Virtual sensing in intelligent buildings and digitalization

Virtual sensing technologies have a huge potential in an informative and reliable sensing environment, which is essential to provide and maintain intelligent services in building life-cycle and digitalization. However, there remains a lack of comprehensive literature reviews and suggestions regarding virtual sensing applications in the building sector. The existing virtual sensing classification affords limited insight into virtual sensor modeling, verification, and calibration in building operations. Therefore, this paper reviews existing virtual sensors and applications and characterizes up-to-date virtual sensing technologies, with a focus on virtual sensors and virtual sensor calibration. First, virtual sensors are classified into built-in and in-situ ones according to their modeling and verification environment. The review also examines how virtual sensors have been used for observation, fault detection and diagnosis, and control in buildings. Second, virtual sensor calibration methods are mainly classified into virtual built-in calibration and virtual in-situ calibration. Finally, key calibration strategies proposed in previous studies are summarized. Through this comprehensive review and a consideration of the field-centric building characteristics and building life-cycle, this review proposes virtual sensor categorizations and highlights research challenges and directions in virtual-sensing-driven intelligent building systems to provide greater insight into the fundamentals and methodologies of holistic virtual sensing. Thus, the results of this review are expected to contribute to enhancing the availability and reliability of virtual sensors in the long-term in order to harness them as a novel sensing system in the context of the building life-cycle and digitalization.