Canberra's Duplicate Image Problem: How the Capital Stacks Up Against Global Peers Tackling Digital Archive Bloat

Canberra's major public institutions are sitting on digital archives clogged with redundant imagery — the same photograph stored dozens of times under different file names, across multiple servers, at significant cost to storage budgets and archival integrity. The Australian National University and the National Archives of Australia, both of which manage collections running into the tens of millions of digital assets, are among the organisations now being pushed to adopt automated duplicate-detection systems that peer cities adopted years ago.

The pressure is practical. Federal agencies in the capital are subject to the Australian Government Records Interoperability Framework, which sets expectations around clean, deduplicated digital holdings. With the ACT government also managing a growing catalogue of urban planning imagery — documenting everything from the Gungahlin Town Centre's rapid expansion to light rail corridor development along Flemington Road — the volume of near-identical files accumulating in shared drives has become an administrative headache that carries real dollar costs.

What Other Cities Are Doing

Amsterdam's municipal archive, Stadsarchief Amsterdam, completed a two-year deduplication project in 2024 that reportedly cleared roughly 18 percent of redundant image files from its public-facing collection — freeing storage capacity and making search results more accurate for researchers. Singapore's National Heritage Board introduced AI-assisted duplicate flagging across its digital repositories in early 2025, integrating perceptual hashing tools that can identify visually identical images even when file formats or compression levels differ. Wellington's Archives New Zealand piloted a similar program targeting its photographic holdings in 2023, with the goal of reducing the administrative labour required to manually reconcile duplicate records before public release.

Canberra does not yet have a single coordinated equivalent. The National Library of Australia's Trove platform, which aggregates digitised content from institutions across the country, has internal deduplication logic built into its ingestion pipeline, but the library has publicly acknowledged — in documentation available on its website — that identical images sourced from multiple contributing institutions can still surface as separate records. For researchers at the University of Canberra's Faculty of Arts and Design, which relies heavily on Trove for visual research, that means manual cross-checking remains part of the workflow.

The Local Bottleneck

Inside the ACT government's own infrastructure, the problem shows up most visibly in planning and land management. The Environment, Planning and Sustainable Development Directorate holds aerial and ground-level photography spanning decades of suburban growth, particularly for corridors like Belconnen and the Molonglo Valley. Officers working on development applications in areas such as Denman Prospect have raised concerns internally — documented in agency workflow reviews — about time lost reconciling duplicate site images lodged by different contractors through the same planning portal.

The ACT government's Digital Strategy 2025–2028, published last year, nominates data quality and deduplication as priorities but does not set a specific target date or budget allocation for image-specific programs. That is a contrast with Wellington's Archives New Zealand, which committed NZ$2.1 million over three years to its digital asset rationalisation effort — a figure that appeared in the New Zealand government's 2023 Budget estimates documents.

For the public servants who make up the bulk of Canberra's workforce — and who are often the ones managing these internal document systems from offices in Barton, Parkes, and the city's Civic precinct — the stakes are mundane but real. Duplicate images slow search tools, inflate cloud storage costs billed to agency budgets, and create version-control confusion when files are retrieved for ministerial briefings or public release under freedom of information requests.

Procurement cycles in the ACT government typically run on 12-to-18-month timelines, meaning any decision made this financial year on image management tooling would not see deployment until mid-to-late 2027 at the earliest. Institutions that want to move faster — ANU's libraries are one example — can act independently, and the university has existing relationships with digital preservation consortia through the Council of Australian University Librarians that could accelerate adoption of open-source deduplication tools. Whether the broader federal precinct catches up to Amsterdam or Singapore's pace, or whether the work gets absorbed into the next round of whole-of-government ICT refresh planning, is the question agencies will likely be answering through late 2026 and into 2027.