Identifying Intrusions in Dynamic Environments Using Semantic Trajectories and BIM for Worker Safety

Abstract

While there exist many systems in the literature for detecting unsafe behaviors of workers in buildings such as staying-in or stepping into unauthorized locations called intrusions using spatio-temporal data. None of the current approaches offer a mechanism for detecting intrusions from the perspective of a dynamic environment where the building locations evolve over time. A spatio-temporal data model that is required to store worker trajectories should have a capability to track a building evolution and seamlessly handles the enrichment of stored trajectories with the relevant geographical and application-specific information sources for studying the worker behaviors using a building or a construction site context. To address this requirement of maintaining the information, which is generated during the building evolution and for constructing semantically enriched worker trajectories using the stored building information. This work reports a system which offers the ability to perform user profiling for detecting intrusions in dynamic environments using semantic trajectories. Later, Building information modeling (BIM) approach is used for visualizing the intrusions from a standpoint of a building environment so that necessary actions can be performed proactively by the safety managers to avoid unsafe situations in buildings.