How Zurich's Digital Archive Problem Became a City-Wide Headache: The Duplicate Image Crisis Explained

Zurich's public institutions are sitting on a digital mess. Across cantonal government departments, the Zentralbibliothek Zürich, ETH Zürich's image archive and the Stadtarchiv on Alfred-Escher-Strasse, hundreds of thousands of digitised photographs, maps and administrative documents exist in multiple copies — sometimes three or four versions of the same file stored in entirely separate systems, each consuming server space and complicating search results for researchers and civil servants alike. The problem has a name in Swiss archival circles: Bilderduplikation, or duplicate image proliferation, and after years of low-level grumbling it has finally forced a reckoning.

The timing matters. Switzerland's federal Data Protection Act revision came into full force in September 2023, tightening rules on how public bodies store and catalogue personal data. Digitised images — portraits, street scenes, identity documents — fall squarely within its scope. Institutions that cannot demonstrate clean, deduplicated records risk compliance failures. For a canton that built its post-Credit Suisse reputation partly on administrative competence and transparency, that exposure is uncomfortable.

How the Duplication Built Up Over Two Decades

The roots go back to the early 2000s, when individual Zurich institutions began digitising independently, with no shared standard and no central registry. ETH Zürich launched its own image database for scientific photography. The Museum für Gestaltung on Ausstellungsstrasse maintained a separate catalogue of graphic design archives. The Stadtarchiv digitised civil records. The Zentralbibliothek on Zähringerplatz ran its own platform. Each project was defensible on its own terms; collectively, they created a siloed patchwork.

When the Canton of Zurich launched the Digitale Verwaltung initiative around 2018, analysts conducting a preliminary audit found that shared historical photographs — particularly those documenting the city's industrial belt along the Limmat and construction records from Zürich-West — appeared in as many as four separate institutional databases with different metadata, different access rights and, in several cases, conflicting copyright attributions. No authoritative figure for the total number of duplicate files has been made public, but internal assessments cited in cantonal budget discussions in 2024 pointed to storage inefficiencies running into the terabyte range across the affected systems.

The financial dimension is real. Zurich's cantonal IT infrastructure costs have grown steadily, and redundant storage is a contributing factor that procurement officials have flagged in successive annual reports to the Kantonsrat. Server consolidation projects at the Rechenzentrum des Kantons Zürich in Rümlang have repeatedly identified duplicate media assets as a drag on efficiency, though the canton has not published a standalone cost figure attributable solely to image duplication.

What a Fix Actually Requires

Deduplication is not simply a matter of running a script. The technical challenge is compounded by the fact that the same image may have been cropped, recoloured or reformatted differently by each institution before ingestion, meaning automated hash-matching — the standard tool for identifying identical files — frequently fails to catch near-duplicates. Specialist perceptual hashing tools, which compare images visually rather than byte-by-byte, are more effective but require significant computational resource and human review to adjudicate competing metadata claims.

The Zentralbibliothek and the Stadtarchiv have reportedly begun a joint working group to align their digitisation standards, a process that archivists describe as painstaking. ETH Zürich's library system, which operates under federal rather than cantonal authority, presents a separate coordination challenge entirely. Any solution that spans both cantonal and federal institutions requires negotiated data-sharing agreements — the kind of inter-governmental groundwork that Swiss direct democracy makes slow but ultimately durable once completed.

For Zurich residents and researchers, the practical next step is to watch for the canton's planned update to its open-data portal, data.stadt-zuerich.ch, expected later in 2026. Officials have indicated the refresh will include improved deduplication filters for image searches. Institutions planning to access historical Zurich imagery for commercial or academic purposes should in the meantime cross-check across multiple portals — the Zentralbibliothek's e-rara platform, the Stadtarchiv's online catalogue and ETH's e-pics database — and document provenance carefully to avoid inadvertent copyright complications under the revised data protection framework.