The hidden cost of duplicate images: what the numbers say about Canberra's digital storage problem

Canberra's public sector runs on documents, data, and digital assets. It also, according to records management specialists and government IT procurement data, runs on enormous quantities of duplicate images — the same photograph, scan, or graphic saved multiple times across departmental servers, shared drives, and cloud platforms. For a city where federal and territory agencies employ roughly 85,000 public servants, the storage overhead is not trivial.

The issue has gained sharper focus this financial year. The ACT Government's Digital Strategy, which sets targets through to 2028, flags data hygiene as a priority for agencies operating under the territory's shared services framework. Duplicate or near-duplicate image files are among the most persistent categories of digital waste identified in storage audits, partly because automated backup systems create copies without any human decision-making involved.

What the numbers actually look like

Storage costs are the clearest place to start. Amazon Web Services lists its S3 standard storage tier at around USD $0.023 per gigabyte per month — a figure that compounds quickly when an organisation holds hundreds of thousands of unmanaged image assets. For context, a single high-resolution photograph from a modern DSLR camera can run to 25 megabytes in RAW format. Multiply that across a communications team that has been uploading event photography for a decade, and a single department can accumulate terabytes of image data before anyone notices.

At the Australian National University in Acton, the library and digital collections teams have grappled with exactly this problem across their institutional repositories. ANU's library holds records stretching back to the university's founding in 1946, and digitisation programs over the past 15 years have introduced systematic duplication — the same archival photograph ingested at different resolutions, under different file names, by different project teams working without a unified asset management protocol.

The University of Canberra, based in Bruce, has faced comparable challenges as its research data management policies have evolved. UC's Research Office has pushed toward FAIR data principles — Findable, Accessible, Interoperable, Reusable — which implicitly require de-duplication as a precondition for proper cataloguing. Without it, search tools return redundant results, storage budgets inflate, and staff waste time manually cross-checking files.

Commercial de-duplication software now advertises storage savings of between 20 and 80 percent depending on the data type, with image libraries typically falling in the 30-to-50 percent range. Applied to a mid-sized ACT government directorate running, say, 50 terabytes of mixed digital assets, that suggests between 15 and 25 terabytes of recoverable space — worth thousands of dollars annually in avoided cloud storage fees, before factoring in the labour cost of managing bloated asset libraries.

Why Canberra's local context makes this urgent

The ACT Legislative Assembly's Public Accounts Committee has previously scrutinised shared services IT spending, and digital asset management sits within that broader accountability frame. Territory government agencies centralise many IT functions through Shared Services, based on Constitution Avenue in the city centre, which means storage inefficiencies in one directorate can propagate across the shared infrastructure that neighbouring agencies rely on.

Gungahlin and Belconnen — the two fastest-growing residential corridors in the territory — have both seen new community facilities open in recent years, each generating their own streams of event imagery, planning documents, and communications assets that flow back into central government repositories. The Gungahlin Town Centre redevelopment alone has been the subject of sustained photographic documentation across multiple agencies, with planning, transport, and community services each maintaining their own image sets independently.

For organisations wanting to address the problem practically, the first step most IT governance frameworks recommend is a content audit using perceptual hashing tools, which identify visually similar images even when file names and metadata differ. The Australian Government's Digital Transformation Agency publishes guidance on data management that touches on deduplication as part of its broader cloud framework, though specific image-library protocols remain the responsibility of individual agencies.

The next ACT Government ICT expenditure review is expected in the second half of 2026. Agencies that have not conducted a systematic image audit before then may find the question raised for them — along with a bill that could have been smaller.