The research aims at providing a state of the art regarding the use of Post-occupancy evaluations (POEs) to optimize the facility management phase of large building stocks. Building occupancy evaluations are common topics in the field of building energy performance studies, with recent attention on the impacts of user actions and movements on building performances. Most part of the research implied the use of sensors to monitor these aspects, providing huge amounts of data regarding large building stocks. The research provides a literature review about: methods, tools, and existing studies regarding the use of sensors to monitor occupancy values at room level and users’ flows; applications of machine learning (ML) techniques to analyze large amounts of sensor data and provide valuable predictive information on facility management; existing types of ML techniques and related feasibility for the presented purposes. In addition, the paper investigates the integration in a BIM approach to visualize occupancy levels and predictive information in the Information Model. Potential applications to facilitate the optimization of cleaning activities and reorganization of spaces in large building stocks are also explored, investigating the setting of a decision support system for facility managers to handle building management and cleaning activities using predictive information.
Full Text (PDF)