Visualizing intrusions in dynamic building environments for worker safety

Abstract A review of the existing literature suggests that intrusions on sites are often overlooked as current construction safety assessment methods are primarily focused on visible consequences such as injuries and deaths. Moreover, traditional Behavior-Based Safety (BBS) methods for improving worker attitude towards adopting safety practices on work sites are not adequate as these methods do not provide quick feedback for changing unsafe worker behaviors in real-time to avoid fatalities. Consequently, geo-localization systems are used for acquiring spatio-temporal trajectories to identify unsafe worker behaviors on construction sites. However, spatio-temporal trajectories lack contextual building or site information. For incorporating the geographical and application-specific context in trajectories, semantic enrichment processes are executed. The semantic enrichment process should include changing contextual information related to the evolving building objects over time. The changing contextual information of building objects is required to be tracked by a system to study the worker behaviors in a dynamic building environment context. After reviewing the existing literature, it is concluded that none of the present systems offer a mechanism to identify unsafe worker behaviors such as intrusions in dynamic environments where the contextual information of building spaces evolves over time in terms of location, size, properties and relationships with the environment. To fill this research gap, a system is proposed; First, to acquire the worker movements using Bluetooth Low Energy (BLE) beacons. Second, to perform the pre-processing as well as semantically enriching the spatio-temporal worker trajectories using relevant updated contextual information of a building. Lastly, to visualize the building locations where intrusions have occurred using Building Information Modeling (BIM) approach. The developed system presents an integration of systems from the data acquisition stage to visualizing the unsafe work behaviors that could serve as a foundation for future research in studying advanced movement-related worker behaviors in dynamic environments by overcoming the spatio-temporal data management challenges.