How Zurich's Digital Archives Ended Up Full of the Same Image Twice — and What Comes Next

Zurich's municipal digital archive, maintained under the Stadt Zürich umbrella, is carrying thousands of duplicate image files — the accumulated result of at least a decade of fragmented digitisation campaigns, incompatible software migrations, and departmental silos that were never properly reconciled. The problem came into sharper focus this spring when Stadt Zürich's IT directorate began a systematic audit ahead of a planned platform consolidation scheduled for late 2026.

The issue matters now because the consolidation carries a price tag and a deadline. The city has budgeted for infrastructure upgrades across its digital services division, and duplicate files inflate storage costs, slow retrieval systems, and complicate the metadata tagging that underpins public-access portals used by journalists, architects, and civil society groups. Every redundant file is a small drag on a system that is about to be asked to do more, not less.

How the Backlog Built Up

The roots of the problem run back to the early 2010s, when individual departments — from Stadtentwicklung to the Amt für Städtebau — began digitising their own photograph and document collections independently, using whatever software was available to them at the time. When the city's central Stadtarchiv Zürich, based on Alfred-Escher-Strasse, later tried to consolidate those collections, it inherited files that had already been uploaded multiple times under different naming conventions, different resolution settings, and sometimes different copyright designations.

ETH Zürich's Chair of Information Science has documented similar fragmentation problems in European municipal archives more broadly, noting that ad-hoc digitisation without a central taxonomy creates compounding redundancy that grows harder to untangle the longer it sits. Zurich was not unusual in following this path — but the city's ambition to build a genuinely searchable, AI-assisted public record system by 2027 means the reckoning has arrived earlier than it might have otherwise.

The situation was further complicated by the migration away from a legacy content management system in 2019. Files transferred during that window were often duplicated automatically by the migration tool when it encountered naming conflicts, a known issue that was flagged at the time but never fully resolved because the remediation work was deprioritised during the Covid-19 years of 2020 and 2021.

What the Audit Found — and What Happens Now

Internal documents reviewed as part of the current audit process show that duplicate rates in some departmental sub-collections run as high as 30 percent by file count, though the citywide figure across the full archive is estimated to be lower. The Stadtarchiv itself holds records stretching back centuries; the duplication problem is concentrated in material digitised after 2010.

The practical consequences are not trivial. Storage on the city's managed cloud infrastructure costs money per gigabyte per month. Beyond cost, duplicate entries create confusion for users of the public-facing portal, where the same photograph of, say, the Lindenhügel or the Rathausbrücke can appear as three separate search results with three different metadata records — sometimes contradicting each other on date or photographer credit.

The remediation plan, as understood from city communications, involves running deduplication algorithms across the full archive before the new platform goes live, then manually reviewing flagged cases where metadata differs between apparent duplicates. A tender for the technical work was published on simap.ch, Switzerland's public procurement platform, in April 2026. Bids closed in mid-June.

For residents and professionals who rely on the archive — historians working in the Kreis 1 reading rooms, urban planners pulling historical maps for projects in Altstetten or Oerlikon, journalists checking image rights — the practical advice is straightforward: flag any search result that appears to be a duplicate using the portal's feedback function, which the Stadtarchiv has confirmed remains active throughout the transition. Each flagged duplicate shortens the manual review queue. The consolidated system, if the timeline holds, should be operational by the first quarter of 2027.