Abstract Buildings are complex dynamic systems composed of sub-systems and components in continuous interaction with human behavior. Occupancy and movement data are crucial to deepen the understanding of the built environment performance. Observation data is paramount in space–use studies. However, there is a lack of automatic observation techniques, which enable the continuous and systematic recording of the mobility behavior, the producing of non-arbitrary registries, the gathering of big data and the emergence of patterns. This study is divided into two companion, although autonomous, papers. The first paper [M. Kuipers, A. Tome, T. Pinheiro, M. Nunes, T. Heitor, Building Space–Use Analysis System — a Multi Location/Multi Sensor Platform, Automation in Construction 47 (2014) 10–23] describe the video-based analysis system and give account of preliminary use patterns obtained by data fusion of video plus RFID inputs; the present paper expand the obtained results and focus on investigating analytical procedures aimed at the study of the functional condition of architectural artifacts, and promotion of a better understanding with the spatial conditions, based on computer vision based tracking. Computer vision allows the simultaneous recording of the user and the spatial container and a full description of the movement behavior. The proposed method for the analysis of the space–use interactions is evaluated in two main atriums of a university department building. It allows to: a) represent, describe and quantify occupancy/co-presence patterns and movement/navigation patterns; and b) establish correlations between the occupancy/movement patterns and the morphological properties of space. Several mobility indexes and its mapping are obtained, such as the number of users, time occupancy, average speed, and users' encounters. The results show how space–use data can be interfaced with spatial analysis tools to arrive at an understanding of the relationship between space–use and building design.
Read full abstract