Trial PointFuse

Digital Twin Capabilities

Steve Salmon, CEO of UK-based PointFuse, says digital twin thinking will transform the architecture, engineering and construction world in the next five years.

What do you believe are the biggest trends in modeling and simulation tools, particularly as it relates to digital twins of physical things like cars and buildings and more ephemeral things like simulating business operations?

Some industries – for example, automotive and aerospace – have moved forward significantly with digital twin technologies, but the architecture, engineering and construction (AEC) market has historically lagged. However, we have seen digitalization of the AEC sector accelerate in recent years, buoyed by technology investors keen to capitalise upon the opportunities to transform a $12 trillion global industry.

Building owner-operators are also increasingly keen to build a digital platform for asset maintenance. A typical building may have a 50-year life-cycle, but we know maintenance costs can surpass the original design and construction costs in under 20 year, particularly if buildings and their systems have not been accurately documented, and if that documentation is not kept up-to-date.

Previously heavily dependent upon 2D drawings and paper-based records, AEC professionals are now increasingly capturing information digitally for handover to the building owner. This ‘digital twin’ of a building is being partly developed using building information modeling (BIM) processes and technologies, but these don’t capture detailed as-built information. Laser-scanning and photogrammetry techniques are therefore being deployed, along with geo-tagging solutions, to provide comprehensive, accurate, asset-specific 3D representations of buildings and their internal systems for use through buildings’ life-cycles.

We are some years away from routinely having live streams of real-time data connecting physical buildings to their digital counterparts, but digital reality capture is preparing the way. Incidents such as the Grenfell fire disaster in London and the Florida condominium collapse are underlining the need for building owner-operators to have a ‘golden thread’ of digital information detailing what was constructed when and by whom for health and safety reasons as well as for efficient building operation and maintenance.

As regulatory requirements grow, we are going to see dramatic growth in demand for reality capture of both new and existing buildings, and meeting this demand will be increasingly dependent upon using technologies that can be easily integrated with existing solutions and platforms.

What have been some of the more interesting algorithmic technical improvements in modeling and simulation to watch and why?
With PointFuse’s mesh approach, users can also dramatically reduce the heavy duty processing involved in converting ‘dumb’ point clouds into useable data both for millimetre-accurate progress monitoring and for as-built information handover. Using lightweight outputs that are in the same file formats as their original models means architects and engineers can now easily associate geotagged data with their 3D models in, say, Autodesk Revit, or with other design documents held in document management platforms such as Procore.

Modelling has also advanced through use of geotagging in the field; subject matter experts in surveying can rapidly create asset tags that can be used for ‘digital twin’ purposes. Meshing and asset identifiers provide a powerful base information layer for asset, facilities and space management, and for monitoring the environmental efficiency of buildings and their systems.

We are also seeing rapid technology advances that reduce the need to deploy highly skilled professionals in field data capture, processing and use. Holobuilder’s 360-degree image capture and simultaneous localization and mapping (SLAM) technologies such as NavVis can, for example, be combined with artificial intelligence (AI) and machine learning (ML) to automate asset identification. And PointFuse quickly converts point clouds to formats that can be readily accessed by designers using their normal tools.

How would you characterize the market and capabilities around simulation tools and engines for engineering, business simulation, and other common use cases related to digital twins?

Digital reality capture, analysis and development of digital twins is still at the ‘early adopter’ stage in the AEC sector. Industry methodologies have not kept pace with technological developments, and – as construction is a notoriously low-margin, cost-conscious industry – many businesses have been reluctant to invest in the hardware, software and skills needed to advance.

However, building owners and architects are waking up to the benefits of investing and working in 3D environments, and we are also seeing new business models that provide cost-effective routes to technology adoption. Hexagon’s Multivista business, for example, is inviting people to set up as franchises to expand the reach of its reality scanning and documentation tools and services.

What do you think are some of the more significant breakthroughs in workflows, particularly as you cross multiple disciplines like mechanical, electrical, thermal, and cost analysis for designing new products?

As I mentioned, the Grenfell fire and Florida building collapse have focused people’s attention on the through-life use and re-use of information. AEC workflows are having to extend beyond the previously separate silos of owners, architects, engineers, surveyors, contractors, specialists, manufacturers and suppliers, facilities managers, etc. Information management platforms such as Procore are now being used to span multi-disciplinary teams, and, along with technologies such as PointFuse, can help tackle cost and data management constraints. This has two effects: first, it cuts the cost and complexity of capturing data; second, it enables economic benefits to flow from smart use of that data.

For example, teams might be adopting off-site fabrication approaches to building delivery. Cost-effective installation of modules is a highly information-dependent activity, so accurately capturing and sharing as-built data throughout earlier construction activity can be critical in avoiding ‘clashes’ or other problems when modules are lifted into place and connected to other systems.

Similarly, when it comes to fitting-out a building, tenants can maximise their use of the lettable space through re-use of accurate as-built data about the existing building, its structure and services. Knowing what’s inside suspended ceiling spaces and behind wall panels avoids expensive additional surveys or investigations to identify if particular options are feasible and viable (a US estimate suggests building owners annually spend 20 cents per square foot just to locate things – in a one million sq ft portfolio, that’s $200,000 a year). Scanning dynamically during construction can therefore help avoid cannibalising lettable area later, and reduce uncertainty about asset locations.

What are your predictions for some of the more exciting developments in simulation and modeling tools over the next 3-5 years, particularly as it relates to digital twins?

Some of the biggest advances will be made from building on advances we are already making through better exploitation of information via shared platforms. As we expand our AI and ML tools, I expect we will see tighter integration between the captured data and the original design models.

The biggest customer demand is for a ‘magic button’ to convert point clouds into a Revit model for as-built data. I see PointFuse as part of the solution, and through applying cloud-based AI and ML, I believe we will soon be creating semantic mesh models where components recognise their environment and their context or relationships with other components.

We will have accurate as-built imagery that is directly related to the owner’s original asset information requirements captured in BIM, rendered in industry standard formats like IFC and FBX, and providing a clear route towards an accurate digital twin of the physical asset.

Above all, we have technologies which are now more affordable and usable than ever, which can be deployed at lower costs, and which deliver significant savings over the lifetime of buildings. With wider industry awareness and adoption, the AEC sector will be in a stronger place to create and manage digital twins for building owners.