Bristol's Battle Against Duplicate Images in Public Records: How It Compares to Amsterdam, Edinburgh and Toronto

Bristol City Council has been working since early 2025 to address a significant but unglamorous problem in its digital infrastructure: thousands of duplicate images clogging planning portals, heritage registers and public-facing property databases. The effort, centred on the council's online planning system hosted through the Bristol Planning Portal on College Green, has so far identified more than 14,000 redundant image files across active planning applications dating back to 2017.

The issue matters now because Bristol, like dozens of mid-sized European cities, has been migrating legacy paper records into searchable digital formats. That process, accelerated by the UK government's Digital Planning Reform agenda launched in 2023, has exposed how badly duplicated content inflates storage costs, slows search results and, in some cases, obscures the most recent and legally binding versions of submitted documents. For residents in Stokes Croft or Bedminster trying to scrutinise a planning application affecting their street, finding the right image in a list of near-identical files is not a trivial inconvenience.

Bristol's approach has involved a two-track system. The council's Digital Services team, working out of the City Hall annexe on Wilder Street, has deployed automated hash-matching software to flag visually identical files. A secondary manual review process, handled in partnership with Knowle West Media Centre—a long-standing community digital organisation in south Bristol—has been piloting public-assisted verification, asking local volunteers to confirm whether flagged duplicates are genuinely redundant or represent different stages of a planning revision. That community layer is relatively unusual among UK cities of Bristol's size.

What Other Cities Are Doing

The comparison with Amsterdam is instructive. The Dutch capital, which began its own digital archive deduplication programme in 2022 under its Stadsarchief Amsterdam, processed roughly 2.3 million image records in an 18-month period and reported a 31 percent reduction in archive storage requirements—figures published in the Stadsarchief's 2023 annual report. Amsterdam used fully automated pipeline tools with no public-participation element, which was faster but drew criticism from heritage groups who argued some historically significant near-duplicate images were incorrectly purged.

Edinburgh took a different tack. The City of Edinburgh Council's Records Management team, working alongside the National Records of Scotland, chose a slower, more cautious approach to its Historic Environment portal, imposing a freeze on deletion of any image file until at least two staff members had independently reviewed it. That process, still ongoing as of June 2026, has kept error rates low but created a backlog of roughly 8,400 unresolved flagged files, according to a progress update published on the council's website in May 2026.

Toronto's approach, implemented through its open-data portal on the city.toronto.ca platform, leaned heavily on machine learning classifiers trained on architectural photograph metadata. The results were fast but the city acknowledged in a 2025 audit report that the classifier had a false-positive rate of around 7 percent on construction-phase images—a meaningful margin when documents carry legal weight in planning disputes.

Bristol's Current Position and What Comes Next

Bristol sits somewhere between Edinburgh's caution and Amsterdam's pace. The council has confirmed that no confirmed-duplicate image will be deleted without a retention log entry and that all removed files will be held in a cold-storage archive on servers managed through NHS Bristol, North Somerset and South Gloucestershire's shared IT infrastructure arrangement—a partnership that has existed since 2021 and extends to several non-health council data functions.

The council has not yet published a completion timeline for the full deduplication exercise, and the scope does not currently extend to the Bristol Museums, Galleries and Archives digital collections held at the Bristol Record Office on Smeaton Road, which holds its own separate backlog of an estimated 6,000 flagged image pairs from digitisation work carried out between 2019 and 2022.

For residents with active planning applications, the practical advice from the council's Digital Services guidance page is to always upload images with distinct filenames that include the application reference number and a version date. That small step, the guidance notes, is the single most effective way to prevent a file entering the duplication pipeline in the first place—and it costs nothing.