Artificial intelligence could help reduce energy consumption by as much as 40% by 2050

Artificial intelligence could help U.S. buildings to significantly reduce energy consumption and carbon emissions, according to a paper by researchers at the Lawrence Berkeley National Laboratory.

AI-driven energy use and carbon emission reductions could grow to 40% and 90% in 2050, respectively, compared with a business-as-usual scenario, the paper says. Absent energy efficiency improvements or policy support, building energy consumption would keep increasing. AI could be applied in four key categories: equipment, occupancy influence, control and operation, and design and construction.

The research focused on medium-sized office buildings in the U.S. It forecasts AI’s potential to reduce energy consumption and carbon emissions through six scenarios. These include a setting that uses current building efficiency as a baseline, as well as two business-as-usual scenarios with and without AI, and three policy-driven scenarios.

In a separate study, the International WELL Building Institute examined AI’s potential impact on building energy consumption earlier this year. IWBI said that by implementing digital twins and AI, operators could use live data feeds to automatically open or close doors and windows, or increase or decrease the strength of fans throughout a building to achieve efficiency gains.