Ex Situ: A Spatial Index of 200,000 Displaced Cultural Artifacts
Ex Situ is an open-source geospatial index that maps museum artifacts as arcs from their origin site to the institution that holds them. The map currently covers 200,552 artifacts across 9 collections, 194 territories, and 6,679 origin sites. It started as a solo MA thesis in 2022 and the creator, Hüseyin Burak Yel, has been building it since.
The infrastructure is designed as an indexer, not a hoster. Artifacts stay linked to their institutional source. Images are kept as URLs pointing to the source institutions. No content is copied or mirrored. The project is “connective tissue between archives that were never designed to speak to each other.”
The data model is intentionally flat to avoid encoding problematic colonial taxonomies. Many museums still categorize items under labels like “Islamic art,” “Asian art,” “ethnological collections”, and “religious objects.” These are categories that reflect the collector’s worldview, not the object’s origin. Ex Situ sidesteps this by prioritizing geographic provenance over institutional categories.
The map renders arcs at three zoom levels. At the global view (zoom 0–4), you see territory-to-institution arcs. At the regional view (zoom 5–9), city/site to institution. At the artifact level (zoom 10+), you get precise coordinates for individual objects. The result is a tool that works for both broad pattern recognition and detailed research.
Currently indexing from 9 collections including the Metropolitan Museum of Art, the Victoria and Albert Museum, the Staatliche Museen zu Berlin, and the Art Institute of Chicago. It only pulls from institutions with open-access APIs, which means the map unavoidably underrepresents museums that keep their data closed. A commenter on Hacker News pointed out that Paris and Moscow look innocent on the map, because their data is not publicly available.
The built-in search lets you filter by origin site or collection, and there is an MD export for researchers who want to download provenance data for a filtered set. There is also a public API at
exsitu.app/api/museum-objects/geospatialthat returns GeoJSON arc data filtered by zoom level, bounding box, institution, or origin country.Because the data comes exclusively from Western and Euro-American institutions, the scope is a deliberate constraint. The creator notes that expanding to Japanese, Russian, or other museum networks depends entirely on whether those institutions provide open-access APIs. The technical pattern for adding a new resolver is documented in the ETL pipeline (fetch, map, geocode, insert), but each new institution means adapting to whatever API format they use.
The project is fully self-hostable and AGPL-3.0 licensed. Stack: Next.js + Deck.gl + MapLibre GL JS on the frontend, Strapi + PostgreSQL/PostGIS on the backend, Python for the ETL pipeline. No proprietary cloud dependencies, no third-party tracking.
The takeaway: Ex Situ turns the relationship between an object’s origin and its current location into a visible, queryable spatial index. It does not host the data or duplicate institutional records. It draws the lines. The map reveals which museums have open data and which do not almost as clearly as it reveals where artifacts came from.
Related TMFNK Content
- The Castle Map: 3,708 Fortresses, One Open-Data Atlas Another open-data spatial atlas, built from Wikidata and Wikimedia Commons instead of institutional APIs.
- SearchWhole.Earth: The Digitally Resurrected Whole Earth Catalog A different kind of open-data catalog; curating interesting things into a browsable index.
- Infrastructure Maps: Four Ways to See the World’s Power Grid A map that makes hidden infrastructure visible, a similar impulse to Ex Situ’s provenance arcs.
Crepi il lupo!