Offenburg University of Applied Sciences
Hochschulschriftenserver der Hochschule OffenburgNot a member yet
6641 research outputs found
Sort by
Wie KI als Co-Leaderin unterstützen kann
No full textFührungskräfte in Verwaltungen agieren mit hoher Verantwortung, aber geringen Stellhebeln. Künstliche Intelligenz als neutrale Co-Leaderin kann hier den Menschen zwar nicht ersetzen, aber stärken – indem sie bei der Urteilsbildung entlastet, Bias sichtbar macht und den Horizont um zusätzliche Handlungsoptionen erweitert
Comparative Study on AI-Augmented Inventive Problem Solving with TRIZ, TRIZ-AI Hybrid, and Autonomous AI: An Inline Coating Measurement Case in Lithium-Ion Cell Production
No full textThis study presents a comparative experimental analysis of three inventive problem-solving approaches: (a) traditional TRIZ methodology applied without AI assistance, (b) a hybrid approach integrating TRIZ with generative AI, and (c) an autonomous AI-powered process without direct human intervention. A real-world engineering challenge for the experiment, an inline coating measurement for lithium-ion-cell-production, was selected by an expert and structured using detailed problem description. During a maximum seven-hour problem-solving phase, three teams applied the mentioned approaches for ideation, and selected the best 10 solution ideas. The ideation outcomes were evaluated based on key performance metrics: the total number of unique ideas, team and individual ideation productivity, as well as the ideas’ usefulness, feasibility, and novelty. The results offer insights into the performance of AI-augmented problem solving, emphasizing differences in ideation productivity, solution diversity, and the impact of human intervention
AI-Aided Anticipatory Protection of Inventions Against Technical Non-infringing Patent Circumvention
No full textThe emergence of Generative Artificial Intelligence (GAI), particularly Large Language Models (LLMs), offers novel opportunities for enhancing intellectual property (IP) creation and protection in engineering. Technical Patent Circumvention (TPC) refers to the legal design-around of existing patents by identifying alternative technical solutions that avoid infringement, including under the Doctrine of Equivalents. Traditionally, TPC is employed after a patent is granted, but it can also be applied proactively during the early stages of innovation to strengthen patent applications against future circumvention. This anticipatory use of TPC can significantly enhance the robustness of patent protection. However, conventional TPC methods, such as those based on the Theory of Inventive Problem Solving (TRIZ), require extensive training and domain expertise, limiting their practicality for preventive application. By leveraging the capabilities of LLMs, AI-aided TPC offers a more accessible and efficient approach to anticipatory patent protection. This study introduces a workflow for AI-assisted patent circumvention, leveraging OpenAI’s ChatGPT and Google’s Gemini to explore and compare various prompting strategies. The results are evaluated from the perspective of a patent attorney, demonstrating that generative AI can efficiently and effectively uncover viable, non-infringing design alternatives across selected case studies. The paper offers a systematic overview of legal and technical aspects of TPC, outlines traditional and TRIZ-based inventive approaches to TPC, and compares the application of the Doctrine of Equivalents in European and U.S. jurisdictions. It also presents a general algorithm for inventive anticipatory patent circumvention
A SIEM-Based Framework for Multi-Layer Data Collection and Anomaly Detection in OT-Networks
No full textThe increasing convergence of Information Technology and Operational Technology networks in Industrial Control Systems has introduced new cybersecurity challenges, necessitating robust anomaly detection mechanisms. Despite progress, a significant gap exists in the literature regarding benchmark datasets that comprehensively incorporate data from both OT and IT networks. Existing datasets often focus on isolated network segments and fail to capture the full spectrum of network, host, and protocol activities, thereby lacking the holistic visibility required to detect sophisticated, multi-stage cyberattacks. To address this gap, this paper proposes a SIEM-based framework for generating a comprehensive dataset by integrating data from two fundamental levels of an ICS architecture: the OT field network and the IT monitoring system. By leveraging a SIEM solution as the core of the framework, data from heterogeneous industrial devices are captured using NetFlow, Auditd, and Zeek integrations to extract network, host, and protocol features, respectively. The collected data, including sensor telemetry, host logs, and protocol events, are systematically aggregated and correlated to provide a global view of normal and anomalous behaviors. This dataset not only facilitates the development and evaluation of a hybrid anomaly detection framework, combining rule-based SIEM alerts with autoencoder-based machine learning at the edge, but also lays the foundation for more effective, SIEM-driven defenses. Federated learning is discussed as a future prospect to further enhance privacy and scalability
Integrating Robotics into University Curriculum: Driving the Future of Robotics Education
