AI-powered modelling is changing construction – but who is vetting the new tools?

Of the many new technologies the construction industry has adopted over the past two decades, virtual construction – including building information modelling (BIM) – is among the most transformative.

 

3D modelling tools such as AutoCAD and Revit have become integral to the architecture, engineering and construction industries over the past 30 years. All building information – material, thickness, shape, size – is stored in a single model, ensuring accuracy and precision across stakeholders. Clients have also benefitted from being able to “see” a building before work begins on it.

 

Virtual construction (VC) builds on advances in traditional 3D modelling by changing how that technology is used in the field once construction begins. Designs created using VC reveal critical details that could previously only be seen once the build began: tolerances, the assembly of components and the order they’re installed in; whether spatial conflicts will arise when, for instance, ductwork is installed close to electrical conduits. It offers installation-ready geometry, weeding out errors, enhancing safety and helping clients understand their space in close detail before work begins.

 

Catching errors early on

 

“If a picture is worth a thousand words, then a 3D model done using VC is worth a million,” says Joe Zobel, Head of MEP & Virtual Construction at Clune Construction. “Every stakeholder can understand the spaces by simply viewing or flying through the model. Any changes can be made with little impact on cost and schedule.”

 

A typical construction project loses between 10 and 15 per cent of its total budget to change orders. For a large, complex build, this can reach around 25 per cent. All told, millions are spent on alterations resulting from what are now, with the advent of VC, seen as avoidable errors.

 

What’s more, with accidents during a build often resulting from insufficient planning, the value of highly detailed modelling increases.

 

“With the upfront planning that VC provides, safety can be done in parallel,” says Zobel. “Visualising the project in 3D allows for efficient installation techniques to be developed with safety in mind rather than deciding what to do during the installation.”

 

An industry leader in tech vetting

 

AI is now enhancing VC primarily by automating model analysis and identifying the things a human might overlook – granular-level inaccuracies in design, for instance – which means clashes and rework are now caught before they happen.

 

Yet, Zobel says, the market has been flooded with AI applications “that may or may not provide something that’s beneficial”.

 

As a result, Clune Construction has dedicated a career path to not only using new technology but vetting it. Any technology deployed to the field that hasn’t been properly road-tested risks compounding the problems it was designed to address. Clune has therefore become something of an industry gatekeeper for VC and AI tools, which look set to transform much of the design-bid-build process.

 

“In the future, AI may be capable of generating complete facility designs, optimising systems and even procuring the required materials and equipment with minimal human intervention,” says Zobel.

 

If that’s the case – and current trends suggest it may be – then proper testing of the tech, as Clune does, will be one of the most valuable new tasks in construction.

 

As AI tools become increasingly common in the construction industry, recognising their limitations is critical, as they are only as effective as the data and information they can access.

---

For more information, visit clunegc.com