This year’s TechFest conference focused on how AI and automation can help the construction industry reduce its carbon footprint
According to the government and environmental organisations such as the UK Green Building Council, the construction sector accounts for about 25 per cent of the country’s carbon emissions. But the hope is that automation and AI-enabled data will help push that figure down.
At TechFest 2024, speakers explored ways AI could help construction cut its carbon footprint – including by reducing embodied carbon on new-builds and enabling retrofit.
The conference, a joint initiative by Construction News and its sister title New Civil Engineer, took place on 27 November at the Hilton London Metropole. It attracted 15 exhibitors and 250 delegates from across the construction and engineering sectors.
In his keynote address, Arup global digital services leader Will Cavendish said adopting AI systems will be crucial in the push to move the sector’s business model towards retrofit and away from new-build construction.
“We need to evolve our business model from one which is about creating new stuff, to one which is more about repurposing and replenishing the assets that we’ve already got,” he said. Cavendish added that the industry faced the additional challenge of using as little carbon as possible.
He noted that “we’ve already nearly exhausted” the carbon emissions budget if we are to stay within a forecast 1.5-degree Celsius rise in global temperatures by 2030, meaning that innovative technology-driven solutions are needed.
“A range of AI systems are already proving highly impactful in their ability to augment human intelligence in meeting these fundamental challenges,” Cavendish added. In his speech, he gave examples of where Arup has used AI for planning and retrofit purposes. “We’re finding AI significantly augments our ability to design greener and lower-carbon cities and city infrastructure,” he said.
“That allows our water engineers, our city designers and our city planners to create different kinds of solutions for city infrastructure challenges, particularly ones that integrate nature and the environment into infrastructure design.”
For instance, in Shanghai, Arup designed a water infrastructure site that “integrated nature and nature-based solutions into it”. He said this decreased the cost and carbon impact of the new infrastructure by 30 per cent each.
Optimising data
The TechFest conference also touched on using automation and machine learning to monitor the carbon footprint of concrete and other materials.
Cavendish said that energy-intensive cement, steel and glass production account for about 20 per cent of global greenhouse gas emissions and are difficult sectors to decarbonise.
But attendees heard about ongoing efforts to tackle the issue of decarbonising construction materials.
For instance, Morrisroe lead engineer James Wibberley detailed how his firm has used data techniques to analyse ready-mix concrete.
“Historically, we want to know compressive strength, slump [and] minimum cement content. But now we’re being asked for the whole recipe of that concrete mix,” he said. “What is the cement replacement blend percentage, for instance.”
Morrisroe has also been working with data to link the cement’s makeup to its performance and its embodied carbon. This has allowed it to improve its mix to minimise its environmental footprint.
“We’re having conversations with architects and clients about how we can improve the performance of a mix, or how we can look at meeting or even beating the benchmark of the embodied carbon targets,” Wibberley said. He added that optimising data is a particular buzzword.
Oceans of data
In the same session, Morrisroe director Dan Bannister discussed how clients have become very interested in how buildings perform environmentally – which can be demonstrated with data. “There are oceans of data out there,” he said.
Clients like Goldman Sachs, he said, are now asking for data on how buildings can be made sustainable.
“They’re really interested in how the building has been performing, in addition to what their space is going to be like. We found ourselves in the past 12 months providing a lot of guidance and information feedback to the developers,” he said.
Cavendish said that automation is being used to deal with the “sheer appalling state of data” related to existing buildings. When it comes to retrofit, it is often too time consuming and difficult to manually analyse the state of the building because there is so much documentation, he explained.
“For most of our existing buildings, the data doesn’t talk to each other, it’s held in completely different coding formats,” he added.
On a project in Australia, for instance, a client had an “absolute Tower of Babel” of information on one site, which was impossible for humans to read and analyse. But by using automation, they were able to see much more clearly what the building’s component parts were made of and produce more accessible data.
“We were able to design automation scripts that automated a number of different asset classifications for the sorting of consistent asset tagging and asset management information,” Cavendish said.
