PhD Position: Earth Observation, Data Science, and AI for Poverty Estimation

About the role

Four-year fully funded PhD at Chalmers University of Technology’s Division of Data Science and AI, jointly supervised with the University of Gothenburg and Linkoping University. The project trains deep-learning models on Sentinel-2 satellite imagery to estimate living conditions across Sub-Saharan Africa and benchmark how well communities track the Sustainable Development Goals. The research is funded by the Swedish Research Council and contributes to open-source poverty estimation software. Starting salary is SEK 34,550 per month. Deadline: 13 June 2026.

Responsibilities

• Train and validate deep-learning models on Sentinel-2, Pleiades, and Landsat imagery to predict household-level poverty indicators.
• Apply explainability methods to identify which satellite features drive predictions.
• Develop and maintain open-source poverty estimation pipelines.
• Publish in venues such as NeurIPS, ICML, and remote-sensing journals.
• Participate in weekly cross-institutional lab meetings.
• Contribute up to 20% teaching or related duties in years 4 or 5 if extended.

Requirements

• Master’s degree (120 ECTS) in computer science, data science, statistics, or a related quantitative field.
• Strong Python skills and experience with deep-learning frameworks (PyTorch or similar).
• Solid grounding in deep learning and computer vision.
• Fluent written and spoken English.
• Ability to work in an interdisciplinary, international team.
• Physical presence in Gothenburg required; valid residence permit by the start date.

Nice to have

• Experience with satellite imagery or geospatial libraries (Google Earth Engine, GDAL, rasterio).
• Background in spatio-temporal modelling or causal inference.
• Prior publications or open-source contributions.

How to apply

Apply through the Chalmers online recruitment portal using the link above. Submit a CV, cover letter explaining your research interest, degree transcripts, and contact details for two referees. The expected start date is flexible within 2026.