How Singapore's Digital Records Crisis Quietly Built Up: The Story Behind Duplicate Image Replacement

Singapore's push to consolidate its sprawling public sector image databases has exposed a problem that accumulated over more than a decade: tens of thousands of duplicate photographs, scanned documents, and archival images sitting across incompatible systems, driving up storage costs and slowing access for citizens and researchers alike. The effort to systematically replace and rationalise those duplicates is now a live operational challenge for agencies running under the Smart Nation and Digital Government Group framework.

The issue did not arrive overnight. It is the direct product of Singapore digitising its public records in waves — first under the 2006 eGovernment Masterplan, then again under successive iterations of the Infocomm Media Development Authority's digitalisation drives. Each wave added files. Few were cross-checked against what already existed. By the time the current Smart Nation 2.0 push began consolidating databases in earnest from 2023 onward, the redundancy problem had become structural.

A Patchwork of Databases Left Behind

The National Archives of Singapore, housed at Canning Rise, holds the country's definitive public record — photographs, maps, government documents stretching back to the colonial era. But parallel collections grew up independently at the National Library Board on Victoria Street, at statutory boards, at Ministry of Communications and Information sub-agencies, and within the Housing and Development Board's own estate documentation systems. Each body digitised what it had, in formats that made sense at the time, without a shared deduplication standard.

HDB alone manages image records for more than one million residential units across estates from Tampines to Buona Vista. When estate officers photograph flats for renovation permits, valuation checks, or defect inspections, those images feed into systems that were never designed to talk to each other seamlessly. The result: multiple versions of the same photograph stored under different file names, timestamps, and metadata tags, each treated by its home system as a distinct record.

The IMDA's Digital Infrastructure Blueprint, published in February 2024, flagged data redundancy as one of three priority problems for Singapore's public sector cloud migration. Moving legacy databases to Government Commercial Cloud — the whole-of-government platform operated in partnership with Amazon Web Services, Google Cloud, and Microsoft Azure — required agencies to audit what they were actually holding. Those audits surfaced the scale of the duplicate problem for the first time in aggregate.

Why Replacement Is Not Simply Deletion

Removing a duplicate sounds trivial. It is not, because Singapore's public records law under the National Library Board Act and the National Archives of Singapore Act requires that any deletion of an official record be authorised through a formal disposal schedule. An image that exists in two places is still, legally, two records until a records manager signs off on which version is canonical. That process takes time and trained personnel — a resource the public sector is competing for against private sector demand.

The practical cost is real. Government Commercial Cloud storage is billed at rates competitive with commercial benchmarks, and redundant image files — particularly high-resolution archival scans that can run to hundreds of megabytes per file — add up fast across dozens of agencies. Industry estimates for enterprise deduplication savings typically range between 30 and 60 percent of storage footprint, though Singapore government figures have not been published.

The Government Technology Agency, known as GovTech, which sits at the centre of the Smart Nation architecture, has been piloting automated deduplication tools since mid-2025 at three agencies, though GovTech has not publicly named which agencies are involved or what results the pilots have produced so far.

For anyone dealing with these systems on the ground — public servants managing estate records in Jurong East, researchers pulling archival images from the National Library at Victoria Street, or businesses applying for permits that require photo documentation — the practical advice for now is to cross-check image submissions against existing records before upload, use standardised file-naming conventions aligned with the IMDA's metadata framework, and flag suspected duplicates through agency feedback channels rather than allowing them to multiply further. The rationalisation effort is underway, but the finish line is not close.