When the Drawings Don't Exist: Using LiDAR to Document What Decades of Modifications Have Buried

Priya had been the facilities manager at a 1960s commercial building in inner Brisbane for six years before anyone asked her a question she couldn't answer.

A structural engineer needed to know the column grid on level three. Not approximately. Exactly. The building had been through three fitout cycles, a mezzanine addition in the 1990s, and a facade upgrade that nobody had fully documented. The original drawings, if they existed, were not in the council archives. The previous owner had no records. The architect who handled the 1994 mezzanine had retired and his files were gone.

Priya spent two weeks chasing paper. She found fragments: a partial floor plan from 1987, a structural detail that might have been from the original construction, and a set of hydraulic drawings that had nothing to do with what she needed. In the end, the engineer had to work from measurements taken by hand, room by room, with a laser distance meter and a notepad.

That process took four days and produced a sketch set that everyone involved privately acknowledged was probably not quite right.

This situation is not unusual. It is, in fact, the norm for a large proportion of Australia's existing building stock.

The Documentation Gap in Existing Buildings

Buildings constructed before the 1990s were rarely documented in a format that survives decades of ownership changes, council record purges, and office relocations. Even buildings with original drawings often have those drawings rendered unreliable by subsequent modifications, none of which were formally documented or, if they were, were never consolidated into a current set.

A 2019 survey by the Australian Institute of Architects found that more than 60 percent of practitioners working on existing buildings reported encountering incomplete or absent documentation at some point in the preceding twelve months. For buildings older than 40 years, the figure was higher.

The consequences are practical and financial. Engineers designing remediation work on an undocumented structure are forced to make assumptions. Assumptions drive contingency allowances. Contingency allowances inflate budgets. And when the work begins and the assumptions turn out to be wrong, the cost of variations can exceed the original contingency entirely.

For heritage buildings, the problem is compounded. A masonry wall that looks uniform from the outside may contain timber lintels, brick arches, infill panels from different eras, and structural modifications that were never drawn. Without knowing what is actually there, any intervention carries risk.

What LiDAR Actually Does

LiDAR stands for Light Detection and Ranging. A scanner emits laser pulses in a 360-degree pattern and measures the time each pulse takes to return after bouncing off a surface. From millions of these measurements, taken from multiple positions around a building, the scanner builds a point cloud: a three-dimensional record of every surface the laser reached.

A modern terrestrial LiDAR scanner captures somewhere between 500,000 and two million points per second. A typical floor of a commercial building can be fully scanned in two to four hours. The resulting point cloud is accurate to within two to three millimetres at distances up to 30 metres.

That accuracy matters. Hand measurements, even careful ones, accumulate error. A wall that is measured as 6,200mm in one room and 6,180mm in the adjacent room creates a 20mm discrepancy that an engineer has to resolve somehow, usually by averaging or by assuming one measurement is wrong. A point cloud doesn't accumulate that kind of error. Every point is referenced to a common coordinate system established by survey control targets placed before scanning begins.

The output is not a drawing. It is a dataset. What you do with that dataset determines its value.

From Point Cloud to Structural Model

The workflow from scan to usable deliverable has several stages, and each one requires decisions about what the data is actually for.

Registration is the first step. Scans taken from multiple positions are aligned into a single coordinate system using the survey targets. Software such as Leica Cyclone or Trimble RealWorks handles this automatically, but the quality of the result depends on how well the targets were placed and how many overlapping scan positions were used. For a complex building with irregular geometry, this stage can take as long as the scanning itself.

Cleaning removes noise: scanner artefacts, moving objects captured mid-scan, reflections from glass surfaces. A raw point cloud of a busy commercial floor will contain people, trolleys, and temporary equipment. These are filtered out before the data is used for modelling.

Extraction is where structural information is pulled from the point cloud. For a building documentation project, this means identifying columns, beams, walls, slabs, and connections. In a well-maintained building with exposed structure, this is relatively straightforward. In a building where everything is covered by ceilings and cladding, the scanner can only document what it can see. Knowing the limits of the data is as important as knowing what the data contains.

Modelling converts the point cloud into a BIM model, typically in Revit or a similar platform. This is not automatic. A technician traces structural elements from the point cloud, building a parametric model that can be used for analysis. The level of detail in the model depends on what the project requires. A condition assessment might need only a massing model with accurate floor-to-floor heights and column positions. A remediation design might need beam sizes, connection details, and slab thicknesses verified by additional investigation.

For the structural engineer, a BIM model derived from a point cloud is fundamentally different from a model built from original drawings. It reflects what was actually built, not what was intended. That distinction matters enormously when the two diverge.

Where Scanning Changes the Engineering Outcome

Consider a facade remediation project on a 1970s high-rise. The engineer needs to know the fixing locations for new cladding panels, which means knowing the position and size of the structural frame behind the existing facade. Without drawings, the options are: open up the facade at multiple locations to measure directly, or scan.

