About Me
During his childhood, he fell in love. It was not that kind of ordinary love; it was different — “the sky love.” His eyes were always looking towards the sky, watching clouds and stars, and that fascination eventually became a decision to become a space engineer.
From Space Engineering to Earth Observation
Islam studied Aerospace Engineering at Istanbul Technical University, where he worked on CubeSats, sounding rockets, and CanSats, and represented the region at the University Rover Challenge in the Utah desert. That hands-on systems background pulled him toward a broader question: how do you actually use space assets to understand what’s happening on Earth?
That question led him to the ESPACE program (Earth Oriented Space Science and Technology) at the Technical University of Munich, and from there into the German Aerospace Center (DLR)‘s Microwaves and Radar Institute, where he built his PhD around a hybrid of physics and machine learning for SAR.
PhD and Current Research
He completed his PhD at ETH Zurich (2021-2025), Chair of Earth Observation and Remote Sensing, supervised by Prof. Dr. Irena Hajnsek and co-advised by Dr. Kostas Papathanassiou and Dr. Ronny Hansch. His thesis, Hybrid AI-Physical Modelling in Interferometric SAR for Enhanced Parameter Retrieval, combined data-driven learning with electromagnetic scattering models to improve forest height retrieval from TanDEM-X InSAR and to correct penetration bias in X-band InSAR DEMs — work validated over tropical biomes (Gabon, Amazon) and the Greenland Ice Sheet, and explicitly noted as relevant to ESA’s Biomass mission. He also combined TanDEM-X interferometric coherence with GEDI lidar waveforms to map forest structure dynamics at continental scale across the Brazilian Amazon, and co-developed the DLR/ESA PolInSAR Training Course infrastructure on ESA’s MAAP platform.
Since April 2025, he has been a postdoctoral researcher at the Universität der Bundeswehr München (UniBW), developing AI-based Earth observation methods that lean on foundation models for SAR and optical data: adapting large vision foundation models to remote sensing tasks, building a hybrid ship detection and segmentation system (YOLO11 + SAM2) for maritime surveillance in Sentinel-1 imagery, and developing a pipeline to generate OpenStreetMap-style vector maps from very-high-resolution optical imagery.
Research Interests
Synthetic Aperture Radar (SAR) and InSAR, forest height and biomass retrieval, hybrid physics-informed machine learning, multi-mission data synergy (TanDEM-X, GEDI, Sentinel-1), foundation models for Earth observation, and — where it all started — spacecraft dynamics and embedded systems.
Still looking towards the sky, just processing what it sends back down.