Structural Health Monitoring of large-scale bridges requires collection, storage and processing of large amounts of data, and must provide distributed concurrent access. In this paper we report on the progress of the design and implementation of a cyber-infrastructure system that is currently being field-tested on a long-span bridge in California and a short-span bridge in Michigan. This system provides remote access for sensor data acquisition systems, data analysis modules, and human operators. The implementation is based on an object-oriented data model description and makes extensive use of code generation to allow for the rapid development and continued improvement of the system. Currently the system provides storage of raw and processed sensor data, finite element models, traffic data, links to PONTIS data, reliability modeling data, and model-based analysis results. Apart from the ubiquitous read/write access, the system also includes an event system that allows data consumers to be triggered by the arrival of new data. In addition to essential backup/restore facilities, the system also includes import/export tools that can migrate data between versions, which is very useful in keeping pace with the continuous improvements that are being made in the design and implementation of the cyber-infrastructure system. The system also provides introspection, as the data model is made available by means of an inspector client interface, which allows the development of generic client tools that can dynamically discover the data model, and present a corresponding interface to the user. Currently available user-interfaces include an editor GUI application, and a read-only web application.