How robots and AI are building the future of construction

National Observer

20-05-2025

The future residents of 230 Royal York Drive will be able to say their home was built by robots.
The nine-storey housing development in Toronto's west end is being built by Intelligent City: a BC-based construction design and technology company specializing in prefabricated mass timber buildings. Intelligent City uses AI software to design and optimize its projects, and robotic arms to actually put the pieces together.
'The more standardized things are, the easier they are to mass produce — and the more custom things are, the slower you are in production,' Intelligent City president Oliver David Krieg told Canada's National Observer.
'You have to find a middle ground, and that's where automation comes in; that's where robots come in,' he added.
Intelligent City is one of a growing number of Canadian companies harnessing the power of artificial intelligence to make the construction industry more efficient and sustainable — at a time when the sector needs to build 3.5 million new units by 2030 to ease the national housing crisis.
Intelligent City's design and manufacturing facility in Delta, BC, is home to a series of AI-powered machines programmed for various tasks, such as applying glue, screws, and nails.
'You can program the robots to basically repeat the same logic, but always with a slightly different variation to it,' Krieg said, explaining that machines programmed to apply glue to specific timber shapes can be reprogrammed for different configurations.
'Robots don't have to do the same thing over and over again. They can change what they do, and they don't care.'
The components are then shipped to the construction site to be assembled (by humans, at least for now) into a building's structure.
'We wanted to make it easier and faster and a little bit more standardized for how we design and deliver housing, especially in urban centres, where we saw the most need and the biggest opportunity,' Krieg said.
'A lot faster'
Aside from the quirky bragging rights of living in a robot-built apartment, the residents at 230 Royal York Drive — which is slated for completion by the end of the year — will also be able to boast that their home was made sustainably.
Using mass timber instead of steel or concrete can reportedly reduce greenhouse gas emissions by 13 to 26 per cent.
And because it's made from trees, the theory goes, it stores carbon that would otherwise be released into the atmosphere. Though it should be noted that the jury's still out on whether or not that last point checks out in practice.
Still, mass timber is naturally fire safe: instead of going up in flames, it chars like a thick log, helping the building retain its structural integrity.
And it's sturdy in an earthquake, which might explain why the BC government has extended the cap on the earthquake-prone province's mass timber buildings from 12 storeys to 18, opening the door for companies like Intelligent City to scale up.
Prefabricated homes cut down on building time by an estimated 30 to 50 per cent, since so many key components can be created offsite in a controlled environment that's not at the mercy of the weather. It's a key plank in Prime Minister Mark Carney's plan to double housing construction to 500,000 units a year.
'If you want to accelerate construction, if you want to make construction more productive, more efficient, and ultimately, higher quality, you take it into a factory,' Krieg said. 'You reduce onsite work and you increase the offsite work, and that makes it go a lot faster.'
Data tracking
'AI has seen huge progress; capabilities are going up exponentially, and every day we're seeing new use cases,' said Wyatt Tessari L'Allié, spokesperson for Canada's Coalition for Responsible AI.
One advantage is its ability to integrate data from sources like satellites, sensors or drones, he said.
'AI can be used to go through a lot of information and catch issues. And then once products are finished, a lot more efficiencies can be found in terms of energy use.'
On the latter point, Montreal's BrainBox AI uses machine learning to monitor and coordinate building heating and cooling systems.
Inspired by autonomous technology found in self-driving cars, BrainBox AI's exclusive AI combines a building's HVAC (heating, ventilation and air conditioning) data with external factors like weather, utility rate structures, and occupancy to learn its individual patterns and needs.
From there it goes into autonomous mode, making decisions — like turning on a fan, a pump, or a heater — in real time.
'That may sound niche, but in terms of energy consumption — especially on a global basis — it's one of the biggest categories out there,' BrainBox AI CEO Sam Ramadori said.
'Therefore it's one of the biggest emitters out there. So our focus is optimizing the systems while they're running.'
While it varies per building, Ramadori said integrating BrainBox AI's tech usually sees an HVAC energy reduction between 13 and 25 per cent.
What sweetens the deal, he added, is that there's no big investment required, either.
