Long-Term Assessment of an Autonomous Wireless Structural Health Monitoring System at the New Carquinez Suspension Bridge

A dense network of sensors installed in a bridge can continuously generate response data from which the health and condition of the bridge can be analyzed. This approach to structural health monitoring the efforts associated with periodic bridge inspections and can provide timely insight to regions of the bridge suspected of degradation or damage. Nevertheless, the deployment of fine sensor grids on large-scale structures is not feasible using wired monitoring systems because of the rapidly increasing installation labor and costs required. Moreover, the enormous size of raw sensor data, if not translated into meaningful forms of information, can paralyze the bridge manager's decision making. This paper reports the development of a large-scale wireless structural monitoring system for long-span bridges; the system is entirely wireless which renders it low-cost and easy to install. Unlike central tethered data acquisition systems where data processing occurs in the central server, the distributed network of wireless sensors supports data processing. In-network data processing reduces raw data streams into actionable information of immediate value to the bridge manager. The proposed wireless monitoring system has been deployed on the New Carquinez Suspension Bridge in California. Current efforts on the bridge site include: 1) long-term assessment of a dense wireless sensor network; 2) implementation of a sustainable power management solution using solar power; 3) performance evaluation of an internet-enabled cyber-environment; 4) system identification of the bridge; and 5) the development of data mining tools. A hierarchical cyber-environment supports peer-to-peer communication between wireless sensors deployed on the bridge and allows for the connection between sensors and remote database systems via the internet. At the remote server, model calibration and damage detection analyses that employ a reduced-order finite element bridge model are implemented.