Integrated bridge management from 3D-model to network level

In bridge management, namely in complex networks like e.g. in urban areas, it is of equal importance to consider the condition of the individual building as well as the effects of maintenance measures on network level. In this paper we introduce a concept to integrate a 3D-model based management of single bridges with a maintenance schedule optimization on network level. The use of 3D-models in building management has some advantages to the mere textual based approach used in most common management systems: The orientation of building inspectors and maintenance planning engineers at navigation through the data is facilitated; the possibility of misplacements and misinterpretation of data is reduced, as all data is attached directly to a 3D representation of the building. Further the current condition (gained from inspection data), as well as prognoses for the future condition trends can be visualized on this 3D-model thus helping to identify weak points in the construction to be subject of special attention in later inspections. The condition prognosis obtained on building level is input data for finding an ideal schedule of maintenance on network level. This schedule is not only subject to the wish to keep all bridges in the network under good condition and therefore safe, but also to some additional considerations by the building manger: For example, the manager may have a limited budget for maintenance measures each year or may want to steady the amount of money spent on maintenance over the years. It may also be of interest to plan maintenance and thus involved (partial) road closures in such a way, that the impact on traffic flow is as low as possible. Additionally, there may be synergies with maintenance measures by other parties, e.g. streetcar operators, which should also be considered. All these considerations make the task of finding a good schedule a constrained multi-objective optimization problem which we solve using advanced heuristic approaches (e.g. ant colony optimization) which are presented in the paper.