Uncertainty-aware visualization and proximity monitoring in urban excavation: a geospatial augmented reality approach

Background: This research aims to improve the urban excavation safety by creating an uncertainty-aware, geospatial augmented reality (AR) to visualize and monitor the proximity between invisible utilities and digging implements. Excavation is the single largest cause of utility strikes. Utility strikes could be prevented if the excavator operator were able to “see” buried utilities and excavator movement, and judge the proximity between them in real time. Geospatial augmented reality (AR) is an enabling technology for such knowledge-based excavation. It synergizes the geospatial utility locations and the excavator movement into a real-time, three-dimensional (3D) spatial context accessible to excavator operators. The key to its success is the quality of the utility location data. Methods: This paper describes a dynamic approach to incorporate positional uncertainties of buried utilities into an uncertainty-aware, geospatial-AR system for real time visualization and proximity analysis. Uncertainties are modeled as probability bands (e.g. spatial bands with certain probabilities of enclosing the “true” location of utilities). Positional uncertainties are derived in real time by referring to its determinant, data lineage, the genesis and processes used to collect and interpret data. Results: A computational framework, and a generic data model and its XML-format implementation are developed and tested. A method is developed to analyze the proximity in the context of positional uncertainties of both the utilities and the excavator movement. Conclusions: This newly created approach is expected to contribute to the safety in urban excavation via the integration of Geoinformatics and construction informatics into an uncertainty-aware, geospatial-AR, with real time visualization and analytical capabilities.