How AI can help cut real estate carbon emissions
JLL Chief Technology Officer, Yao Morin, explains how nascent AI technologies could enhance the scale and speed at which we can retrofit older buildings
The adoption of technology to help solve climate issues is now a given. The majority (85%) of commercial real estate leaders expect that their business will increase their real estate technology budget over the next three years, according to JLL's recent survey. Investments in artificial intelligence (AI) and sustainability solutions are expected to have the biggest impact.
But, while AI’s potential to transform the global economy has captured the world’s imagination, less talked about is how the adoption of AI will help tackle carbon emissions in real estate, which is responsible for roughly 60% of global emissions in major cities. The new technologies can enable public and private organizations to create effective sustainability strategies for single buildings and entire portfolios.
A major area where we’re going to see this unfold, and one that is underestimated, will come amid efforts to retrofit buildings. This is no small task. Globally, over 1 billion square meters of office space require retrofitting by the middle of this century. That means retrofitting rates must rise significantly from around 1% of stock per year today, to at least 3% of stock per year to meet the target.
Achieving this will take more than just installing new heating and air conditioning control systems or putting solar panels on rooftops. The solutions needed are far more complex. They involve systematically examining the current state of buildings, their structures, materials, energy and utility systems and operations. Essentially, it’s about absorbing and analyzing vast amounts of data at scale and speed, making AI perfect for the job.
AI at work
To effectively retrofit buildings, large amounts of information must be scrutinized. And, as no two buildings are the same, each property needs its own approach. This is where the speed of AI and large language models (LLMs) comes into play. For instance, going through construction documents and maintenance records can help limit the need for new equipment and materials, thereby reducing waste. And, examining engineering systems can help identify easy solutions, such as whether a simple sensor add-on could solve the problem, as opposed to a bigger overhaul.
This work, when done manually, is difficult and time-consuming. But AI can do it faster and more efficiently by digesting the documents, extracting the necessary information and quickly organizing it into a standard format that’s ready to use.
Take, for instance, plans to retrofit the World Trade Centre in Brussels, which, when completed, is set to be the most energy-efficient building in the European Union’s capital. During its construction, almost 95% of all the materials and existing equipment will be recovered, reused or recycled, with 65% of the existing buildings maintained. About 30,000 tonnes of concrete from selective demolition will be reused on-site. Another 1,000 tonnes of products and materials, including wood paneling, carpets and insulation, will be reused on-site or in other locations.
A project of this scale would benefit hugely from using AI to help streamline efforts allowing valuable human resources to be deployed elsewhere.
Monitoring in real-time
Project planning and monitoring construction sites in real-time are also areas where AI can add huge value. AI can use photos to generate 3D models of sites and layer it into a construction plan, helping developers build faster and with less risk by tracking project progress and optimizing construction schedules. Combining existing site data and images with generative AI capabilities can also be particularly helpful for visualizing how new structures will fit into existing surroundings.
Then, there’s supply chain tracking. Real estate must take into account supply chain carbon emissions, which are notoriously hard to track. This is where AI can help assess the pros and cons of different retrofitting options, rapidly analyzing the entire supply chain impact and total carbon involved, while standardizing reporting and documentation.
To be sure, there are still major challenges ahead. Retrofitting processes are complicated and AI can’t solve it overnight. In many cases, we’re also talking about old buildings, where less data may be readily available to feed into AI models, thereby requiring more human investigation and effort before AI can be of real use.
Yet, while AI doesn’t necessarily simplify the entire process, it can certainly help complete each task more efficiently, delivering better outcomes in the process. We have a long way to go if we’re to reach net zero carbon in real estate, but with more tools emerging all the time, we now have solutions available to get us started on the path.