Skip to main content

Getting a reality check – putting the ‘as-built’ into digital construction

By [email protected] - 9th May 2018 - 09:32

We are currently seeing another shift in the use of GIS, not only to mapping in 3D and real-world visualisation, but also greater convergence with the traditional ‘CAD’ market and Building Information Modelling (BIM). Managers of infrastructure and other location-based assets now want all data in one place, so they don’t have to deal with different sets of data held in different systems.

What is 4D BIM? 3D models are widely used throughout the AEC and MEP industry sectors and provide comprehensive design data. 4D BIM contains more information about the project that helps address the ‘what if’ scenarios by enabling tools such as clash analysis and works scheduling. By driving efficiencies in its ability to disseminate the right information to the right people in the right format at the right time, 4D BIM is crucial in minimising project unknowns and delays.

Key challenge

As-built data for new build construction and renovation is increasingly the requisite extension to the 4D BIM collaborative process and represents a key differentiator. The key challenge is that ‘Reality Capture’ technologies create unwieldy point clouds that are difficult to manage and share and lack of intelligence when combined with design data.

Adding real-world as-built information throughout the 4D BIM process can eliminate potential risk, by providing snapshots by trade at every stage of the construction process. As well as providing a way to accurately prove compliance, it also allows for comparison between actual and planned design milestones. As such, clashes can easily be flagged and progress monitored. This means more accurate and collaborative information sharing, with any amendments to the bill of materials or time scales being made available throughout the process.

Margins are very tight on construction projects, so every workflow process needs to be costed and evaluated. The addition of frequent as- built data capture may seem like adding unnecessary extra work into an already stretched schedule, but advances in Reality Capture techniques mean that the mass capture of site data is no longer time consuming, nor invasive. The use of laser scanners or photogrammetry from UAVs allow entire sites to be captured in hours, with a 3D point cloud delivered the same day. By utilising the existing ground control networks, these point clouds can be millimetre-accurate representations of exact site conditions at that exact moment.

Start small

When contractors see savings, efficiencies and the reduced risk during the build, technology such as this becomes an integral part of the construction process. It is imperative to start small, maybe on one trade or part of the build, as there is nothing like success to promote adoption

Methods for utilising this raw data have been traditionally cumbersome and offset the efficiencies in the capture. Using point clouds directly is often not achievable, especially for large scale projects, as they can be too large to work with effectively. It can take weeks to manually model directly from the point cloud into a useful format, a process which, in itself, is highly subjective and requires highly trained personnel. By the time a final model is ready, construction would have progressed and the model outdated.

Spatial analytics is one of the main capabilities of GIS that enables designers and project managers to analyse the impact of their design in the conceptual stage … a process dubbed Geodesign. Taking this into consideration, the importance of providing the ‘as-built’ data in the right data structure offers obvious efficiencies in terms of data operability and integration with third party platforms hosting the design data. Importantly, it also provides a base for which analytical processes can be used to harvest more intelligence from such rich sources of data.

Bridging the gap

Technical specialists have, for some time, grappled with the problem of bridging the gap between Reality Capture and Digital Construction. To address it, software experts at Pointfuse decided to completely rethink the way point cloud data has been processed. The solution was to convert points on the same plane into a series of separable meshes based on their geometry, which can be referred to as a segmented mesh model. The resulting software automatically converts point clouds into high fidelity, intelligent mesh models that are highly optimised. By reducing the size of working data by a hundred-fold, it makes the models significantly easier to use and share, as compared to the original point cloud.

Key to the Pointfuse process, however, is removing human involvement in the modelling process, thereby eliminating a significant bottleneck and removing any chance of subjectiveness. By delivering the same result over and over again, it makes Pointfuse models repeatable and directly comparable … an ideal characteristic for accurate project validation.

Once converted into a segmented mesh model, it is possible to select and classify groups of objects within the Pointfuse software, either to isolate relevant features, such as walls and floors, or remove noise such as people and debris. When exported using the industry standard ifc. or fbx. formats, this intelligent mesh model maintains this classification so it is easily recognisable within a BIM environment. The result is a cleaner mesh, in a recognised format, that contains only those features that are required, with none of the noise inherent to point cloud data.

Clash detection … and avoidance

So, with Pointfuse it is possible to create an accurate, highly detailed model in just a few hours, and with regular, up-to-date models achievable in a single day, it is now possible to employ as-built data as part of a standard 4D BIM workflow. This allows detailed clash detection analysis by comparing design information with real world scanned data. Potential design clashes can be then identified and rectified before a contractor arrives on site, avoiding costly hold-ups. After all, it is estimated that each clash can cost $17,000† to resolve once construction has begun, a cost that can quickly multiply into millions on big budget schemes.

† O'Donnell & Naccarato, 2012. Clash Detection in BIM Modelling. Association of Construction & Development.

Mark Senior is the Business Development Manager at Pointfuse ( based in Maidenhead, Berkshire

Download a PDF of this article