MACHINE LEARNING PRACTICES DURING THE OPERATIONAL PHASE OF BUILDINGS: A CRITICAL REVIEW

Abstract

Machine Learning (ML) is gaining attention in civil engineering especially within operational phase of building life cycle. This phase is crucial for managing every energy aspect while ensuring occupant comfort. Previous ML experiments have explored occupant behavior, occupancy estimation, load prediction, defect detection, and Heating, Ventilation, and Air Conditioning (HVAC) system diagnostics. However, challenges such as ML transferability and limited literature on ML components for the operational phase hinder broader industry adoption. This critical review aims to assess the potential of ML in building operations, focusing on energy consumption, big data control, reinforcement learning, and thermal comfort modeling. By identifying knowledge gaps, the study recommends further research to leverage ML for sustainable energy consumption and occupant comfort. It highlights ML’s promising role in striking a balance between energy efficiency and occupant wellbeing.

Keywords

• big data control
• energy consumption
• operational phase
• reinforcement learning
• thermal comfort modeling