Adelaide Designers and Archivists Tackle the Duplicate Image Problem This Week

South Australian design studios, university archives and government communications teams spent this week confronting one of the most persistent frustrations in digital asset management: the duplicate image. Whether catalogued twice after a server migration, re-uploaded by separate staff members, or copied across shared drives without a naming convention, duplicate images quietly bloat storage budgets, slow production workflows and create legal headaches when licensing details are attached to only one version of a file.

The timing matters. Adelaide's Lot Fourteen precinct on North Terrace — home to the Australian Space Agency headquarters, several defence-tech startups and a growing cluster of creative industries — has added dozens of organisations to its tenant roster over the past 18 months. Each new team arriving at the old Royal Adelaide Hospital site brings its own image library. When those libraries merge or share a common server, duplication compounds fast.

What Changed This Week

On Tuesday, the University of South Australia's library and digital collections unit — based at the City West campus on Hindley Street — circulated an internal guidance note to affiliated research centres about adopting perceptual hashing tools for duplicate detection. Perceptual hashing compares images by their visual fingerprint rather than their file name or metadata, meaning a JPEG and a PNG of the same photograph are flagged as duplicates even if they carry completely different file names. The technique has been standard in large tech companies for years but is only now filtering into mid-sized institutional workflows in Australian cities.

Separately, a collective of independent graphic designers operating out of the Glenside co-working precinct on Fullarton Road held a practical session on Wednesday walking members through open-source duplicate-finder software. Attendance was capped at 24 participants and the session sold out, which organisers said reflected genuine demand rather than novelty. The session covered not just detection but deletion protocols — specifically, how to decide which version of a duplicated file to keep when resolution, colour profile and embedded metadata differ between copies.

State government communications, meanwhile, is navigating its own version of this problem. The Department for Trade and Investment, which coordinates marketing assets across programs including the hydrogen jobs plan and the Olympic Dam expansion project, manages image libraries that draw from multiple agencies and contracted photographers. Consolidating those into a single, deduplicated repository has been a standing item in digital asset management reviews for at least two years.

Why the Numbers Are Striking

Industry figures from the Digital Asset Management Society's 2025 annual survey — covering organisations across Australia and New Zealand — found that duplicated files typically account for between 20 and 35 percent of total storage in unmanaged digital libraries. For a mid-sized organisation storing 10 terabytes of visual content, that translates to roughly 2 to 3.5 terabytes of redundant data. At current Adelaide data-centre pricing of approximately $80 to $120 per terabyte per month for managed storage, the direct cost of unaddressed duplication runs into thousands of dollars annually before accounting for staff time spent searching through redundant files.

The problem is sharper for organisations working on AUKUS-related defence contracts at Osborne Naval Shipyard and the Lot Fourteen precinct, where version control on technical imagery is tied to contractual and security obligations, not just production convenience. A duplicated image with conflicting metadata in that environment is not merely inefficient — it can represent a compliance risk.

For Adelaide organisations looking to act now, practitioners recommend starting with a full audit using free tools such as dupeGuru or open-source scripts built on the ImageHash Python library before investing in enterprise solutions. Establishing a clear file-naming convention — ideally incorporating date, project code and photographer credit — prevents new duplicates from accumulating after the initial clean-up. The Glenside collective plans to run a follow-up session in August focused specifically on metadata standardisation. Details are expected on its mailing list later this month.