Automating closed-loop structural safety management for bridge construction through multisource data integration.

• An integrated framework automates closed-loop management of structural safety during bridge construction is proposed. • Goes beyond mere structural member generation of previous time-dependent structural model generation method. • Enables dynamic comparison and tracking of simulation results and monitoring data for risk assessment. • The approach was validated by an integrated on-site practice in a large-scale bridge. Structural safety during construction is vital to engineering success of large scale bridges. However, difficulties in time-dependent structural modeling and data fragmentation of different engineering and management systems remain unresolved, hindering the plan, do, check, and adjust (PDCA) loop for structural safety management during bridge construction. In this paper, an integrated framework for closed-loop management of structural safety based on multisource data integration is presented. The proposed framework consists of a bridge safety information model (BrSIM), algorithms for data integration and semi-automatic time-dependent structural model generation, and methods for structural safety warning and assessment. The proposed BrSIM and algorithms integrate data related to 3D products, schedule, structural simulation and monitoring from various engineering systems, which covers the main data for structural safety management during construction. Meanwhile, automatic calculation and generation of static loads and constraints of a structural model based on 3D product information and monitoring data are also considered. Demonstration in the construction of a long-span bridge shows that with the proposed framework, it is possible to visualize the construction process, generate time-dependent structural models and simulate, monitor and assess the structural safety dynamically. Thus, the structural safety management loop is automated and fully closed. Furthermore, by tracking and simulating the changes of structural performance over time, and comparing the difference between simulation results and monitoring data, earlier detection and better evaluation of potential structural risks are achieved. Moreover, efficiency of information modeling and sharing is improved and effective management and decision-making are achieved with the proposed approach. [ABSTRACT FROM AUTHOR]