Berlin's Digital Archives Are Full of Ghost Images — Officials and Experts Want That Fixed Now

Berlin's public-facing digital systems are carrying thousands of redundant, duplicated image files — and officials, archivists and tech professionals say the problem has become too expensive and too disruptive to ignore any longer. The Senatsverwaltung für Stadtentwicklung, Bauen und Wohnen confirmed earlier this year that its digital planning portal alone was running with an estimated 40 percent image redundancy rate, slowing document retrieval times and complicating the housing permit workflows the SPD-led coalition has prioritised as central to easing the city's housing crisis.

The timing matters. Berlin is in the middle of an ambitious push to digitise planning applications, resident registration and transport documentation — a programme partly funded through the Digitalpakt infrastructure budget that the federal government extended through 2025. With those funds now expiring and a transition period underway, city departments are under pressure to deliver clean, usable digital environments rather than bloated systems carrying years of duplicated visual data.

What the Experts Are Saying

Technology professionals working with Berlin's public sector describe the duplicate image problem as a legacy of rapid, uncoordinated digitisation. When multiple departments separately scanned physical records, uploaded photographs from public events, or migrated older databases, they frequently pulled the same source image into different folders under different file names — creating redundancy that is invisible to end users but costly in storage and processing terms. The Zentraler IT-Dienst Berlin, the city's central IT management body, has been working since late 2024 on a unified asset management protocol that would flag and remove duplicate files across municipal servers.

Archivists at the Landesarchiv Berlin on Eichborndamm in Reinickendorf have noted the specific challenge of historical photograph collections, where automated deduplication tools frequently misidentify near-identical images — say, two photographs taken seconds apart at the same event — as exact duplicates and delete one. Getting that distinction right requires human oversight layered on top of algorithmic tools, a resource-intensive combination that smaller district offices in Neukölln and Lichtenberg say they currently lack the staffing to provide.

The startup community around Prenzlauer Berg and the Factory Berlin campus on Rheinsberger Strasse has taken notice. Several companies specialising in computer vision and document management have pitched the Senatsverwaltung on AI-assisted deduplication pipelines, arguing that machine learning can now distinguish between genuinely identical files and near-duplicates with accuracy above 95 percent — a threshold archivists say would be acceptable for non-historical administrative documents.

Costs, Timelines and What Comes Next

Storage costs for municipal Berlin digital infrastructure are not trivial. City budget documents reviewed for the 2025 fiscal year indicated that data storage contracts with third-party providers had grown year-on-year, with unoptimised image libraries cited among the contributing factors. District-level IT managers have said internally that a successful deduplication drive across all twelve Bezirke could reduce those costs meaningfully, though no single authoritative figure for potential savings has been published.

The BVG, Berlin's public transport operator, has its own parallel concern. The authority's internal documentation system — used to manage everything from infrastructure inspection photographs on the U-Bahn network to bus depot records — has reportedly been audited for redundancy as part of a broader efficiency review tied to the operator's multi-year investment programme running to 2030.

Practically, residents and businesses interacting with digital planning tools through the Berlin Service Portal on berlin.de are unlikely to notice a change immediately. The deduplication work happens on the backend. What officials and advocates say should eventually improve is search speed, document accuracy and the reliability of planning records — all of which matter directly to anyone waiting on a housing permit, a building inspection sign-off or a business licence in a city where bureaucratic delays have become a persistent political grievance. The Senatsverwaltung has indicated a phased review of its image asset holdings is scheduled to conclude before the end of the third quarter of 2026.