This thesis will focus on developing methods to protect privacy while using Large Language Models (LLMs) in automotive requirements processing. As LLMs process sensitive data, such as customer information and design specifications, ensuring that this data remains private is crucial. The research will explore anonymization techniques like data masking and pseudonymization, which hide sensitive details while keeping the data useful for LLM tasks. It will also focus on ensuring that anonymized data retains its meaning and effectiveness in the context of automotive requirements processing. For application please send me an email with title "Master Thesis Application: Data Anonymization LLM ". Please also attach your resume and transcript of records in the email. A motivation letter is NOT required.

Large language models (LLMs) have transformed how we interact with technology, allowing users to communicate with systems using natural language. Ubuntu, a popular open-source operating system, provides a powerful environment for exploring the use of LLMs as intelligent agents. This master topic focuses on using LLMs as agents that can directly control and manage Ubuntu. The goal is to develop an LLM-based agent that can perform tasks such as configuring system settings, installing software, troubleshooting issues, and managing resources—entirely through natural language commands. This research will explore how LLMs can streamline system management, making Ubuntu more intuitive and user-friendly, while also enabling greater automation and efficiency for users at all levels. For application please send me an email with title "Master Thesis Application: An LLM-based Ubuntu User". Please also attach your resume and transcript of records in the email. An motivation letter is NOT required.

Are you interested in topics such as Power to X, methanol production, CO2 capture and waste treatment? In this master's thesis, you will use Aspen Plus to simulate and optimize plant processes.

Zielsetzung Das primäre Ziel dieses Projekts ist die Erforschung, wie LLMs zur Generierung sicherheitskritischer Szenarien im CommonRoad-Framework eingesetzt werden können.

Using the Laplace approximation in Bayesian deep learning for uncertainty estimation in the Airbus vision-based landing system. For details, see the attached file.

Wir suchen einen engagierten Masteranden, der die Programmiersprache ALDiSP weiterentwickelt. Sie studieren Informatik oder ein verwandtes Fachgebiet, verfügen über fundierte Kenntnisse in C++ und im Compilerbau und bringen idealerweise auch Erfahrungen mit LLVM, Bison und Flex mit? Dann sind sie der optimale Kandidat. Freuen Sie sich auf ein innovatives Forschungsumfeld mit modernster Technik, intensiver fachlicher Betreuung und Unterstützung bei wissenschaftlichen Publikationen.

This Master’s thesis explores circular economy strategies for biogenic waste in Bavaria using Life Cycle Assessment (LCA). The research builds on an ongoing project, analyzing biomass flows and innovative utilization pathways like pyrolysis and gasification. The student will conduct a literature search, perform an LCA, and compare current vs. future waste management strategies. Applicants should have a background in sustainability, familiarity with LCA/MFA, and an interest in bioeconomy.

When working with robotic assistance systems in the operating theater, surgeons and assistant physicians may have to adopt unergonomic postures. As part of the project, a VR environment is to be created that enables the experimental investigation of human-robot interactions in a virtual clinical environment. As part of the project, the course of an operation is to be recreated with the help of infrared tracking and haptics and evaluated in a user study. The study builds on preliminary student work. The technical basis is provided by the Unity development environment and the Vicon infrared camera system. At the end of the work, a functional 3D environment is to be created that can be used for ergonomic studies.