Undergraduate student in Geodesy Engineering at Universitas Diponegoro, Semarang, Indonesia. Working on geospatial machine learning for land-use classification and environmental monitoring.
- Multi-sensor remote sensing (Sentinel-1 SAR, Sentinel-2 multispectral)
- Supervised classification for land-cover mapping
- Geospatial data engineering with Python (rasterio, geopandas, GDAL)
- Spatial databases (PostgreSQL + PostGIS)
End-to-end remote sensing pipeline mapping industrial land cover in the Karawang industrial corridor, West Java. Combines Sentinel-1 SAR and Sentinel-2 multispectral imagery via XGBoost classification to produce 1,063 industrial polygons over a 122 km² area of interest. Includes documented human-in-the-loop refinement and an explicit limitations section.
Headless SfM Engine
Async REST orchestrator for resource-constrained binary tools, demonstrated with COLMAP for Structure-from-Motion reconstruction. Single-tenant FastAPI service with subprocess timeout, payload validation, and VRAM-bounded preprocessing.
Stack: Python, rasterio, geopandas, scikit-image, XGBoost, ESA SNAP
- Deep learning for semantic segmentation in remote sensing
- Reproducible geospatial ML pipelines (test design, validation rigour)
- NASA ARSET training modules in applied remote sensing
| Domain | Tools and Methods |
|---|---|
| Remote sensing | Sentinel-1/2, Landsat, ENVI, ER Mapper, ESA SNAP |
| Geospatial ML | XGBoost, Random Forest, SVM, scikit-learn, scikit-image |
| Geospatial Python | rasterio, geopandas, GDAL, shapely |
| Databases | PostgreSQL, PostGIS |
| Geodetic computation | Map projections, least squares adjustment, satellite geodesy |
Universitas Diponegoro — Semarang, Indonesia
Bachelor of Engineering in Geodesy Engineering (in progress)
- Email (academic): devonrama@students.undip.ac.id
- Email (personal): devonrama.w@gmail.com
- Location: Semarang, Indonesia