Opening up a facade is expensive, disruptive, and only gives you point measurements. Scanning the interior face of the facade, combined with GPR scanning to locate embedded steel, gives you a continuous record of the frame geometry across the entire elevation. The engineer can design fixings to the actual structure, not to an assumed structure.

At the Victory Hotel project, TRSC used LiDAR scanning on a 170-year-old building where the original construction drawings had never existed in any formal sense. The building had been modified repeatedly across its life, and the relationship between visible elements and structural elements was not obvious from inspection alone. The point cloud, combined with material investigation, allowed the engineering team to build a model of the existing structure that was accurate enough to support both the structural analysis and the heritage conservation strategy. You can read more about that project at [/preview/trsc/projects/victory-hotel](/preview/trsc/projects/victory-hotel).

For the 140 William Street project in Melbourne, a 39-page investigation report was built on a foundation of accurate existing conditions data. When you are advising on a heritage facade with a $230,000 engagement scope, the cost of getting the geometry wrong is not a rounding error. Accurate documentation is not a luxury in that context. It is the basis on which every subsequent decision rests. See [/preview/trsc/projects/140-william-street](/preview/trsc/projects/140-william-street).

The Practical Limits

LiDAR is not a solution to every documentation problem, and it is worth being clear about what it cannot do.

A point cloud documents surfaces. It does not document what is behind those surfaces. A column that appears in the point cloud as a 400mm square element might be reinforced concrete, or it might be a steel section encased in concrete, or it might be a hollow form that was filled in a later modification. The scan tells you the geometry. It does not tell you the material properties, the reinforcement configuration, or the condition of the internal structure.

This is why LiDAR works best as part of an integrated investigation, not as a standalone exercise. At TRSC, point cloud data is typically combined with GPR scanning to locate embedded reinforcement, half-cell potential testing to assess corrosion risk, and material sampling for laboratory analysis where the condition of the structure is uncertain. The geometry from the scan feeds into the structural model. The material data from the investigation populates the model with properties that allow actual analysis.

Scanning also requires access. A basement carpark that is in use 24 hours a day presents a different access challenge than an empty floor of a vacant building. Scheduling scanning around operational constraints adds time and cost to the programme. For buildings where access is genuinely difficult, the scan may need to be broken into multiple sessions, which increases registration complexity.

And point clouds are large files. A full building scan can run to hundreds of gigabytes. Managing, storing, and sharing that data requires infrastructure that not every project team has in place. TRSC uses SharePoint-integrated documentation workflows that allow point cloud data and derived models to be accessed by all project stakeholders, but this requires agreement on formats and access protocols at the start of the project, not after the scan is complete.

What Owners and Architects Should Know Before Commissioning a Scan

If you are considering LiDAR scanning for a building documentation project, a few practical points are worth keeping in mind.

First, define the deliverable before the scan. A point cloud is not a drawing set. If you need drawing-ready documentation, you need to specify that at the outset, because the modelling work that converts a point cloud into a usable drawing set is a separate scope item with its own cost and programme.

Second, understand what the scan cannot see. If critical structural elements are concealed behind finishes, the scan will document the finishes, not the structure. Agree with your engineer upfront on where additional investigation is needed to supplement the scan data.

Third, think about ongoing use. A point cloud and BIM model created for a remediation project has value beyond that project. It becomes the as-built record for future maintenance, future modifications, and future investigations. If the model is built to a standard that supports long-term asset management, the cost of creating it is amortised across every future project that draws on it. If it is built to minimum scope for a single project, that opportunity is lost.

Finally, consider the alternative cost. The four days Priya's engineer spent measuring level three by hand, producing a sketch set that everyone knew was approximate, cost money and introduced risk into every decision that followed. A LiDAR scan of the same floor would have taken half a day and produced data accurate to three millimetres. The comparison is not always that stark, but for complex buildings with multiple modifications and no reliable documentation, the economics of scanning are usually straightforward.

The Record That Should Have Always Existed

There is something slightly ironic about the fact that the technology to create accurate, permanent, three-dimensional records of buildings now exists and is accessible, while the buildings that most need documenting are the ones that were built before anyone thought to keep proper records.

But that is the situation. A significant portion of Australia's commercial and heritage building stock exists without reliable documentation, and every year that passes without documenting it is another year of modifications, maintenance work, and undocumented changes that make the gap wider.

Scanning does not recover what was never recorded. It creates a new baseline: an accurate record of conditions as they exist today, from which future decisions can be made with confidence rather than assumption.

For building owners managing assets without reliable drawings, for architects working on alterations to undocumented buildings, and for engineers who need to analyse structures they cannot fully see, that baseline is not a technical nicety. It is the foundation of every decision that follows.

TRSC integrates LiDAR scanning and BIM documentation into structural investigation and remediation projects across Queensland, New South Wales, and Victoria. For buildings where the drawings do not tell the full story, or do not exist at all, the process of creating an accurate record is often the most valuable first step. More information is available at [https://trsc.com.au](https://trsc.com.au).