Optimising Energy Performance of buildings through Digital Twins and Machine Learning: Lessons learnt and future directions

Abstract

In recent years, the energy efficiency of buildings has received increasing attention due to climate change mitigation goals, and higher energy costs. This paper explores the integration of 3D models, IoT sensors, Digital Twins (DT), data-driven modeling, and Artificial Intelligence (AI), particularly Machine Learning (ML) algorithms, to enhance energy performance prediction and optimisation in existing buildings. By leveraging real-time data from IoT sensors, DTs provide a comprehensive digital representation of buildings, facilitating intelligent monitoring and control for enhanced energy efficiency and occupant comfort. This paper presents the development and application of a data-driven DT for an office building in Norway, focusing on energy performance prediction. Through a case study, specific outcomes and insights are gathered regarding the feasibility and benefits of this approach, together with its inherent limitations. The results highlight that significant advancements in energy efficiency could be achieved through predictive modeling and intelligent control strategies. In future, adaptation of these technologies requires addressing key challenges and advancing methodologies for broader implementation. By identifying and addressing these challenges, the integration of IoT sensors, DTs, and AI holds considerable scope for optimising building energy performance and advancing sustainability objectives.