Data-driven approaches to improve Indoor Environmental Quality: A Systematic Literature Review.

Abstract

Indoor environmental quality (IEQ) refers to the conditions within a building and is a key factor in occupant comfort. To improve these IEQ, heating, ventilation, and air-conditioning (HVAC) systems are currently being used in indoor environments in buildings. However, inefficient systems to improve IEQ lead to high energy consumption, which is directly connected to building-energy-related CO₂ emissions. Unfortunately, these emissions are inextricably linked to global climatic issues. Data-driven technologies show a positive future in solving climate-related targets. Technologies such as artificial intelligence (AI), machine learning (ML), and digital twin (DT) are being explored to improve energy performance while improving IEQ. Therefore, this paper aims to systematically review the applications of data-driven technologies for improving IEQ and energy efficiency. A comprehensive literature search using the PRISMA method to retrieve publications related to data-driven and IEQ. The findings highlight common areas of data-driven applications in IEQ, including the detection and prediction of IEQ factors, IEQ controls, building energy management, and HVAC system controls. Commonly used ML and AI techniques identified, including deep learning, neural networks, KNN, SVM, decision trees, and multi-objective algorithms. AI in combination with BIM and IoT techniques was employed in the studies to develop DT models for real-time building monitoring. Further, the review highlights barriers to data-driven and digitalisation approaches in buildings that impede the robust approaches to IEQ improvement.