02.10.2025

[Master Thesis] On-body AI-guided visualization of medical imaging volumes

This research investigates the visualization of medical volumes, like computed tomography (CT) scans, on a patient's body. This is to be done with the help of virtual reality glasses, overlaying the scan over a user's virtual body. The goal is to provide an easy-to-understand way in which doctors and patients can communicate about upcoming medical procedures with visual support.
read more

Kontakt: hex-thesis.ortho@mh.tum.de, constantin.kleinbeck@tum.de

01.10.2025

Bachelor- und Masterarbeiten im Brandschutz

Sie suchen gerade ein Thema für eine Bachelor oder Master Thesis?

Für das Wintersemester sind an der Professur für Brandwissenschaft und Brandingenieurwesen noch Themen zu vergeben:

Kontakt: manuel.kitzlinger@tum.de

01.10.2025

Bachelor- und Masterarbeiten im Brandschutz

Sie suchen gerade ein Thema für eine Bachelor oder Master Thesis? An der Professur für Brandwissenschaften und Brandingenieurwesen sind noch Themen für das Wintersemester zu vergeben:

• Bachelorthema: Aufzeigen des Zusammenspiels und der Unterschiede bei der Verwendung von Ereignis- und Fehlerbäumen zur Beurteilung von Brandschutzmaßnahmen am Beispiel ovn Sprinkleranlagen
• Masterthema: Konsistenzprüfung von Annahmen mit Indexmethoden bei der Verwendung von Fehlerbäumen zur Beurteilung von Brandschutzmaßnahmen
• MAsterthema: Modellierung und Implementierung von Staubewertung und Diagrammdarstellungen für makroskopische Räumungsmodelle

Kontakt: manuel.kitzlinger@tum.de

01.10.2025

Agilox Scenario Setup

[Localization] Development of a Test Platform for Indoor Radar Localization

We are currently building a sensor rig to equip an industrial robot with measuring capabilities. The robot is able to autonomously operate and gather measurement data. Using this, we want to build an extensive data collection of radar data in industrial settings, which will aim to accelerate the development of radar-based approaches in more diverse environments.

Kontakt: vasilije.rakcevic@tum.de

23.09.2025

[Master Thesis] In situ N-glycomics for tumor characterization

Are you curious about how molecular details hidden in tissues can reveal key insights into disease? This thesis project offers you the chance to explore N-glycomics in tumors through mass spectrometry imaging.

Kontakt: juliana.goncalves@tum.de

19.09.2025

BASAMA-Ethics-Aware Decision-Making with Large Language Models

The primary objective of this project is to develop a framework for ethics-conditioned decision-making using Large Language Models (LLMs) and to evaluate their principle adherence and counter-reasoning robustness.
read more

Kontakt: yuan_avs.gao@tum.de

19.09.2025

BASAMA-Talk2Drive: A Conversational Preference-Aligned Driving Framework Using Multi-Modal Large Language Models

The primary objective of this project is to develop a conversational, preference-aligned driving framework that (i) elicits and tracks driver expectations via natural dialog and (ii) outputs high-level, safety-checked control commands for the motion planner. We will evaluate alignment quality, safety, comfort, and robustness to phrasing variations.
read more

Kontakt: yuan_avs.gao@tum.de

19.09.2025

BASAMA-Learning from BEV and Dash-Camera Views: A Multimodal Large Language Model for Quantitative Risk Assessment

The primary objective of this project is to develop a Multimodal Large Language Model (MLLM) agent that learns from both BEV images and dash-camera views during training, and is able to perform quantitative, spatio-temporal risk assessment from dash-camera views alone at inference.
read more

Kontakt: yuan_avs.gao@tum.de

19.09.2025

Master Thesis / IDP: Evaluation and Optimization for LLM Inferences on RISC-V based Tenstorrent AI Accelerators

We are looking for motivated computer science students with research interests in LLM inference and computer systems.
read more

Kontakt: binqi.sun@tum.de

17.09.2025

Data-Driven Prediction of Surgery Times from Clinical and Operational Features

This thesis develops data-driven methods to predict surgery durations using hospital datasets. By applying statistical and machine learning models to clinical and operational features, it aims to improve scheduling accuracy, optimize resource allocation, and reduce delays. A reusable prediction pipeline will provide ongoing, practical support for hospital operations.
read more

Kontakt: sidra.rashid@tum.de