D

igital twins have the potential to change how buildings are designed, functioning as circular links between the physical and digital worlds. Sensors within a building take regular measurements that are fed into the digital environment to generate a virtual replica. This digital twin is regularly updated and can generate alerts when issues are detected.

The technology could also be vital in maintaining digital records of buildings, as well as assessing the structural integrity of materials over time. Similar technologies are already used in industries such as automotive and oil & gas, to monitor assets and predict maintenance.

From BIM to digital twin

While digital twins are still some way off from becoming a regular feature in construction, building information modelling (BIM) is already widely used by architects and developers. BIM contains detailed construction data, but it is not regularly updated to reflect to real-world events. This could change when BIM is used as part of a digital twin, which receives regular feedback from the physical building to update the virtual model accordingly.

“A big part, I think, will be actually taking feedback from the operational parts of asset and helping that inform new designs,” explains Simon Evans, digital energy leader at Arup, who works closely with the UK’s National Digital Twin programme led by the Centre for Digital Built Britain.

“Then, later on down the lifecycle in construction and operations, something might happen that maybe wasn't predicted in the design phase. If we can close that loop and say: ‘Can we inform our future architectural and buildings based on knowledge we have from the past or existing operating assets?’ That'd be fantastic. But also continue to optimise and improve the operational asset based on insights derived from the digital twin.”

There is a misconception that a digital twin is a specific type of technology you can buy off-the-shelf, but rather, it is a combination of technologies, Evans explains.

“While many companies offer specific software for use in the digital twin space, the reality is that the concept is more of a methodology and an approach, rather than a product or technology, which is underpinned by full-lifecycle information and data management,” he adds.

Collecting the right data at the right time

Scott Brownrigg Architects has a dedicated arm of its businesses called the ‘digital twin unit’. The company has collaborated with Vietnam-based Atlas on several international infrastructure projects involving digital twins, such as Hong Kong International Airport. The companies are also working on a student housing project in the US.

“A well-constructed BIM model is a foundation for a good digital twin,” says Ana Matic, director of digital development at Scott Brownrigg Architects. “If you already have a well-structured BIM model for your building, you will then be able to then connect the digital twin systems to it without much trouble, but also you can even brief your teams to deliver a BIM model that predicts some of the systems that will be integrated.”

There is a perception that a digital twin requires multiple sensors and devices to collect data continuously in real time. And while this is technically possible, it can be counterproductive. There is a risk of collecting too much data that is not acted on correctly, which could also decrease network speeds and clog up storage.

“It’s more appropriate to refer to it as right time, rather than real time,” adds Evans. “There are not many applications that require truly real-time data. Instead, it’s about collecting at the right frequency based on the problem you are trying to answer.”

Before installing sensors and software for a digital twin, it is vital that organisations first decide what specific data they need and what they need it for.

“The data has to be structured,” Matic explains. “We can all collect data and have a lot of it, but it's whether we're being able to use it appropriately, and also, are we getting the right sort of outputs from it? Because you need to understand what the data is telling you and whether you're actually able to push it in the right direction for it to help you make decisions. We need to be able to set up systems that are crunching a lot of data, but also getting quite practical and pragmatic results out of that.”

Artificial intelligence and machine learning are technologies can help make sense of the data collected. Another helpful technology is edge computing, which will allow users to determine exactly what they need to know in advance and only collect that precise data. The roll-out of 5G will dramatically increase data gathering speeds.

“If it's more targeted, and edge computing is leveraged correctly to filter, to sample, to digest and then provide just what is relevant, then it's going to be a natural part of it,” adds Evans. “This is why I think that there's no one technology that will solve all this. It's always about a constellation of technologies talking with each other.”

Scott Brownrigg Architects has collaborated with Atlas on several international infrastructure projects involving digital twins, such as Hong Kong International Airport. Credit: Scott Brownrigg Architects | Design Strategy Unit

Will large projects be the driver of change?

It is not uncommon for clients to request architects change designs to reduce costs. However, there is a belief that digital twins will allow architects to better persuade clients of the merits of their designs, as they have the data to back up their arguments and can display the benefits. This will be particularly important for maintainable assets.

“Convincing your client to, for example, get a much more environmentally sustainable ventilation or heating system will definitely be part of that because you will be literally able to see that through the development of the digital twin handover,” says Matic. “And you'll be able to understand how much money you might be able to save.”

The construction industry has a reputation for being slow to embrace change and preferring to stick with established practices. Yet there are signs that this is starting to shift.

"A lot of contractors are now getting very organised and very good at asking for things and being able to brief us in terms of BIM and digital deliveries,” explains Matic. “Also, they're getting better at monitoring their own digital delivery – scanning on-site and all that kind of stuff.

“But the push to deliver already structured digital twins, I think, will come from the industry, almost from the facility management industry backwards into construction. Because more and more larger clients will be able to see benefits from that. The challenge is really how we do that for smaller clients and how we do that for existing stock and smaller projects.”

Digital twins are currently used for infrastructure and public projects as large organisations are already aware of the benefits. These organisations also have the funding and resources, with a digital twin being a logical extension of their existing IoT network.

The digital twin concept will likely be understood and adopted first as part of these larger infrastructure projects and then gradually spread to smaller companies in the construction industry.

“Part of the industry-wide and international move is to standardise a lot of the approach that we're all using so that it becomes almost like an international language,” says Matic. “And we can actually tell each other why we're doing certain things in BIM and then how that helps the later maintenance of the buildings and the digital twin.”