That new capability, enabled by chatbots and agentic AI, was set out by Samsung chiefs during a panel discussion at the recent UK Construction Week London event (12-14 May). ‘Intelligent Buildings: Plan, Design and Develop for How We Live, Work and Play, Now and for the Future’ explored the opportunities presented by smart buildings, from before construction begins through to their eventual occupation.
Possibilities are increasing as AI technologies become more energy-efficient, said Timothy Hospedales, vice president and head of the Europe AI Centre at Samsung. This means that once-costly processes running on the cloud can now be done in real time on embedded devices.
Domestic use of AI will encourage construction firms to integrate smart tools earlier into a project, said Jamie Harrison, Samsung head of strategic partnerships. “I think we’re going to see it fall further back into the construction process, where you’re going to see environmental monitoring, energy monitoring actually happening on that building site from the moment it’s set up.”
For people at home or workers on construction sites, the main advantage of new AI tools is the opportunity to ask questions and get answers in “plain English”, Harrison said. “It’s that change that really is going to shape the future,” he said.
“We hear reports every day about ‘We’ve collected loads of data off this site about emissions, we’ve collected loads of data off this site about how efficient this building is performing’. But it’s about what you can actually do with that data…
“You can ask an AI and say: ‘Where might I generate revenue from this site? Where might I be able to optimise to increase my energy efficiency?’ And I think that kind of ‘democracy of AI’ is a real gamechanger.”
The technologies offer further capabilities during operation of a building, allowing owners and managers to optimise not just efficiency and reliability, but also the comfort and enjoyment of all occupants.
Harrison gave the example of a group of pupils visiting a building on a school trip; the operator could “have a conversation” with the building through AI, he said, asking it to adjust the temperature as needed in a particular room, and to schedule a check from maintenance staff later in the day.
“We’ve got the data, we’ve got a shared system,” Harrison said. “I’m really excited to see what happens when you can have a conversation with that system and actually treat the building more as your partner, and start to see what feedback you get – say: ‘Have you also thought about this?’ That’s where I think we’re heading, and that really isn’t too far away.”
Agentic AI allows the setting of nuanced goals rather than simplistic tasks, Hospedales added. “The goal might be to maximise user comfort while minimising energy cost, for example,” he said.
“Then you have a new part, about planning. So the agent says, ‘OK, what plan should I come up with that achieves this goal? Maybe I adjust the blinds, according to where the Sun is. Maybe I buy and sell electricity from the grid at the right time of the day, according to the status of my building’s batteries and the current power price. Maybe I dispatch the cleaning robot to the place where the schoolkids have just been.’ Whatever it is, it’s making up plans to figure out how to solve those goals it is given.
“And then the third piece is the memory those agents have. Maybe at the beginning, it doesn’t know what schoolkids visiting means, and then eventually it learns that schoolkids are people who make more mess than average users. And so then it needs to adapt to that accordingly, in the future plans it puts together.”
As record-breaking heatwaves become seemingly annual events and critical global systems face potential collapse, climate resilience is another significant focus for AI-enhanced buildings.
The topic has “overtaken” net zero conversations in infrastructure operation, said panel speaker Professor Sean Smith, chair of future construction at the School of Engineering, University of Edinburgh. “While [operators] had sensors and other things in their infrastructure and their own buildings before, they’re now stepping that up,” he said.
Drawing on his experience as a member of Climate Ready Infrastructure Scotland, which covers telecommunications, water, power, energy and canals, he said the cost of monitoring equipment was preferable to the cost of insurance, repairs, maintenance and downtime from climate-related disruption.
“Suddenly, the value of that data and [those] sensors is coming back,” he said. “10 years ago it was useful information – now it’s critical information.”
Integration of satellite technology, such as for wildfire monitoring, is further enhancing smart management of infrastructure, he said.
The future of AI in buildings will be “predictive resilience”, said speaker Matthew Dunne, senior consultant at global property and construction consultancy Rider Levett Bucknall, “actually looking ahead at what will come up, as opposed to just designing for now and optimising systems for how they perform in the present”.
He added: “So whether that relates to climate risk, whether that relates to carbon performance, future data requirements, even things like our supply chains, I think resilience is going to be the keyword with AI going forward… a lot of difficult-to-foresee elements can be managed upfront.”
Want the best engineering stories delivered straight to your inbox? The Professional Engineering newsletter gives you vital updates on the most cutting-edge engineering and exciting new job opportunities. To sign up, click here.
Content published by Professional Engineering does not necessarily represent the views of the Institution of Mechanical Engineers.