Abstract #

The increasing demand for sustainable building operations has increased the integration of advanced IoT technologies in modern office environments. This paper presents a comprehensive study on the transition from raw data to actionable insights in understanding energy consumption patterns within a smart building located in Berkeley, California. The dataset, used over three years from more than 300 smart sensors, encompasses whole-building and end-use energy consumption, HVAC system operating conditions, indoor and outdoor environmental parameters, and meteorological data. This study focuses on developing interpretable predictive models to decode intricate energy consumption patterns. By prioritizing explainability, our research aims to bridge the gap between raw sensor data and meaningful insights. We investigate the significance of different patterns to enhance predictive accuracy while uncovering latent insights within the raw data. Through a combination of data-driven methodologies and Explainable Artificial Intelligence (XAI) techniques, we identify key factors influencing energy usage, thereby providing actionable intelligence for optimizing building operations. Our findings contribute to the broader discourse on sustainable building practices, aiming to reduce energy use, lower costs, and minimize carbon emissions.

Citation #