From theory to ontology: Representing people in building performance simulation models

Abstract

Building information modeling (BIM) can be viewed as an evolutionary process; whereby digital representations of buildings become increasingly detailed. Similarly, the main concern of building performance simulation (BPS) has been the virtual enactment of buildings‘ behavior. Naturally, the bulk of efforts to couple BIM with BPS have been circled around the methods by which geometric and semantic building properties captured in BIM are inherited by BPS. But these properties have been predominantly of a static nature. However, whereas classical representations of buildings may mostly involve static elements, buildings themselves function under highly transient boundary conditions, both external (i.e., microclimatic conditions) and internal (occupants’ patterns of presence and behavior). As a consequence, impulses to augment BIM models in view of dynamics have frequently stemmed from developments in BPS. One such impulse pertains to the critical challenge of representing people in digital models of the built environment. The early developments in BPS focused on the modeling methods for physical phenomena such as heat transfer and structural dynamics. In contrast, the development of high-resolution models of occupants has been a comparatively recent concern. In this context, the present contribution explores three layers of discourse necessary for a theory-driven occupant-centric ontology: i) foundational theories (sources of domain knowledge of occupants‘ perception of and behavior); ii) ontology (a shared schema to capture the classes and attributes necessary for an interoperable representation of occupants’); iii) application (the computational implementation of ontologically streamlined processes pertaining to occupants‘ perception of and behavior in buildings).