AI for building energy modeling: A transformation

With the increasing volume of data from buildings and affordable powerful computing, artificial intelligence (AI) has been explored in various applications for building energy modeling (BEM), including collecting input data, creating and tuning energy models, managing simulation runs, and extracting insights from large volume of simulation output to inform decision making across a building’s life cycle for energy efficiency, demand flexibility, climate resilience, and occupant comfort and health. However, significant challenges remain to address, including AI-ready data, selecting fit-for-purpose AI models or tools, BEM workforce training, standard benchmark datasets and methods. This perspective article describes how AI is transforming BEM workflows and the larger ecosystem focusing on four major AI themes of data, models, computing, and applications, highlighting the associated opportunities, challenges, and future trends.