Open Access

Autonomous Driving (AD) has advanced significantly in recent years, yet widespread deployment remains limited. One of the most enduring challenges in Autonomous Vehicle (AV) development is planning a safe, comfortable, and efficient motion in complex, real-world environments. This thesis addresses motion planning across three distinct domains: urban shuttles, passenger vehicles, and truck-trailer systems. It contributes practical insights and novel approaches toward scalable autonomous mobility. The first part of this work presents an integrated motion planning framework for the Continental Urban Mobility Experience (CUbE) driverless shuttle. Extensive real-world testing over several years highlights the system’s robustness and underscores the importance of long-term validation in urban settings. Key innovations include a multi-layered planning stack and a data-driven motion forecasting approach that enhances interaction with human traffic participants. The second part investigates the behavior of human drivers in understructured traffic environments. Those are areas that fall between well-defined road systems and fully unstructured spaces. A novel analysis framework is introduced for mining patterns from naturalistic trajectory datasets, enabling AVs to better blend into human traffic and navigate ambiguous scenarios with improved predictability and safety. The final part of the thesis explores Deep Reinforcement Learning (DRL) for planning and controlling complex truck-trailer maneuvers. A DRL-based approach is developed and evaluated in simulated environments, demonstrating the method’s potential to handle the nonlinear dynamics of articulated vehicles. These contributions advance the field of motion planning by combining theoretical insights, system-level integration, and empirical evaluation. They offer pathways for improving AV behavior across diverse platforms and use cases, ultimately supporting the broader adoption of AD technologies.

UI/UX-Gestaltung ist für die Mensch-Computer-Interaktion bei Webanwendungen entscheidend. Hohe Nutzererwartungen und starker Wettbewerb erfordern Usability-Messungen, um Nutzer langfristig zu binden. Die Heuristic Evaluation (HE) stellt diesbezüglich eine effiziente Messmethode dar, ihr Erfolg hängt jedoch stark von der Erfahrung der Evaluatoren ab – Experten finden bis zu 50 % mehr Probleme als Novizen. Um diese Performanzlücke bei der Anwendung der HE zu schließen, werden in diesem Paper Eye Movement Modeling Examples (EMMEs) hinsichtlich ihrer Wirksamkeit untersucht. Dabei werden die Blickbewegungen und Kommentare von Usability-Experten aufgezeichnet, während sie eine Website anhand von Jakob Nielsens zehn Heuristiken bewerten. Dies visualisiert Expertenstrategien und kognitive Prozesse und macht die Heuristiken didaktisch nutzbar und für Novizen nachvollziehbar. Fragebogenbefunde zu den EMMEs bestätigen, dass diese als hilfreich und lernfördernd wahrgenommen werden, was ein tieferes Verständnis ermöglicht und den Einsatz in anderen Domänen unterstützen könnte

This study uses holistic models of image perception originating from radiology and psychology to analyze and interpret eye movements during code reviews in the C++ programming language. The study design is based on former experiments, but is supplemented by approaches from expertise research. The study utilizes a sample of 34 subjects whose eye movements are recorded by a Tobii Pro Spectrum 600 Hz. The results show that the holistic models of image perception are suitable for application to source code. In addition, it can be observed that the code reviews are conducted in phases, which are characterized by certain strategies (e.g. scan, error detection, ...). Furthermore, experience-related differences can be detected between experts and novices, which emphasize that experts use elaborate strategies and have a comparatively better ability to collect and process information from source code

Eye-tracking and questionnaires are typically treated as separate methods for measuring usability and user experience (UX). Recent studies show that machine learning models trained solely on eye movements can predict pragmatic and hedonic quality ratings. Building on this, this study examines which gaze patterns predict usability and UX and whether models can generalize across stimuli. Five models were trained on eye movements from 121 users browsing six websites. A feature-importance analysis revealed that saccadic patterns, such as regressions and successive forward movements, are more associated with UX, whereas longer consecutive saccades are indicative of usability. When trained separately for each website, the best-performing models achieve Matthews Correlation Coefficient scores of 0.751 and 0.780, with only small negative effect sizes on holdout data. Trained across websites, holdout scores dropped to 0.196 for usability and 0.338 for UX, suggesting that cross-stimuli generalizability is limited and, at best, achievable for hedonic interaction aspects.

This paper presents the design and implementation of the digitisation and digital subsystems of a physical true random number generator (PTRNG). The system is being developed in alignment with the KRITIS!M project at the Regensburg University of Applied Sciences. The system digitises analogue noise signals and processes them using a modular firmware architecture based on the Zephyr RTOS. The subsystem comprises two custom hardware platforms: One is dedicated to high-speed analogue-to-digital conversion, while the other handles data processing, statistical validation, and external communication via TCP/IP. The design emphasises extensibility, robustness and real-time operation. The firmware employs a mediator-based buffer management system that ensures clean separation and coordination between the acquisition, post-processing, validation and networking components. While challenges remain in optimising the analogue front end, the digital components, however, operate reliably and form a solid foundation for the future integration of cryptographic post-processing and full system validation against established standards, such as NIST SP 800-90 and BSI AIS 31.