Monthly Register — January 2025

Watch a Dubai submission go out and the last thing into the portal is usually the model. Someone exports the federation, zips it, finds the slot, uploads it — and then goes back to the work that actually decides whether the pack is accepted, which is checking that the sheet index agrees with the PDFs. The model goes in because the gate asks for one. The sheets go in because the sheets are what gets stamped. Few of the people doing this would describe it as BIM at permit. That is nonetheless the name it has been given.

January 2024 was the month the Gulf permit conversation stopped being optional on a large share of live jobs. Not “are we doing BIM?”, which had been a marketing sentence for a decade, but “does the permit pack include a model?” Saudi programmes and Dubai submissions were where that question landed hardest, and the industry treated it as a switch being thrown. A year on, the honest reading is more modest, and more useful.

What moved

What typically changed at permit stage is the channel and the checklist. Not the object that gets reviewed.

More jobs now have a portal slot, a transmittal line or a pre-submission list that asks for a model file. The BIM execution plan is attached. A clash report travels with the pack, usually as a PDF. Level-of-development language appears in the cover letter. Someone is named as the information manager for the submission. None of that is nothing. A gate that used to accept only sheets now often refuses a pack with no model attached, consultants have added an export to the ritual, and client project-management offices have started asking to see the federation before they will countersign.

What typically did not change is what the reviewer marks, and what the site builds from.

The stamped thing is still a set of sheets. Title blocks, revision clouds and a suitability code on a PDF remain the language of the permit, and coordination comments still come back as mark-ups on those sheets. The model that satisfied the portal is often a stripped or frozen federation — good enough to open, not the live design model, and not the issued set the contractor will plot from. Ask on an ordinary job whether the permit model and the construction model are the same object and you tend to get a pause before the answer, which is answer enough. A year of BIM at permit did not, by itself, make them one object.

The new-cover problem is the one document controllers can still name without a circular in hand. Same drawing pack. Same sheet index. A cover that now says the submission is model-based. The PDF has a new first page, the register has a new container, and the plotter has not noticed.

Jurisdictions still do not share a filing system. Dubai, Abu Dhabi and the Saudi municipal channel are not one portal with three logos, whatever the bid narrative implied, and a project that proceeds as though they were will fail at the second gate even if it sailed through the first. That is hardly a discovery. It is the condition the anniversary made harder to ignore, because more jobs are now trying to reuse one BIM pack across more than one authority, and because the people who warned about it at tender are rarely the people still on the job at submission.

ISO 19650 vocabulary travelled with the mandate talk. Suitability codes turned up on title blocks that had never carried them, which is a genuine gain and cost almost nothing. The container discipline those parts assume travelled more slowly, as it always does, because it needs somebody's time rather than a template change. On a large share of jobs the issued set was still a folder of PDFs with a model file parked beside it.

That is the movement worth recording from the year. Not a keynote, not a product. A year in which the gate asked for a model and the job, more often than the slides admitted, answered with a drawing pack and an export.

What changed in the standards

Very little, if you insist on a new number. Nothing arrived in the ISO 19650 series in January to change what a BIM execution plan should be pointing at, and the UK BIM Framework did not issue a replacement text to mark the Gulf anniversary. The working documents remain the published parts already in circulation. They were not new this month; they were the language in which permit-stage BIM was being described, which is a different fact and a less exciting one.

The anniversary is the standards event, thin as it is. Where a jurisdiction has told the industry that a model, or an information container, belongs in the permit, that sentence changed the checklist. On the evidence of ordinary jobs it did not replace the drawing pack as the thing that carries the stamp. Anyone claiming otherwise is describing an execution plan rather than a submission.

One thing to watch

Two weeks from now the slides will be closed, and the test will not be whether anyone can still date the thing to January 2024. It will be whether document controllers are still arguing about what “BIM at permit” actually named: a file in a portal, or a published state that is the same object the job will build from. That argument belongs beside federated common data environments on a gigaproject, and beside the anatomy of building from a superseded drawing — not beside a mandate recap.

So watch the cover sheet. If the model and the sheet index disagree and the stamp sits on the sheets, the job has not moved as far as the execution plan claims. If the permit federation cannot be opened by a party that was not in the authoring licence, the gate has accepted a file it cannot reuse. If three authorities are being fed one pack with the logos swapped, the second rejection is already in the post.

And if you are the appointing party, ask for one line in the appointment: either the permit submission and the construction issued set are the same container in the same state, or somebody writes down why they differ and who reconciles them. It costs a paragraph now. It costs considerably more as a claim later.