PDF to Revit conversion — flat drawings rebuilt as a coordinated model

What we start from

Drawings that exist only as flat output. Exported PDFs with no source file behind them, raster scans of paper prints, photographs of drawings pinned to a wall, or a mixed set of all three at different scales and from different decades. You do not need to sort the material before sending it, because working out which sheet supersedes which is part of the job.

Why this is not the same as converting a DWG archive

A DWG archive, however contradictory, contains measurable geometry. A PDF set does not. Every wall thickness, every level, every setting-out dimension has to be read off the sheet as a written number or inferred from what surrounds it, and each inference is a decision that could be wrong. The modelling time is comparable. The interpretation time is not, which is why this is priced as its own service.

Where the effort actually goes

Reconciliation. Two sheets of the same floor that disagree by 150 mm, a section that does not match the plan it was cut from, a level datum that changed between packages without anyone renaming the file. We resolve those conflicts where the material allows it, and where it does not we flag them in writing rather than modelling a guess and letting you discover it on site.

Twenty-three years of building design and construction is what makes that judgement possible. Knowing which of two contradictory drawings is likely to be the built condition is not a Revit skill.

What you get

A Revit model at LOD 200–350 as agreed, built in your template, with levels, grids and wall types set up so the model carries the project rather than illustrating it. Plans, sections and elevations are generated from the model and coordinated, not drafted independently. Every element whose position was inferred rather than read is listed in the accuracy note, so the model can be relied on where it is reliable and checked where it is not.

Accuracy and what it depends on

A conversion is only as accurate as what it was converted from. From a dimensioned vector PDF we work to the dimensions given. From scans and photographs we work to stated dimensions first and scaled geometry second, and we say so. If a site visit or a fresh measured survey is what the project actually needs, we will tell you that instead of selling a model built on unreliable input.

Can you build a Revit model from scanned PDF drawings?

Yes. Vector PDFs, raster scans of paper prints and photographs of drawings are all workable input. Dimensions written on the drawing take priority over scaled geometry, and where the source cannot answer a question we flag it in writing rather than modelling a guess.

How accurate is a model converted from a PDF?

As accurate as what it was converted from, and no more. The limits are written into a short accuracy note that comes with the model, listing every element whose position was inferred rather than read from a dimension.

Why does this cost more than converting a DWG archive?

Because a DWG contains measurable geometry and a PDF does not. Every dimension has to be read as a written number or inferred from context, and the interpretation time is what separates the two services.

What level of detail do you deliver?

LOD 200 to 350, agreed before we start. The model is built in your template, with levels, grids and wall types set up so that it carries the project rather than illustrating it.

What comes back at the end?

An RVT model, plans, sections and elevations generated from that model and coordinated with each other, an IFC export, a PDF set, and the accuracy note. Native files, not just outputs.