PhD Position in Scalable AI for Earth Observation

This PhD position at Forschungszentrum Jülich’s Jülich Supercomputing Centre (JSC) focuses on advancing AI and advanced computing techniques for Earth observation (EO). The role is based in the Simulation and Data Lab for AI and Machine Learning for Remote Sensing, where you will investigate how AI workflows can efficiently handle large-scale, heterogeneous EO data. Your research will explore trade-offs between predictive performance, computational efficiency, and scalability, leveraging high-performance and cloud computing environments. The position also includes opportunities to explore hybrid quantum-classical approaches. You will collaborate with interdisciplinary teams, including geoscientists, computing specialists, and international partners, to validate and integrate your methods into operational workflows. Enrollment in the PhD programme at the University of Iceland is included, with annual research stays in Reykjavík.

• Conduct literature reviews and formulate research questions at the intersection of AI, Earth observation, and advanced computing.
• Develop and evaluate AI methods for processing multisource Earth observation data, including geospatial foundation models and representation learning.
• Design reproducible data and experimental workflows for model training, adaptation, and evaluation on selected Earth observation applications.
• Investigate MLOps pipelines for scalable training and inference on parallel, distributed, and GPU-accelerated systems.
• Benchmark developed approaches against established methods, assessing predictive performance, generalizability, and computational efficiency.
• Collaborate with domain experts and computing specialists to validate methods and integrate research prototypes into shared software workflows.
• Publish findings in scientific journals, present at international conferences, contribute to open-source research software, and prepare a doctoral thesis.