No full textThis contribution presents an innovative educational approach for acquiring advanced competencies in robotic simulation and programming through a real-world industrial case study within the context of higher education. The Work-Life Robotics Institute at Offenburg University of Applied Sciences offers lectures in which theory and practical applications are combined. As part of the module "Advanced Work-Life Robotics", students have the opportunity to directly apply theoretical knowledge from the accompanying robotics lecture to practical scenarios. This takes place within the framework of an application-oriented laboratory project focused on the implementation of realistic automation solutions. Students acquire essential 3D printing skills that allow them to independently design and produce gripper components and workpiece fixtures in line with rapid prototyping principles. The objective of the module is the complete design and realization of a partially or fully automated process for an actual industrial product where the students plan, simulate, and implement the automation solution with a high degree of independence. This approach promotes, in addition to comprehensive technical skills, the development of soft skills such as teamwork, methodological competence, and problem-solving abilities
Schulen im Dialog mit der Öffentlichkeit
No full textIm Bereich der Öffentlichkeitsarbeit von Schulen ist derzeit sehr viel in Bewegung. Wir haben „bewegte Zeiten“ im Umfeld der Schulen, wenn Schülerinnen und Schüler sich verstärkt Gedanken machen, was die weltpolitische Lage mit ihren Zukunftsplänen zu tun hat, die mit Diensten, Pflichtjahrdiskussionen und einer Wiederbelebung der Wehrpflicht Themen auf die Tagesordnung setzen, die vor fünf Jahren für die meisten Schülerinnen und Schüler beim Übergang ins Erwachsenenleben lediglich eine mögliche Option im Bereich freiwilliger Dienste darstellten oder überhaupt nicht im Fokus der jungen Menschen auftauchten.
Zugleich verändert sich das System Schule durch neue Herausforderungen von Finanznot und Personalmangel bis hin zu gestiegenen Anforderungen durch gesellschaftliche Veränderungen in denm Bereichen Migration sowie, Familien- und Sozialstrukturen. Hinzu kommen die technischen Veränderungen, die die Schule in ihrem pädagogischen Kerngeschäft berühren, wie die erweiterte Digitalisierung und die neuen Möglichkeiten der Künstlichen Intelligenz
Becoming River
Full text linkBecoming River is a sensory ethnography along the course of the river Murg in the Black Forest. A media-ecological probe – equipped with cameras, microphones, and sensors – drifts from the spring that is the river’s source to its confluence with the Rhine. The object, constructed from alluvial material, is both an actor and a piece of passive flotsam. Over the course of several months, it weaves a narrative in which observations and events become entangled
Quo vadis advanced footwear technology research?
Full text linkSince the introduction of advanced footwear technology (AFT), distance running performances have improved significantly, as evidenced by a series of world records and general improvements in both elite and recreational running performances. 1 In their recent systematic review and meta-analysis, Stephen et al. 2 synthesize evidence on how AFT as a whole, as well as longitudinal bending stiffness (LBS) and midsole energy return as key constructional features, affect running economy (RE) and ankle mechanics during running. 2 Their systematic review demonstrates that AFT improves RE by∼ 2.7% at an average running speed of 14.5 km/h. In contrast, their synthesis of the currently available literature suggests that neither LBS nor energy return alone significantly improves oxygen consumption or alters ankle mechanics
Evaluation of NIST Lightweight Cryptography Algorithms for Industrial Internet of Things
No full textLightweight cryptography (LWC) has gained abundant attention lately, especially when discussing security in Industrial Internet of Things (IIoT) applications and other systems with constrained resources. IoT has raised various security issues, highlighting the necessity for protection mechanisms. Traditional cryptographic algorithms, such as AES, require significant memory and processing power. LWC, on the other hand, promises security solutions that can meet the needs of IIoT devices with constrained resources. This paper systematically investigates how well the selected algorithms from the NIST CAESAR competition (Ascon, TinyJambu, Schwaemm, Xoodyak, Elephant, Photon-Beetle, and Grain) perform in deeply embedded systems, such as ESP32 microcontrollers. The main contributions of this study include a comprehensive evaluation, including throughput, performance on various data sizes, an initialization time of LWC algorithms, memory usage, CPU utilization, and randomness analysis, along with the calculation of a target function based on critical metrics in IIoT environments. The findings were that ASCON outperformed TinyJambu regarding cryptographic statistical analysis and performance at the expense of highly required memory. When data sizes are less than 64 bytes, TinyJambu, similar to Ascon, offers outstanding throughput and efficiency in memory utilization compared to other LWCs. When comparing LWC results to AES, the throughput, latency, and other metrics of Ascon, TinyJambu, Schwweam, Gift, and Xoodyak were significantly higher. The outcomes of LWC showed encouraging performance, highlighting the lightweight cryptographic algorithm’s potential for use in upcoming IIoT applications
