The Role of Artificial Intelligence in Sustainable Building Operations
Artificial intelligence (AI) holds a promise of improving the world—but it’s also capable of contributing significantly to climate change. As the buzziest recent tech innovation, AI has seemingly endless applications. For buildings, its use cases include automated controls for tenant comfort, energy management, building security and more, all of which can reduce the costs and carbon emissions associated with building operations and improve tenants’ experience at the same time.
Alongside AI’s benefits, however, building controllers also need to consider its potential impact to their organizational sustainability efforts. AI-enabled technologies require serious computing power for model training, processing and data storage. Gartner predicts that at current rates, AI could be responsible for as much as 3.5% of all global electricity consumption by 2030. That would be more than what the world’s entire human workforce currently uses.
This could become an issue for building owners in the near future as they try to bring down their electricity use to meet corporate sustainability goals and comply with possible regulations on energy use. This outside pressure on building owners is increasing as governments around the world begin to mandate scope 3 carbon emissions reporting for certain corporate organizations.
Building owners may already be familiar with scopes 1 and 2 reporting: that would be sharing the emissions produced directly by company-owned sources and those from purchased energy sources like electricity. Scope 3 reporting would cover emissions coming from a company’s value chain. This would include the emissions produced by vendors, including the technology used in a company’s buildings portfolio.
This shouldn’t scare building owners from adopting AI technologies: they’re an important tool that can help buildings operate at their most optimal levels. AI is a new industry and despite the many opinions going around about its potential, it won’t singlehandedly end climate change nor doom the world to the worst possible outcomes. The reality is, AI researchers and developers are increasing their understanding of its carbon footprint every day and innovating their way toward reduction strategies. With a full understanding of the state of sustainability in AI, building owners can trust its ability to support their own initiatives while preparing for future environmental regulations.
The changes needed to reduce AI’s carbon emissions are a facilities challenge in and of themselves. AI processors are housed in data centers, which have immense cooling needs in order to offset the heat generated by the thousands of servers they host indoors. As AI scales, so do the data centers that house them. Technologies like HVAC and water cooling, alongside tactical space configurations, help keep temperatures down—which, in turn, means a lot of power is used to mitigate the effects of immense energy consumption.
To combat this issue, data centers are typically located in cooler areas of the world. That means certain countries’ energy usage and sustainability efforts are disproportionately impacted by their presence and the growth of AI. In turn, these countries are the driving force behind regulations pushing data centers to reduce their environmental impact. Germany, for example, has the second-most data centers in the world and is passing regulations focused on reducing data centers’ harms on their environment. One proposed law, the Energy Efficiency Act, would require that data centers use 100% renewable energy by 2027 and jumpstart efforts to capture and reuse residual heat, which could be used to warm other structures in lieu of electricity.
The good news is that in response to AI’s sustainability gaps and upcoming regulations, there are a number of initiatives being made by AI players to drastically reduce these issues. Amazon, a major data center operator, is moving toward 100% renewable energy for power, extending the lifespans of their servers, and working to become water positive by 2030 so that more water is returned to communities than is used by their facilities for cooling. Amazon also has a goal to have net zero carbon emissions by 2040. Microsoft’s Azure data centers have a more ambitious goal to be carbon negative by 2030 with similar pathways to achieve it as Amazon. If more data centers follow these giants’ footsteps, then AI’s carbon footprint will decrease even more quickly and become less of a concern for building owners.
Building owners should take advantage of AI for building management while understanding its duality as a major sustainability tool that has its own carbon footprint to keep track of. Globally, building operations account for a quarter of all carbon emissions, and using AI to automatically manage lighting, HVAC and more are just a handful of ways that AI can address this issue. But as industries worldwide prepare to publicly report more extensive sustainability data, AI users should understand what their technology vendors are doing to reduce their carbon emissions.
With certain smart investments—for example, choosing a vendor based on their pathway toward emissions reductions, or looking into updated methods for calculating scopes 1, 2 and 3 emissions — building owners will find themselves ahead of the curve for both building management and sustainable practices.
