AI-Driven Building Automation Systems for Energy Optimization and Predictive Maintenance: A Quantitative Case Study of Smart Building Performance Enhancement

Abstract

This study examined the effectiveness of artificial intelligence (AI)-enabled building automation systems in improving energy performance and maintenance reliability through a quantitative case-study design integrating Internet of Things (IoT) sensing, machine learning inference, intelligent heating, ventilation, and air-conditioning (HVAC) optimization, and predictive maintenance analytics. The research analyzed twelve months of pre-implementation baseline data and twelve months of post-implementation operational data drawn from a 185,000-square-foot commercial office building accommodating approximately 950 occupants and supported by multiple HVAC assets. The findings revealed that AI-enabled controls substantially reduced building energy consumption, with monthly reductions ranging from 36.2% in the heating-dominant winter to 51.7% during the peak cooling month, and an overall annual energy reduction of 45.3%. Energy savings were strongly associated with cooling-degree intensity, indicating that machine-learning-based setpoint optimization, occupancy-aware scheduling, and weather-responsive control delivered their largest gains under high thermal load. Predictive maintenance analytics reduced maintenance expenditures by 34.8%, lowered emergency work orders from 62 to 24, decreased equipment failures from 31 to 11, and reduced unplanned downtime from 196 to 78 hours, demonstrating a shift from reactive to condition-based intervention. Regression analysis indicated that cooling-degree days, occupancy variability, and predictive-model lead time jointly explained a substantial proportion of the variance in monthly energy reduction, while correlation analysis confirmed strong relationships between anomaly-detection accuracy and downtime reduction. Overall, the results demonstrated that integrating AI with IoT-enabled building infrastructure produced measurable improvements in both energy efficiency and operational reliability, offering empirical support for the broader adoption of intelligent automation across commercial, institutional, and public building portfolios.