PhD Position: Mathematics & AI for Topology-Correct 3D City Reconstruction
I’m recruiting a PhD student to join my group at the Digital Twin Cities Centre (DTCC), working on how to reconstruct analysis-ready, topology-correct 3D city models from multi-modal geospatial data.
A lot of current “3D city modeling” looks good visually, but breaks down the moment you want to do anything serious with it—simulation, decision support, robust measurements, or scalable pipelines. In this project, the goal is to push beyond visually plausible geometry and instead create LOD3 urban surface meshes that are watertight, manifold, free of self-intersections, and have consistent connectivity—so the resulting models can actually be used.
What the project is about
The research sits at the intersection of applied mathematics, geometry processing, and modern machine learning. You’ll work with heterogeneous data such as LiDAR point clouds, aerial orthophotos, street-level imagery, and map/cadastral information, and develop methods that combine learning-based interpretation with topology-aware mesh construction and refinement.
On the machine-learning side, you’ll explore learning-based models for multi-modal 3D reconstruction (for example sparse 3D CNNs and related architectures) and build pipelines for training, evaluation, and large-scale inference.
On the geometry and mathematics side, you’ll design topology-aware meshing and refinement algorithms that support hierarchical levels of detail, and you’ll connect the methods to theory: error measures and bounds, robustness to noise, and conditions that guarantee topological validity. The aim is not just to “make it work,” but to make it principled and reliable.
Why this is a great PhD topic (in my opinion)
You get to work on something that is both deeply mathematical and very real: the outcomes are meant to integrate into the DTCC Platform as open-source components and be validated on real city data with partner use cases in mind. That means you’ll spend time on theory, but also on software that is engineered well enough that others can actually use it.
And if you like the idea of seeing your work land in practical applications—CFD wind comfort, heat/solar analysis, flood and noise simulations, infrastructure planning, AR/VR visualization—this project is built for that.
Who I’m hoping to find
I’m looking for someone with a strong foundation in mathematics and real comfort with coding. You should have (or be close to completing) a Master’s degree in Mathematics. Beyond that, what matters most is that you enjoy tackling hard problems, you can work independently when needed, and you also like collaborating.
Experience with any of the following is a plus: 3D geometry processing, meshing/remeshing/mesh repair, machine learning for 3D or multi-modal data, numerical analysis/optimisation, geospatial data, GPU/HPC scaling, or research-grade software engineering.
How to apply
All formal details—eligibility, employment conditions, salary, required documents, and the submission portal—are on the official page. If the project description above resonates with you, I strongly encourage you to apply via that link:
https://www.chalmers.se/en/about-chalmers/work-with-us/vacancies/?rmpage=job&rmjob=14476&rmlang=UK
The application deadline is February 23, 2026.
If you have questions about the research direction and fit, feel free to reach out (my contact details are on the official page). I’m happy to briefly discuss the project and what kind of background would be a good match.
I’m really looking forward to seeing applications from people who want to combine mathematics, AI, and geometry in a way that produces city-scale, analysis-ready digital twins.