Surveyors tune in to 3D laser scanning

Advances in 3D laser scanning and digital photogrammetry are transforming the way refurbishment and renovation projects are delivered in BIM. James Gregory, head of BIM at surveying consultancy Plowman Craven, explains the impact these technologies can have during design and tendering.

What types of BIM project can benefit from detailed 3D surveys?

Scan-to-BIM is typically applied on refurbishment or conservation projects, where there is a need for an accurate model of existing structures that are not easily modelled by hand in software.

However, surveyors are now widening their scope to apply 3D laser scanning on new build projects, where they want a record of existing adjacent buildings, buildings to be demolished and the topography of the site. These aspects can have an impact on verified views and other planning work.

What scanning technology is typically deployed?

We try to scan in every way possible, on the ground and in the sky, to capture all the terrestrial and aerial data needed to produce a complete point cloud dataset for the project and the foundation for a highly detailed BIM model.

Terrestrial tripod-based laser scanners, such as those produced by Faro and Leica, are highly accurate and able to record interior spaces and exterior facades, or hard-to-reach mechanical and electrical installations or plumbing, concealed behind walls or ceiling tiles.

Mobile scanners, worn as a type of backpack or moved on a trolley with wheels, can speed up the process, but do not capture data as accurately. In addition, high resolution photos taken by drones fitted with high resolution cameras can be stitched together by software in a process called photogrammetry. This can be used to can generate a second point cloud, covering all hard-to-reach external areas of the building.

How does the ‘to-BIM’ part of Scan-to-BIM work?

A point cloud is just a mass of “dumb” dimensional points, modelling is what gives it intelligence. 3D components in BIM should effectively “know” what they are and where they are in the model and have data and attributes tagged to them, providing designers with much more detailed information to work with than static 2D plans, sections and elevations.

The point cloud file is first transformed into smart geometric components in BIM authoring tools such as Autodesk Revit or Bentley Microstation. This can be a time-consuming task, especially when working with heritage buildings that don’t follow straight lines. As a result, we have developed various bespoke tools to help speed up the process.

Algorithm-based modelling tools, such as Edgewise Building, by ClearEdge3D, can streamline the process by automatically recognising and extracting items from the point cloud, such as pipework or structural steel, and transforming them into 3D components.

When the model is complete, non-graphical information can be added, then accessed by simply clicking on the object. For example, selecting a ceiling can bring up information on the presence of asbestos drawn from the asbestos report, while a tree can have arboricultural data added, such as the tree type, water demand, and root area.

The more intelligence you build in, the greater the need to push it away from Revit to avoid slowing down the software. We tend to display it in a model viewer, such as Navisworks Freedom, which is free to download making it easy for anyone to jump in and take a look around.

What are the benefits of Scan-to-BIM for clients and designers?

When applied during pre-design, Scan-to-BIM enables the client to create a detailed asset record of all existing information to send to the designers (architects and engineers).

Rather than receiving a tender pack with thousands of drawings and surveys covering environment, ecology, unexploded ordinance etc designers get a solid model dataset to work from as a “single source of truth”.

Scan-to-BIM can significantly de-risk the entire project. I formerly worked as a structural engineer on refurb projects where the use of 2D plans meant we had to assume the structure behind walls and ceilings, then during construction we would uncover various anomalies and had to redo the design several times which cost the client a fair sum, simply because there wasn’t enough data at the outset.

A holistic dataset can save the client a significant sum of money in design development and iterations.

Arguably the biggest advantage is having this strong visual model to help understand and explore the building without having to repeatedly visit the site. This is important on sites considered dangerous, such as those with asbestos-ridden buildings or access challenges.

A detailed 3D model can immediately highlight any challenges for planning to give designers a significant head start.

We supply BIM models embedded with fully classified data, compliant with relevant standards, to help enable design rather than limit it. A 2D plan in a tender pack will give an idea of where the walls and window locations are, but not the specifics. By modelling in 3D we can give a better representation of where each feature sits and its adjacencies.

Is this the most common survey procurement route taken by clients?

Only those clients who appreciate the value a BIM survey can bring, when compared to a traditional survey. Unfortunately, the reality is that many clients do not have the technical knowledge or “BIM Champion” to be able to adequately specify BIM surveys and so more often we steer and support the client in the survey scoping phase.

It’s more common for surveyors to be approached at a later stage in design by the lead architect, in support of the client, with a specification for a model of the building.

This can happen anywhere from RIBA Stage 0 to Stage 3, before or after planning. Much of the time architects work in 2D before planning then move to 3D afterwards.

What’s the future of Scan-to-BIM?

The further we explore and innovate, the more we realise that the possibilities are endless. Some projects have started purchasing as-built surveys to be used for the ongoing operation of the asset.

We are now beginning to investigate how we turn these static models into dynamic models, integrating elements like live monitoring data (taken from sensors in the actual building) and bringing the model to life.