising application areas for large language models (LLMs) within a broader effort to develop human-centered, GenAI-supported KM solutions. Based on two structured workshops with 23 experts from 14 German manufacturing companies, the study reveals key barriers to knowledge retention and highlights two prioritized LLM use cases. The findings provide concrete design implications for LLM-enabled KM systems and contribute practical insights into addressing demographic change in industrial contexts.
Integrating heat pumps into district heating systems is a crucial pathway toward decarbonizing urban heat supply and increasing the share of renewable energy in the heating sector. However, diverse temperature requirements of district heating networks impose distinct challenges on heat pump technologies. This study presents a systematic thermodynamic analysis of three fundamental closed-loop heat pump cycles - vapor compression, reverse Brayton, and reverse Stirling - for district heating applications. It considers representative temperature levels and transients of environmental and industrial heat sources, along with varying district heating supply temperatures to reflect typical operating conditions.
In aging societies across western industrialized nations, the loss of expertise due to retiring skilled workers presents a critical challenge for industry. That is especially true on the shop floor, where much of the knowledge is tacitly gained through years of hands-on experience rather than formal documentation. This study explores current knowledge management (KM) challenges and systematically identifies high-potential applications for large language models (LLMs) as part of a broader research initiative aiming to develop human-centered KM solutions supported by generative artificial intelligence (GenAI). We conducted two structured workshops with 23 participants from 14 German manufacturing companies. Three core barriers and two prioritized LLM use cases were identified, contributing specific design recommendations for LLM-supported KM systems for companies. The results advance the understanding of GenAI-assisted knowledge retention in industrial settings and provide a practical foundation for addressing the demographic shift through intelligent, technology-driven solutions.
Durability of proton exchange membrane fuel cells (PEMFCs) in single-cell setup is often assessed using potentiostatic test protocols under inert cathode atmosphere, such as the US DOE’s accelerated stress tests (ASTs). Although these voltage-controlled protocols impose consistent electrochemical stress, they do not accurately replicate real-world operating conditions, for instance due to the lack of product water formation in the cathode or due to the generation of heat. In contrast, stress testing under hydrogen and air, which is commonly employed for PEMFC durability testing on stack level, reflect the hybrid potential- and current-controlled degradation modes which are typical of mobile applications. This is usually done galvanostatic, but can also be realized in a voltage controlled manner. There are notable differences in the ageing behaviors between those two methods: Potentiostatic testing protocols, in which the voltage remains constant throughout the test duration, induce consistent electrochemical stress regarding potential driven degradation modes such as platinum dissolution, Ostwald ripening, carbon corrosion and chemical membrane degradation due to hydrogen peroxide formation. In contrast, galvanostatic testing protocols result in a gradual decline in potential with time, thereby reducing potentially driven electrochemical stress. However, in galvanostatic mode, water production remains constant (unlike in potentiostatic mode), which is known to contribute equally to fuel cell degradation. This dynamic of varying potential driven and humidity driven stress significantly affects catalyst and membrane durability, underscoring the need to understand the degradation behavior specific to each testing method. The present study compares the impact of load-cycle and voltage-cycle protocols on PEMFC degradation. Electrochemical characterizations, such as cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS), are used to identify the ageing behavior of the individual components. All tests were performed using commercial Gore® PRIMEA® membrane electrode assemblies (MEA) in a Baltic quickConnect high amp test cell with an active area of 12 cm². Our findings highlight the importance of testing under hydrogen and air to better replicate operational conditions and enhance PEMFC durability in practical applications.
Der demografische Wandel verstärkt den Arbeits- bzw. Fachkräftemangel in der produzierenden Industrie, wodurch das Thema Wissensmanagement für viele Unternehmen an Bedeutung gewinnt. Insbesondere die Erhebung und Konservierung impliziten Wissens stellt eine Herausforderung dar. Die vorliegende Studie beleuchtet, inwieweit Large Language Models (LLMs) im Bereich der Wissenserhebung mit Experteninterviews sinnstiftend unterstützen können. Drei Experten testen und bewerten den entwickelten Chatbot. Die Ergebnisse des Interviews sind vielversprechend, die Zusammenfassung hingegen zeigt Verbesserungspotenziale