Curriculum Vitae
Erik Duncan
Geospatial ML Researcher & Data Scientist
Founder of Geocontinua Technologies and Senior Faculty Specialist at the University of Maryland, working across geospatial AI, remote sensing, and applied earth observation research.
Experience
Founder and Managing Director
Feb 2026 — PresentGeocontinua Technologies
- Founded a geospatial AI consultancy, leveraging experience in research and end-to-end geospatial projects.
- Current projects involve developing operational ML products for humanitarian demining organizations in Ukraine and Iraq.
Senior Faculty Specialist
Jun 2024 — PresentUniversity of Maryland, Department of Geographical Sciences
- Korea Aerospace Administration (KASA) CAS500-4 Satellite Program (Apr 2026 — Present): led development of new, multi-stage compositing algorithms for the CAS500-4 satellite and coordinated with KASA on co-development and integration into their data processing pipelines.
- Co-I, U.S. Air Force / Department of Defense STTR (Jul 2024 — Apr 2026): “Revolutionizing Space-Based ISR through Decentralized Systems and In-Orbit ML Computing for Near-Real-Time Intelligence.” Adapted machine learning capabilities to run onboard optical imaging satellites under strict computational and power constraints; led development of high-speed segmentation algorithms and the code for onboard deployment; developed, validated, and adapted hyperspectral, high-resolution, and very-high-resolution (VHR) models. Achieved first successful in-orbit demonstration — Jan 2025.
Faculty Specialist
Oct 2021 — Jun 2024University of Maryland, Department of Geographical Sciences
- Co-I, NASA Rapid Response (Mar 2023 — Jul 2024): “Artillery Craters and Unexploded Ordnance Mapping in Ukraine using High Resolution Satellite Imagery.” Large-scale data processing and deep learning implementation in conflict-affected areas of Ukraine; delivered final report in 2024, with an associated scientific publication.
- IARPA SMART Program: project member of the University of Maryland / Kitware team, focused on VHR image processing and machine learning for cloud compensation in multi-source VHR imagery.
- NASA Harvest + NASA LCLUC (Mar 2022 — Present): agricultural damage assessment for the NASA Harvest project, liaising with FAO and the Ukrainian government on delivery of agricultural damage information; established data-sharing agreements informing Ukrainian agricultural policy (2023); multi-year NASA LCLUC-funded remote sensing monitoring of Ukraine.
Satellite Mapping Analyst
Jun 2018 — Jul 2019Geological Survey of Canada | Natural Resources Canada, Ottawa, Canada
- Estimated carbon and water volumes in Arctic permafrost landforms using mathematical, remote sensing, and vector-based modeling approaches.
- VHR image processing and analysis for geotechnical hazards along the Tuktoyaktuk Highway project in the Yukon Territory.
Skills
Machine Learning
- Coding, deploying, tuning, and validating deep learning frameworks: UNET, Ultralytics (YOLO), ResNet, Random Forest, and others.
- Proficient in translating technical AI development into non-technical language for stakeholders.
- Researched a Neural Architecture Search genetic framework for deep learning model development (2025).
Data Science
- Python-based scientific computing and geospatial data processing — raster/vector analysis, array and tabular workflows.
- Large-scale data pipeline development leveraging multiprocessing and HPC/cluster super-computing.
- Statistical and mathematical modeling for physical/environmental estimation problems.
- Time-series and change-detection analysis across multi-year satellite imagery archives.
- Model validation, benchmarking, and evaluation against operational performance requirements.
Other Skills
- Academic research and technical reporting: Co-I on federally funded programs (NASA, DoD/Air Force), including end-to-end grant execution.
- Stakeholder management with government and international partners (FAO, Ukrainian government, KASA).
- Client-oriented problem solving and business development.
- Public speaking and technical communication (AGU, GEOINT, ESA Living Planet, and others).
Education
M.Sc. Geoinformatics
Sep 2019 — Sep 2021University of Copenhagen, Department of Geosciences
Thesis: “Detecting Artillery Craters in Conflict Zones with Semantic Segmentation Convolutional Neural Networks.”
B.Sc. Geomatics, with Honours
Sep 2015 — Nov 2018Carleton University, Department of Geography and Environmental Studies
Publications
- 2023Detection and mapping of artillery craters with very high spatial resolution satellite imagery and deep learning
- 2025Monitoring cropland cultivation, abandonment, fallowing, and recultivation dynamics with regard to conflict intensity in war-affected Ukraine
Awards & Honors
- 2025Departmental Award — Outstanding Research; contributed to NASEO (Neural Architecture Search for Earth Observation Onboard Processing), presented at EDHPC.
- 2024FAO meeting with Ukrainian government, Rome; US GEOINT Conference presentation, Kissimmee, FL; NASA SCERIN meeting, Chania, Greece; Departmental Award — Outstanding Research.
- 2023Oral presentation at AGU 2023, CA; NASA Harvest meeting and presentations, Strasbourg, France; NASA LCLUC meeting, College Park, MD; World Bank ICAS IX Conference, Washington, DC; Trans-Atlantic Training Workshop, Brno, Czech Republic; GICHD Innovation Conference, Geneva, Switzerland.
- 2022ESA Living Planet Symposium poster presentation on UXO hazard mapping; AGU 2022 poster presentation, Chicago, IL.
- 2021Global University Climate Forum, Team “Yes We Canteen.”
- 2020NASA Space Apps Challenge regional winner: project “AfterFire”; European Space Agency Launchpad participant.
Languages
English — NativeFrench — B1/B2Dutch — A2/B1