How Johannesburg's Public Records Ended Up Flooded With Duplicate Images — And Why It Matters Now

The City of Johannesburg's digital infrastructure is carrying a quiet but costly burden: thousands of duplicate images embedded across public databases, municipal planning portals and heritage record systems, many of them traceable to a single decade of rushed, poorly coordinated digitisation between roughly 2010 and 2020. The problem did not arrive overnight. It accumulated through budget cuts, departmental silos and a succession of IT migration projects that moved files without cleaning them first.

The issue matters now because Joburg is in the middle of its most ambitious data consolidation push in years. The ANC-DA coalition governing Gauteng has placed municipal efficiency at the centre of its joint mandate, and the City of Johannesburg Metropolitan Municipality is under pressure to demonstrate that its internal systems are fit for purpose. Duplicate image files slow processing speeds, inflate storage costs and, in the case of planning and land-use management, can trigger errors when two near-identical photographs of a site are treated as separate evidentiary records.

Where the Problem Took Root

The origins lie in the early 2010s, when several City departments independently contracted vendors to digitise paper records. The Johannesburg Development Agency, headquartered on St Andrews Road in Parktown, ran its own scanning programme for inner-city regeneration project files. The City's Spatial Planning and Urban Design department, based at the Metropolitan Centre on Braamfontein's Loveday Street, operated a parallel effort. Neither system spoke to the other. Files were uploaded to separate servers, and when a central data warehouse was eventually attempted around 2017, automated scripts merged repositories without deduplication protocols in place.

The Soweto Heritage Trust, which manages documentation for culturally significant sites across Orlando, Meadowlands and Kliptown, flagged the issue internally as early as 2019, when archivists noted that some photographic records of Orlando Towers and the Walter Sisulu Square precinct appeared multiple times under different file names and metadata tags. The duplication was not malicious — it was the residue of well-intentioned but uncoordinated work.

The Joburg Metrorail reform programme, which picked up pace after the Passenger Rail Agency of South Africa entered a restructuring period, added another layer. Station condition reports photographed by different inspection teams were uploaded to a shared drive without a master naming convention, generating a secondary layer of duplicates that sat alongside the original municipal archives.

The Cost of Cleaning Up

Storage is not cheap at municipal scale. Commercial cloud storage rates in South Africa as of mid-2026 sit at roughly R0.23 per gigabyte per month for standard tiers, according to published pricing from local providers. A database carrying even 500 gigabytes of redundant image files generates unnecessary expenditure month after month. More critically, the City's Geographic Information Systems unit — which supports everything from infrastructure maintenance mapping in Sandton to flood-risk modelling in low-lying parts of Diepsloot — loses processing efficiency when queries return duplicate results.

The Gauteng Department of e-Government has been working since early 2025 on a province-wide deduplication framework under its Digital Transformation Strategy. That programme targets all three metros — Johannesburg, Tshwane and Ekurhuleni — and is scheduled to complete a first audit phase by the end of the third quarter of 2026. For Johannesburg specifically, the audit is expected to cover planning, public safety and heritage image repositories held across no fewer than eleven distinct departmental systems.

For residents and businesses, the practical stakes are real. A property developer submitting a building plan at the City's Development Management office on Jorissen Street, Braamfontein, can face processing delays if the cadastral images attached to a site's historical file are flagged as conflicting duplicates requiring manual review. Community organisations in Soweto applying for heritage site maintenance grants have reported similar friction when photographic evidence submitted in prior applications resurfaces as an apparent anomaly in the system.

The deduplication audit, once complete, is supposed to produce a single authoritative image for each record, with redundant files archived rather than deleted outright — a safeguard against losing genuinely distinct photographs that happen to look similar. City departments have been advised to freeze new bulk uploads to shared repositories until the audit's first phase concludes. The window to fix this is narrow: a second major data migration, linked to the City's broader SAP system upgrade, is pencilled in for early 2027.