Instead of the hefty expenses associated with constant system changes or reprogramming — or hiring a team of people to come and evaluate a building's efficiency, which can take weeks — BrainBox AI taps into a building's existing HVAC control system and instantly starts to feed its data to servers.
In newer builds, that connectivity happens with no physical installation at all: a software driver is simply downloaded into the building's system. Where that's not possible, an edge device around the size of a laptop is installed to connect the control system to the servers.
It's all about scalability. After all, air conditioning alone accounts for nearly four per cent of the world's greenhouse gas emissions.
'A big mission of BrainBox is to solve the emissions that buildings generate,' Ramadori said. 'We wanted to scale fast, and scaling fast means plugging into buildings that are already there.'
The company's technology is used in over 15,000 buildings globally, primarily in the US, but also in Canada, Australia, Europe and the Middle East. Its success led to its acquisition by Trane Technologies, a Dublin-based sustainable tech firm, in January.
BrainBox will remain headquartered in Montreal, using the city as a base for reaching into new markets.
'Montreal and Toronto are two major AI hubs globally,' Ramadori said.
'There is a lot of interest in the team and how it continues to innovate. So they [Trane] are definitely interested in continuing to invest in the team based in Montreal.'
Waste less water
BrainBox AI isn't the only Canadian company focused on being easily grafted on to new or existing buildings.
Vancouver-based Orca Water produces ultrasonic water meters and AI-driven sensors to track and understand a building's water usage.
'We started with the idea that water should be able to be measured anywhere it is used,' CEO Kerry Chin said.
'We know that's not the case right now. And we wanted to do this in a non-intrusive way, meaning you don't need to cut the pipe to put a measurement device inside.'
Orca Water's sensors are attached to the outside of a pipe, and use high-frequency sound to measure water flow.
From there, machine learning is used to analyze and understand water usage, identifying trends (such as where the majority of a building's consumption is coming from, be it the shower or the garden) and problems (which can be crucial in helping spot leaks before they turn into floods).
This is especially useful when applied to a multi-unit building, allowing individual tenants to better understand their own water waste.
'Typically, buildings are not piped to support individual metering — they are just piped to support whatever is convenient for construction,' Chin said.
'You can't tell which user or which tenant or which business is using what, because everything is connected. That affects conservation, because if you can't measure an individual unit's usage, you can't give people data to help them do better.'
Like BrainBox AI, Orca Water is primarily targeting existing multi-tenant buildings, because that's where the team feels they can affect the quickest change. Many of the company's customers are in Canada, though it does have an eco hotel in Austria that's currently using its technology.
AI challenges
The integration of AI does not come without concerns.
The rare elements used to create AI-powering microchips are mined in ways that can take an ugly toll on the environment in the Global South. Even just training an AI model can use thousands of megawatt hours of electricity and generate hundreds of tons of carbon dioxide.
Still, Tessari L'Allié said the construction sector doesn't have to worry just yet.
'In terms of the use cases for construction, I would be surprised if they're past energy intensive,' he said. 'That could change. But in the context of the construction industry, I think it's mostly beneficial [to use AI].'
Another big fear is that machine learning will replace jobs, rendering humans obsolete. And in some industries, that threat is very real (just look at all of the consumer brands using ChatGPT as their copywriters).
But when it comes to the construction sector, the issue isn't a lack of jobs — it's a lack of labour.
An RBC report predicted more than 500,000 new construction workers are needed to build millions of new homes by 2030. At the same time, the construction industry could see up to 250,000 workers — particularly labourers and carpenters — retire.
That's where AI is not just helpful, but necessary. Javier Glatt, co-founder and CEO at CadMakers, calls himself a 'techno optimist' on that front. CadMakers is a construction technology company that creates digital twins of future building projects, helping designers identify problems and inefficiencies before construction actually starts.
The software doesn't utilize AI, but Glatt still believes in its value.
'In the construction industry, generally speaking, over the last few years, we haven't had enough people to do the jobs,' he said.
'We can't get enough people to do all the things we need to do.'
Given that, AI can be a powerful tool not just for sustainability, but for economic growth – especially in a housing crisis.
'AI saves people time; it supercharges productivity,' Glatt said. 'So I think it's going to be net positive.'