HiWi Position: Synthetic 3D bridge model generation using Python

06.05.2026, Student assistants, internships, student research projects

Looking for a HiWi student with experience in Python programming and bridge architecture/engineering.

Synthetic point cloud datasets are essential for training and evaluating deep learning-based models for bridge infrastructure. Our chair has developed a parametric pipeline using the CadQuery library (Python) and the Helios++ simulation tool to generate synthetic bridge point clouds of RC slab bridges, with geometric variability grounded in German design documentation.

However, a gap remains between the bridges currently modeled and those present in publicly available real-world datasets (Ruodan LU, 2018) as well as real bridges scanned by our chair. The tasks of this HiWi position are to close that gap by extending the existing pipeline to incorporate additional bridge typologies and geometric configurations that better reflect these real-world counterparts.

Tasks

• Review the existing synthetic generation pipeline (CadQuery, Helios++) and understand the parametric bridge models currently implemented, including their variability logic and scanning simulation setup.
• Analyze publicly available bridge point cloud datasets and locally acquired scans to identify missing bridge types, component geometries, and structural configurations.
• Extend the parametric modeling pipeline in Python to incorporate new bridge designs.
• Validate the generated synthetic point clouds against real datasets.
• Maintain clean, well-documented, and modular Python code integrated into the existing pipeline.

Requirements

• Background in civil engineering or architecture, with knowledge of bridge design and German design standards.
• Strong Python programming skills and ability for spatial geometric reasoning.
• Basic knowledge of BIM concepts and CAD tools.
• Experience with CadQuery, point clouds, or simulation tools is a plus.
• Ability to work independently for 8 hours per week; remote work possible.

Application

Please send your CV and current transcript to salman.ahmed@tum.de .

Additionally, include 5 bullet points in your email explaining why you are a good fit for this role.

Kontakt: salman.ahmed@tum.de