Doctoral Candidate: 3D Scene Reconstruction and Multi-Modal Sensing for Digital Twins of Stone-Built Heritage

About the role

The Chair of Geosensors and Engineering Geodesy at ETH Zurich is looking for a doctoral candidate to work on 3D mapping, scene reconstruction and multi-modal sensing for information-rich digital twins of heritage buildings. The position is part of the collaborative project HC3, which aims to build interoperable 3D digital twins of stone-built monuments together with the sensing and monitoring methods that provide their data, so that deterioration can be modelled and predicted under climate change. You will work alongside two doctoral students from the Chair of Physical Chemistry of Building Materials and a postdoctoral researcher, with dr. Tomislav Medic as your direct supervisor.

Key responsibilities

• Plan and carry out multi-sensor mapping campaigns of monuments, including LiDAR and imaging technologies on ground-based and UAV platforms
• Calibrate, register and fuse point clouds and imagery into unified 3D representations (point clouds, meshes, neural radiance fields, 3D Gaussian splatting)
• Develop deep learning methods that segment scenes into individual building elements (stone blocks) for automated digital twin (HBIM) generation
• Develop machine learning methods that infer building material properties such as stone deterioration type and severity from multi-modal data and time series
• Integrate the data into digital twins and extended reality, validate methods against conservation expert assessments and laboratory experiments, publish results and contribute to teaching upon agreement

Required profile

• MSc degree (completed or about to be completed) in geomatics, remote sensing, computer vision, robotics, computer science or a related field
• Interest in interdisciplinary work with material scientists and conservation practitioners
• Excellent English; French or German is an advantage

Required skills

• Solid background in 2D/3D perception (point cloud and image processing, photogrammetry, 3D reconstruction) and parameter estimation, including machine and deep learning
• Good programming skills in Python or C++
• Hands-on experience with laser scanning (LiDAR) or imaging systems is an advantage
• Willingness to do fieldwork on monuments is a must

What we offer

• A fully funded four-year doctoral position in an international, dynamic team
• State-of-the-art sensing equipment and a project with potential to evolve beyond its initial scope
• ETH Zurich's infrastructure, continuing education and a salary according to ETH regulations

Applications are submitted online only, with a CV, cover letter describing a past result and your research vision, two references and a copy of the MSc diploma.