How Zurich's Digital Archives Ended Up Drowning in Duplicate Images — and What Went Wrong Along the Way

Zurich's public digital archives contain tens of thousands of photographs that nobody can find, because the same image has been uploaded, relabelled, and stored multiple times across at least four separate municipal systems. That is the core problem now forcing Stadt Zürich's Stadtarchiv to undertake its most significant data-cleaning operation in two decades.

The issue did not emerge overnight. It is the cumulative result of a decade-long digitisation push that accelerated sharply after 2015, when the city began scanning historical planning documents to support the Wohnungsnot housing shortage debate. Different departments — Amt für Städtebau, Tiefbauamt, and the communications office at Stadthaus Zürich — each commissioned their own ingestion pipelines, with no shared metadata standard and no central deduplication layer. By 2022, internal audits flagged the problem formally for the first time, but budget negotiations delayed any remediation work.

The Roots of the Problem: Parallel Systems, Parallel Chaos

The city's image infrastructure currently spans at least four platforms: the Stadtarchiv's own DAM (digital asset management) system, a separate repository maintained by ETH Zürich's professorship for digital preservation under a long-standing civic partnership agreement, a cloud-based media library managed by the Zürich Tourism organisation on Bahnhofstrasse, and a legacy FTP-based store that the Amt für Kultur has never fully migrated away from. Each system grew independently, ingesting content from contractors, photographers, and scanning bureaux without cross-checking what already existed.

The results are measurable. A 2024 internal review — details of which circulated among archivists at a closed workshop held at the Zentralbibliothek on Zähringerplatz — found that roughly 34 percent of image files across the main Stadtarchiv system were either exact duplicates or near-duplicates differing only in resolution or file format. That figure climbed to over 40 percent in the construction-documentation subset, precisely the category that had grown fastest during the Hardturm and Zürich-West redevelopment years.

Storage costs, while not publicly disclosed by the city, follow a pattern familiar across European municipal archives: the Swiss Federal Archives in Bern have noted in published guidance documents that unmanaged duplication typically inflates storage expenditure by 20 to 30 percent over a ten-year horizon. For Zurich, with its premium data-centre pricing, the financial case for remediation has become difficult to ignore.

Why the Fix Has Taken So Long

The delay reflects something structural about how Zurich runs large digital projects. Swiss direct democracy means budget line items above certain thresholds require either Gemeinderat approval or, in some cases, a public vote. A remediation project involving procurement of new deduplication software and a multi-year data governance contract crossed that threshold, creating a political queue that only cleared in late 2025 when the relevant Stadtrat committee approved the expenditure as part of a broader digital infrastructure package.

The chosen approach draws on perceptual hashing — a technique that identifies visually similar images even when filenames and metadata differ — combined with a human review layer staffed by archivists from the Stadtarchiv's team at Alfred-Escherstrasse. A pilot covering the pre-1950 photographic collection ran from January to March 2026, with the full rollout scheduled to begin in the fourth quarter of this year.

For researchers, journalists, and city planners who rely on these archives daily — the Stadtarchiv serves an average of several hundred external queries per month — the practical effect of the current mess is slower search results, inconsistent licensing metadata, and occasional retrieval of a lower-resolution version of an image when a higher-quality original exists somewhere in the system. The upcoming remediation will not solve every problem at once, but it will, administrators say, establish a single canonical record for each image and retire redundant copies to a quarantine store before eventual deletion.

The broader lesson for other Swiss cities watching from Geneva, Basel, and Bern is straightforward: digitisation without governance is just analogue chaos at faster speeds and lower visibility. Zurich is now doing the unglamorous work of cleaning up what a decade of well-intentioned but uncoordinated progress left behind.